jpad journal

AND option

OR option



J. Kaye1, P. Aisen2, R. Amariglio3, R. Au4, C. Ballard5, M. Carrillo6, H. Fillit7, T. Iwatsubo8, G. Jimenez-Maggiora2, S. Lovestone9, F. Natanegara10, K. Papp3, M.E. Soto11, M. Weiner12, B. Vellas13, and the EU/US CTAD Task Force


*EU/US/CTAD TASK FORCE: Sandrine Andrieu (Toulouse); Matthew Barton (Raleigh); Randall Bateman (Saint Louis); Monika Baudler (Basel); Joanne Bell (Willmington); Kevin Biglan (Indianapolis); Adam Boxer (San Francisco); Sasha Bozeat (Basel); Claudine Brisard (Issy les Moulineaux); Miroslaw Brys (Indianapolis); Marc Cantillon (Gilbert); Bill Chan (Beijing); Ivan Cheung (Woodcliff Lake); Min Cho (Woodcliff Lake); Julia Coelho (San Francisco); Shobha Dhadda (Woodcliff Lake); Daniel Di Giusto (Basel); Michael Donohue (San Diego); Rachelle Doody (Basel); John Dwyer (Washington); Michael Egan (North Wales); Rianne Esquivel (Malvern); Wendy Galpern (New Jersey); Harald Hampel (Woodcliff Lake); Jason Hassenstab (St Louis); David Henley (New Jersey); Joseph Herring (North Wales); Carole Ho (South San Francisco); Michael Irizarry (Woodcliff Lake); Keith Johnson (Boston); Geoffrey Kerchner (South San Francisco); Gene Kinney (South San Francisco); Shailaja Korukonda (Woodcliff Lake); Lynn Kramer (Woodcliff Lake); Jaren Landen (Cambridge); Ishani Landri (Woodcliff Lake); Lars Lannfelt (Uppsala); Valérie Legrand (Nanterre); Manoj Malhotra (Woodcliff Lake); Eric McDade (St Louis); Francisco Nogueira (South San Francisco); Gerald Novak (New Jersey); Gunilla Osswald (Stockholm); Susanne Ostrowitzki (South San Francisco); Amanda Paley (New York); Martin Rabe (Woodcliff Lake); Rema Raman (San Diego); Elena Ratti (Cambridge); Laurie Ryan (Bethesda); Stephen Salloway (Providence); Peter Schüler (Langen); Hiroshi Sekiya (Malvern); Jiong Shi (Las Vegas); Melanie Shulman (Cambridge); Eric Siemers (Zionsville); John Sims (Indianapolis); Kaycee Sink (South San Francisco); Reisa Sperling (Boston); Joyce Suhy (Newark); Jina Swartz (London); Pierre Tariot (Phoenix); Edmond Teng (South San Francisco); Jacques Touchon (Montpellier); Martin Traber (Basel); Andrea Vergallo (Woodcliff Lake); Judy Walker (Singapore); Jon Walsh (San Francisco); Alette Wessels (Indianapolis); Haichen Yang (North Wales); Wagner Zago (San Francisco); Kenton Zavitz (Cambridge).

1. Layton Aging and Alzheimer’s Disease Center, School of Medicine, Oregon Health and Science University, Portland, OR, USA; 3. Alzheimer’s Therapeutic Research Institute, University of Southern California, San Diego, CA, USA; 3. Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 4. Boston University, Boston, MA, USA; 5. University of Exeter, Exeter, UK; 6. Alzheimer’s Association, Chicago IL, USA; 7. Alzheimer Drug Discovery Foundation, New York, NY, USA; 8. University of Tokyo, Tokyo, japan; 9. Janssen Pharmaceutical Company, High Wycombe, UK; 10. Eli Lilly and Company, Indianapolis IIN, USA; 11. Toulouse University Hospital, Toulouse, France; 12. University of California, San Francisco CA, USA; 13. Gerontopole, INSERM U1027, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France

Corresponding Author: Jeffrey Kaye, Layton Aging and Alzheimer’s Disease Center, School of Medicine, Oregon Health and Science University, Portland, OR, USA,

Published online June 28, 2021,



The 2020 COVID-19 pandemic has disrupted Alzheimer’s disease (AD) clinical studies worldwide. Digital technologies may help minimize disruptions by enabling remote assessment of subtle cognitive and functional changes over the course of the disease. The EU/US Clinical Trials in Alzheimer’s Disease (CTAD) Task Force met virtually in November 2020 to explore the opportunities and challenges associated with the use of digital technologies in AD clinical research. While recognizing the potential of digital tools to accelerate clinical trials, improve the engagement of diverse populations, capture clinically meaningful data, and lower costs, questions remain regarding the stability, validity, generalizability, and reproducibility of digital data. Substantial concerns also exist regarding regulatory acceptance and privacy. Nonetheless, the Task Force supported further exploration of digital technologies through collaboration and data sharing, noting the need for standardization of digital readouts. They also concluded that while it may be premature to employ remote assessments for trials of novel experimental medications, remote studies of non-invasive, multi-domain approaches may be feasible at this time.

Key words: Alzheimer’s disease, clinical outcomes, digital tools, remote assessments.



The 2020 COVID-19 pandemic disrupted clinical trials worldwide as health care facilities became overwhelmed and lock-down conditions shuttered clinical trial sites, forcing both study staff and trial participants to stay home (1–3). Nearly every ongoing clinical study has felt the impact, although the degree of impact varies. Alzheimer’s disease (AD) studies have been acutely affected for multiple reasons, including: 1) the increased vulnerability to COVID-19 of people with or at risk of AD due to their advanced age and high prevalence of comorbid conditions; 2) the effects of even mild cognitive decline on the ability of participants to comply with the operational changes in how trials are conducted (e.g., social distancing from study staff and study partners, changes in study protocols); and 3) effects on clinical endpoints due to isolation, confusion, and behavioral and psychological symptoms of dementia (BPSD) such as agitation, apathy, and depression (1, 4). Other disruptions common to all clinical trials including those for AD interventions include: 1) reduced clinical capacity; 2) increased costs due to the need for increased personal protective equipment (PPE) and infrastructure for altered intervention delivery; 3) delays in ethics approvals as institutional review boards (IRBs) are swamped with COVID-19 related protocols; 4) reluctance of patients and participants to be in contact with clinics and health providers; and 5) suspension of patient screening and recruitment.
Study sponsors have responded to these challenges in different ways; for example, by mailing drugs to trial participants to ensure participants have access to interventions; limiting clinic visits through the establishment of alternative trial sites; allowing a brief hiatus from dosing; and implementing home infusions. Funders have also responded with innovations. In an analysis of their funding portfolio during the latter half of 2020, the Alzheimer’s Drug Discovery Foundation (ADDF) found that about 70% of grantees expected their studies to be delayed, and some requested no-cost extensions. Delays may harm not only budgets but data analyses as well, since data interpretation can be muddled due to long-term disruptions, missing data, and protocol deviations; with clinical studies more affected than preclinical studies. Observational studies have also been affected. For example, Swedish BioFINDER2 halted enrollment, and the Rush University aging cohort studies stopped performing autopsies due to lack of PPE. In response to these challenges, the Alzheimer’s Association introduced a Rapid Program in Dementia (RAPID) funding program to provide bridge funding for Association-funded research during the pandemic.
In March 2020, the U.S. Food and Drug Administration (FDA) issued guidelines on the conduct of clinical trials during the pandemic and updated this guidance in December 2020 (5). The guidance aims to protect the safety of trial participants while ensuring good clinical practice and trial integrity. It also recognized the likely use of remote assessments and new technologies for data acquisition.
In recent years, digital technologies that enable remote assessment and gather real-world granular data on a variety of cognitive and functional outcomes have received increased attention and interest in the AD drug development community (6). Recognizing the urgency of developing these technologies due to the pandemic, the EU/US CTAD Task Force met in November 2020 to address the topic, bringing together an international group of clinical investigators from academia and industry along with funders. They explored emerging digital technologies that are being used in clinical research as well as the challenges that have yet to be addressed to realize the potential of these technologies. First and foremost, they recognized that global challenges such as the COVID-19 pandemic require global solutions.


Opportunities and challenges to make trials more robust using digital technologies

Even if there were no pandemic, clinical trials as currently conducted are too slow and costly to adequately respond to a disease such as Alzheimer’s, which currently affects around 50 million people worldwide and for which there is no approved disease-modifying treatment (7). The difficulty of conducting clinical trials for AD treatments is exacerbated by the use of conventional outcome measures that are insensitive to change early in disease and sensitive to environmentally-induced variation. By capturing alternative outcome measures in real-world settings, digital devices have the potential to mitigate these problems, thereby expediting trials, lowering costs, and enabling more effective clinical trials. Digital tools can also promote inclusiveness and social justice, increase the pool of participants, decrease sample size, and increase trial effectiveness by improving the generalizability and representation of vulnerable populations. Smartphones hold particular promise in the realm of digital tools since they are among the most penetrating technologies available, with over three billion users worldwide (8) .
However, converting trial outcomes from in-person to remote measures is also associated with many challenges. Among the most critical aspects to be considered are usability and the reliability and fidelity of technology-related measures (9). Consumer-oriented devices and apps must be scaled for use in clinical trial settings and confounds introduced by remote assessment must be addressed. User training and support will be critical to reduce user errors. Technologies themselves should not drive their use in clinical trials; rather the use case should drive the technology.
In the AD space, digital tools have shown particular promise in capturing subtle cognitive and functional changes in preclinical AD. Yet there remain questions regarding their ability to serve as outcome measures due to questions around feasibility, reliability, and validity.


Implementing digital devices in AD clinical studies

As the COVID-19 pandemic has necessitated moving clinical trials from specialized centers to home-based virtual assessments using both passive and active measures, investigators are learning lessons that have the potential to accelerate all clinical trials for dementia. Indirect measures of behavior, function, and physiology are being explored as cognitive and functional endpoints for clinical trials by individual research groups as well as technology companies. For example, Elektra Labs has created the Atlas platform, which takes a data science approach to clinical trials through the application of a wide selection of connected sensors and physiological and behavioral markers.
Many research groups that have developed and/or implemented digital technologies in clinical studies presented data and plans to the Task Force, providing lessons learned as well as strategies for moving forward:


The Collaborative Aging Research Using Technology (CART) Initiative funded by the US National Institutes of Health and the Department of Veterans Affairs, has developed a multi-functional digital assessment platform that is technology agnostic, use case flexible, sharable, and scalable (10). The platform, an extension of the Oregon Center for Aging & Technology (ORCATECH) assessment system at Oregon Health and Science University has been deployed to 301 older adults for up to three years in four US regions. It enables unobtrusive, long-term assessment across multiple domains (e.g., mobility, cognitive function, sleep) through the use of many different sensors and connected devices (e.g., electronic pillboxes, bed mats, actigraphs, online reporting on laptops, tablets or smartphones). The platform is currently deployed by multiple groups in the United States and Canada, with plans to extend the platform to France (in the community-wide INSPIRE study based at Toulouse University Hospital) and Australia. Testing in two cohorts — a low-income housing cohort in Portland, Oregon, and an African-American Minority Aging Research Study (MARS) cohort in Chicago – gathered data spanning the time period of the pandemic demonstrating increased loneliness and decreased physical activity (Figure 1). In a proof-of-concept trial, the platform has also demonstrated feasibility for monitoring and predicting agitation among residents of memory care facilities (11), suggesting that it may be possible to demonstrate the efficacy of treatments for agitation with sensors that monitor life-space activity.

Figure 1. Pervasive Computing and Continuous Monitoring Demonstrate Effects of COVID-19 Pandemic on Loneliness and Physical Activity

In two separate cohorts, digital data streams provided by the CART ORCATECH platform demonstrated changes in mood (through online self-reported questionnaires) and physical activity (through actigraphy) that coincided with the imposition of stay-at-home orders because of the pandemic.



The public private partnership Remote Assessment of Disease and Relapse – AD (RADAR-AD), funded by the Innovative Medicines Initiative (IMI) has embraced an “internet of things” (IOT) approach similar to the CART/ORCATECH system to measure early decline in cognitive function using low friction digital devices across a range of challenges. A study is planned that will enroll 200 participants over age 50 including those with preclinical AD, MCI due to AD, mild-to-moderate AD, and age and gender matched controls. RADAR earlier launched a multi-site prospective cohort study to determine the feasibility, usability, and acceptability of remote measurement technology (RMT) for major depressive disorder (MDD) (12).
Remote digital measurements that may be used in RADAR-AD include adaptations of multiple instruments to be administered using a table or smartphone interface such as the Amsterdam-instrumental activities of daily living questionnaire (A-IADL-Q) (13) to assess orientation, planning skills, and memory; and the Patient-Reported Outcomes Measurement Information Systems (PROMIS) general health measures to assess quality of life, disabilities, pain, and fatigue (14).
The RADAR base app, using a smartphone and passive sensors, will assess mobility, social communication, time at home, general activity, and sleep duration and quality. A residential movement detector will also assess gait and mobility at home and time spent in different areas of the home. The Mezurio smartphone app will capture high-frequency data related to cognition, including sleep, planning, memory, mood and anxiety, voice and speech, and typing and swiping (15).

Harvard Aging Brain Study (HABS)

In its search for sensitive measures of subtle cognitive decline in preclinical AD, The Harvard Aging Brain Study (HABS) developed an iPad-based Computerized Cognitive Composite (C3) and subsequently a smartphone-based tool for remote assessment of memory. The C3 on the iPAD comprises 3 components: a Behavioral Pattern Separation task, a Face-Name Associative Memory Exam, and the Cogstate Brief Battery. Development of the C3 included its use in the A4 study, a secondary prevention trial in clinically normal older adults with abnormal levels of amyloid who were randomized to receive placebo or solanezumab. This initial study showed that when the drug was administered in screening period, the amyloid-positive group performed worse on C3 compared to the Preclinical Alzheimer Cognitive Composite (PACC) (16, 17). The C3 has also been shown to provide reliable data when participants complete the assessments at home on an iPad (18). The remote use of C3 allows for more frequent assessments, which provide new information about subtle cognitive changes (19).
HABS has also developed the Boston Remote Assessment for Neurocognitive Health (BRANCH), which combines two paired associative learning tasks, an associated memory test, and a continuous visual recognition task accessible on a smartphone. In a recent study, BRANCH was shown to be correlated with scores on the paper-based preclinical Alzheimer’s cognitive composite (PACC5) (20).
HABS investigators have also been using BRANCH to assess learning curves using data from clinically normal older adults who completed 5 consecutive days of BRANCH. Early evidence suggests that a steeper learning curve is associated with better performance on the PACC, although the results were not statistically significant. Nonetheless, capturing learning curves using a smartphone over several days appears to be feasible and valid and may be useful to prescreen potential trial participants for biomarker status and identify those at greater risk for imminent cognitive decline and most likely to benefit from treatment. Cognitive learning curves may also serve as a paradigm for exploring psychopharmacokinetic profiles of medications to assess for efficacy over relatively short time periods.

The Trial-Ready Cohort for Preclinical and Prodromal Alzheimer’s Disease (TRC-PAD)

The Trial-Ready Cohort for Preclinical and Prodromal Alzheimer’s Disease (TRC-PAD) was established to accelerate drug development for AD by ensuring timely recruitment of targeted individuals into optimally designed preclinical and prodromal trials. TRC-PAD gathers multiple types of data including digital data from wearables and sensors through its informatics platform (TRC-PAD IP). In concert with TRC-PAD, the Alzheimer’s APT Webstudy registry was established along with an analytics platform, a referral management system, and a data management system (21).
The APT Webstudy has been working to optimize digital smartphone-based tools to assess cognition, such as the Cogstate Brief Battery. The study plans to implement this tool in the AHEAD 3-45 secondary prevention trials in early preclinical and preclinical AD and in the CT1812 trial in early AD (22). APT is also exploring alternative remote testing sensitive to amyloid elevation.

WHO – Integrated Care for Older People (ICOPE) Monitor

In its World Report on Healthy Ageing, the World Health Organization (WHO) redefined healthy aging in terms of functional ability, which is a product of an individual’s intrinsic capacity (IC) interacting with the environment (23). In order to implement integrated care for older people (ICOPE), an international partnership of researchers have developed a tool for assessing IC (24) as well as a mobile app available on smartphones that enable self-assessment by individuals to provide health care professionals with data on intrinsic capacity (25). Initial results from more than 5300 older individuals in the ICOPE Digital Cohort report high levels of IC impairments in multiple domains, including in vision, hearing, psychological health, and cognition. Recognizing that additional cognitive assessments are needed, ICOPE data will be combined with the Alzheimer’s Prevention Trial (APT) Webstudy data beginning in 2021. The ICOPE integrative approach from WHO provides the opportunity to detect memory complaints or decline, taking into consideration declines in other ICs, such as vision, hearing, mood, nutrition.

Brain Health Registry

With a goal of improving clinical trial recruitment, the Brain Health Registry (BHR) was established in 2014 as a website and online registry that collects digital data from potential clinical trial participants and study partners through self-report questionnaires and neuropsychological tests. Among the over 70,000 individuals enrolled in BHR, there have been over 106,000 referrals and 5,000 enrolled in other clinical studies. Through co-enrollment, BHR and clinical trial data can be linked.
BHR is also involved in methods development and validation of new online methods, including an electronic Clinical Dementia Rating (CDR) scale, a digital financial capacity instrument, and other digital assessment tools. In support of blood biomarkers studies, BHR has also collaborated with Quest Diagnostics to scale up blood collection using remote identification through the registry.

Other Studies Incorporating Digital Assessment Tools

Many other clinical research programs are exploring using digital measures of cognition and function (e.g., web- or mobile- device based cognitive testing, automated speech and language assessment, or use of sensor-embedded home-based device responsives). For example, the Framingham Heart Study (FHS) is developing a smartphone platform with multiple applications including digital voice. Voice (spoken utterances and language) is measured through device agnostic technologies and FHS research shows that linguistic and acoustic measures can be used as biomarkers of cognitive function (26–28). However, while speech and language assessments are likely to be very sensitive, additional work is needed to determine the best way to capture speech samples.
The Deep and Frequent Phenotyping (DFP) study in the United Kingdom has also incorporated digital measures from wearables and smartphones to capture highly granular data on gait, memory, navigational ability, and other measures that may serve as proxies for cognition and function (29). The study aims to identify an optimal set of markers for patient stratification.


Potential benefits and challenges associated with incorporating digital tools in trials

Digital devices offer the potential to expedite clinical trials, reduce the sample sizes or duration of studies, lower the cost of acquiring and analyzing data, enable the recruitment of more diverse and representative populations, reduce the need for specialized raters, reduce rater burden, and increase the clinical meaningfulness of outcome data collected to determine efficacy of an intervention. During a pandemic, digital tools also may provide the means for remote assessments, which could mitigate problems associated with closure of clinical trial sites and the safety concerns associated with in-person assessments and interventions.
Combining cognitive and functional endpoints in clinical trials is well established through the use of questionnaires in symptomatic stages of disease; and in preclinical disease, reports from both participants and study partners regarding high-level functional activities appear to provide valuable information on disease trajectory. Pairing digital cognitive measures with digital functional assessments may offer increased sensitivity in preclinical disease. The HABS team has shown that functional inventories can be administered remotely through the use of REDCap software, saving time and reducing data errors without sacrificing data quality.
The technologies employed in various digital devices provide additional potential advantages:
• They allow data to be time stamped, which enables assessment of everyday function.
• They can improve participant compliance by providing study participants and partners with reminders to take medications and engage in study-related activities.
• Audio and video recordings can provide a level of biometric validation to ensure that study participants themselves (rather than study partners) are providing data.
• Passive sensing of activities such as driving and computer use have demonstrated unique personalized patterns that can be captured using relatively simple algorithms.
• Passive sensors reduce participant burden, thus potentially reducing drop-outs and producing outcomes that are easily understandable and relevant in a real-world context.
• Meta-data such as time taken to complete a questionnaire or remember to take medication on time may also reflect changes in episodic memory, a key domain affected in AD.
• Passive assessment of sleep may prove useful since disrupted sleep has been shown to be a mediator of amyloid deposition in many studies.

To achieve the promise of digital devices will require addressing many challenges including those related to reliability, validation, user acceptability, regulatory approval, privacy and security. While regulators have indicated a willingness to consider digital data, much work remains to be completed before digital measures are accepted as primary or secondary outcomes. Indeed, the FDA issued guidance on the conduct of trials during a pandemic which includes increased flexibility around trial procedures intended to maintain participant safety, but less flexibility around efficacy measures. Task Force members also questioned whether regulators would accept remotely acquired data where there is no assurance that study participants and not study partners completed questionnaires or cognitive outcome assessments.
While digital registries may enable outreach to otherwise under-represented populations, engagement and retention of trial participants through digital means alone may be insufficient. Indeed, studies have reported high compliance rates that they attribute to personalized rather than anonymous digital engagement of study participants (30).
Transitioning cognitive assessments to smartphones may enable more frequent and sensitive measurement among large numbers of participants in a short period of time. However, smartphones may introduce additional challenges with regard to data fidelity and security. Other technological challenges related to digital data include the difficulty of analyzing continuous data and deconvoluting diverse data captured by devices. Addressing these challenges will be advanced by wider collaboration with the computer science and data science community.
Digital readouts are diverse and non-standardized, which results in impediments to data aggregation and sharing of data, both of which are essential to advance the field. Collaboration among technology developers is needed to ensure that raw data from which readouts on different devices are derived will be made available to the research community. Privacy concerns represent an additional roadblock to data sharing. Federated data management platforms have been developed that provide researchers with access to multiple types of data from multiple sources (31). The Alzheimer’s Disease Data Initiative has also established the AD Workbench (ADWB) platform to promote data, information, and tool sharing across the AD research community (32), although no digital biomarker data are yet incorporated in this relatively new effort.


Moving Forward

Despite the challenges of conducting clinical studies during the pandemic, the AD clinical research community remains committed to advancing current studies and launching new studies. Indeed, Task Force participants suggested that the COVID-19 experience may fuel preparations for better clinical trials that mitigate both the risks associated with the pandemic and more common complications of clinical trials for example by broadening the reach of recruitment efforts.
Among the possibilities discussed by the Task Force was the potential to design completely digital or hybrid trials combining virtual/telemedicine and physical visits. For example, telemedicine consultations in the pre-screening phase of a trial could help verify inclusion and exclusion criteria, provide information to potential trial participants, and obtain electronic consent. Virtual visits could also be used for trial components that do not require technically invasive procedures. The INSPIRE study team at Toulouse University Hospital tested this strategy following the first COVID lockdown in France and found a high rate of acceptance among participants (33).
Digital devices may provide new analytical methods as well, for example, by enabling assessment of intraindividual change in a high frequency way. Digital tools implemented during the pandemic may also provide increased understanding to the effects of daily stress, anxiety, and poor sleep on cognitive outcomes. They may also enable continued assessment of cognitive function in participants post-infection with COVID-19.
To ensure that clinical studies continue through the pandemic and its aftermath, the Task Force recommended that regulators and institutional review boards provide guidance yet maintain flexibility with regard to changes in applications/devices and statistical plans and allow for virtual and home-based visits and remote monitoring. Developers must consider alternate or additional trial sites, and funders should expect unanticipated modifications to protocols and funding needs.
The Task Force concluded its session with a discussion of the field’s readiness to conduct remote clinical trials. Task Force members urged caution in employing remote assessments for trials of experimental medications; the digital measures at this time may be most productively embedded in parallel with conventional measures in early-stage Phase 1 and 2 trials, demonstrating their fidelity or even superiority to current commonly used clinical outcome measures. Enthusiasm was expressed for using digital technologies in remote studies of non-invasive interventions such as studies of vitamins, exercise, and mind-body or multidomain approaches. Trials of repurposed drugs for which safety data exists may also be feasible as entirely remote trials. For example, because of the COVID-19 pandemic, the Alzheimer’s Disease Cooperative Study (ADCS) is attempting to conduct the PEACE-AD (prazosin for disruptive agitation in Alzheimer’s disease) trial (34) as a remote-only trial.


Acknowledgements: The authors thank Lisa J. Bain for assistance in the preparation of this manuscript.

Conflicts of interest: The Task Force was partially funded by registration fees from industrial participants. These corporations placed no restrictions on this work. Dr. Kaye reports grants from Merck, Eisai, Genentech, and Abbvie; and serves on a Data Safety Monitoring Committee at Eli Lilly. Dr. Aisen reports grants from Janssen, Eli Lilly, Eisai, NIA, the Alzheimer’s Association, and FNIH; and consulting fees from Biogen, Roche, Merck, Abbvie, Lundbeck, Proclara, and Immunobrain Checkpoint. Dr. Amariglio declares there are no conflicts. Dr. Au is on the scientific advisory board of Signant Health and a consultant to Biogen. Dr. Ballard reports grants from Arcadia pharmaceutical company and Lundbeck; personal fees from Arcadia, Lundbeck, Roche, Otsuka, Novartis, Eli Lilly, Suven, Sunovion, ADDEX, and Exciva; personal fees and other from Synexus, Novo Nordisk; and consulting fees from Biogen. Dr. Carillo declares there are no conflicts. Dr. Fillit declares there are no conflicts. Dr. Iwatsubo declares there are no conflicts. Dr. Jimenez-Maggiora declares there are no conflicts. Dr. Lovestone is an employee of Janssen Medical Ltd and a co-founder of Akrivia Health Ltd . Dr. Natanegara is an employee of Eli Lilly and Company. Dr. Papp declares there are no conflicts. Dr. Soto declares payment as consultant or advisor from Avanir, Acadia. Dr. Weiner reports grants from Siemens, Biogen, and Johnson & Johnson, NIH, California Department of Health, University of Michigan, Hillblom Foundation, Alzheimer’s Association, State of California, Kevin and Connie Shanahan, GE, VUMc, American Catholic University, The Stroke Foundation, and the Veteran’s Administration; has served on advisory boards for Eli Lilly, Cerecin/Accera, Roche, Alzheon Inc., Merck Sharp & Dohme Corp, Nestle/Nestec; and has consulted with or acted as a speaker to Cerecin/Accera Inc., BioClinica, Nestle/Nestec, Genentech, FUJIFILM-Toyama Chemical, and T3D Therapeutics; and holds stock options with Alzheon Inc, Alzeca, and Anven. Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen, outside the submitted work.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Morris MC, Tangney CC, Wang Y, et al. MIND diet slows cognitive decline 1. van Dorn A. COVID-19 and readjusting clinical trials. The Lancet. 2020 Aug 22;396(10250):523–4.
2. Fleming TR, Labriola D, Wittes J. Conducting Clinical Research During the COVID-19 Pandemic: Protecting Scientific Integrity. JAMA. 2020 Jul 7;324(1):33–4.
3. Sathian B, Asim M, Banerjee I, Pizarro AB, Roy B, van Teijlingen ER, et al. Impact of COVID-19 on clinical trials and clinical research: A systematic review. Nepal J Epidemiol. 2020 Sep;10(3):878–87.
4. Brown EE, Kumar S, Rajji TK, Pollock BG, Mulsant BH. Anticipating and Mitigating the Impact of the COVID-19 Pandemic on Alzheimer’s Disease and Related Dementias. Am J Geriatr Psychiatry. 2020 Jul;28(7):712–21.
5. Food and Drug Administration. Conduct of Clinical Trials of Medical Products During the COVID-19 Public Health Emergency Guidance for Industry, Investigators, and Institutional Review Boards. U.S. Department of Health and Human Services,; 2020.
6. Gold M, Amatniek J, Carrillo MC, Cedarbaum JM, Hendrix JA, Miller BB, et al. Digital technologies as biomarkers, clinical outcomes assessment, and recruitment tools in Alzheimer’s disease clinical trials. Alzheimers Dement (N Y). 2018;4:234–42.
7. World Health Organization. Dementia [Internet]. 2020 [cited 2021 Jan 6]. Available from:
8. Statista. Number of smarphone users worldwide from 2016 to 2021 [Internet]. Statista. 2020 [cited 2021 Jan 26]. Available from:
9. Lindauer A, Seelye A, Lyons B, Dodge HH, Mattek N, Mincks K, et al. Dementia Care Comes Home: Patient and Caregiver Assessment via Telemedicine. Gerontologist. 2017 Oct 1;57(5):e85–93.
10. Beattie Z, Miller LM, Almirola C, Au-Yeung W-TM, Bernard H, Cosgrove KE, et al. The Collaborative Aging Research Using Technology Initiative: An Open, Sharable, Technology-Agnostic Platform for the Research Community. Digit Biomark. 2020;4(Suppl 1):100–18.
11. Au-Yeung W-TM, Miller L, Beattie Z, Dodge HH, Reynolds C, Vahia I, et al. Sensing a problem: Proof of concept for characterizing and predicting agitation. Alzheimers Dement (N Y). 2020;6(1):e12079.
12. Matcham F, Barattieri di San Pietro C, Bulgari V, de Girolamo G, Dobson R, Eriksson H, et al. Remote assessment of disease and relapse in major depressive disorder (RADAR-MDD): a multi-centre prospective cohort study protocol. BMC Psychiatry. 2019 Feb 18;19(1):72.
13. Stringer G, Leroi I, Sikkes SAM, Montaldi D, Brown LJE. Enhancing ‘meaningfulness’ of functional assessments: UK adaptation of the Amsterdam IADL questionnaire. International Psychogeriatrics. undefined/ed;1–12.
14. Tatsuoka C, DeMarco L, Smyth KA, Wilkes S, Howland M, Lerner AJ, et al. Evaluating PROMIS Physical Function Measures in Older Adults at Risk for Alzheimer’s Disease. Gerontol Geriatr Med [Internet]. 2016 Sep 5 [cited 2021 Jan 8];2. Available from:
15. Lancaster C, Koychev I, Blane J, Chinner A, Wolters L, Hinds C. Evaluating the Feasibility of Frequent Cognitive Assessment Using the Mezurio Smartphone App: Observational and Interview Study in Adults With Elevated Dementia Risk. JMIR Mhealth Uhealth [Internet]. 2020 Apr 2 [cited 2021 Jan 8];8(4). Available from:
16. Papp KV, Rentz DM, Maruff P, Sun C-K, Raman R, Donohue MC, et al. The Computerized Cognitive Composite (C3) in A4, an Alzheimer’s Disease Secondary Prevention Trial. J Prev Alzheimers Dis. 2021;8(1):59–67.
17. Papp KV, Rentz DM, Orlovsky I, Sperling RA, Mormino EC. Optimizing the preclinical Alzheimer’s cognitive composite with semantic processing: The PACC5. Alzheimers Dement (N Y). 2017 Nov;3(4):668–77.
18. Rentz D, Dekhtyar D, Sherman J, Burnham S, Blacker D, Aghjayan S, et al. The Feasibility of At-Home iPad Cognitive Testing For Use in Clinical Trials. 2015 [cited 2021 Jan 8]; Available from:
19. Samaroo A, Amariglio RE, Burnham S, Sparks P, Properzi M, Schultz AP, et al. Diminished Learning Over Repeated Exposures (LORE) in preclinical Alzheimer’s disease. Alzheimers Dement (Amst). 2020;12(1):e12132.
20. Papp KV, Samaroo AH, Chou H-CL, Buckley RF, Rentz D, Sperling RA, et al. Repeated memory-based assessments: Implications for clinical trials and practice. Alzheimer’s & Dementia. 2020;16(S6):e038143.
21. Jimenez-Maggiora GA, Bruschi S, Raman R, Langford O, Donohue M, Rafii MS, et al. TRC-PAD: Accelerating Recruitment of AD Clinical Trials through Innovative Information Technology. J Prev Alzheimers Dis. 2020;7(4):226–33.
22. Grundman M, Morgan R, Lickliter JD, Schneider LS, DeKosky S, Izzo NJ, et al. A phase 1 clinical trial of the sigma-2 receptor complex allosteric antagonist CT1812, a novel therapeutic candidate for Alzheimer’s disease. Alzheimers Dement (N Y). 2019 Jan 23;5:20–6.
23. Beard JR, Officer A, de Carvalho IA, Sadana R, Pot AM, Michel J-P, et al. The World report on ageing and health: a policy framework for healthy ageing. Lancet. 2016 May 21;387(10033):2145–54.
24. Takeda C, Guyonnet S, Sumi Y, Vellas B, Araujo de Carvalho I. Integrated Care for Older People and the Implementation in the INSPIRE Care Cohort. J Prev Alzheimers Dis. 2020;7(2):70–4.
25. Sanchez-Rodriguez D, Annweiler C, Gillain S, Vellas B. Implementation of the Integrated Care of Older People (ICOPE) App in Primary Care: New Technologies in Geriatric Care During Quarantine of COVID-19 and Beyond. J Frailty Aging. 2020 May 6;1–2.
26. Thomas JA, Burkhardt HA, Chaudhry S, Ngo AD, Sharma S, Zhang L, et al. Assessing the Utility of Language and Voice Biomarkers to Predict Cognitive Impairment in the Framingham Heart Study Cognitive Aging Cohort Data. J Alzheimers Dis. 2020;76(3):905–22.
27. Thomas NWD, Beattie Z, Marcoe J, Wright K, Sharma N, Mattek N, et al. An Ecologically Valid, Longitudinal, and Unbiased Assessment of Treatment Efficacy in Alzheimer Disease (the EVALUATE-AD Trial): Proof-of-Concept Study. JMIR Res Protoc. 2020 May 27;9(5):e17603.
28. Lin H, Karjadi C, Ang TFA, Prajakta J, McManus C, Alhanai TW, et al. Identification of digital voice biomarkers for cognitive health. Exploration of Medicine. 2020 Dec 31;1(6):406–17.
29. Koychev I, Lawson J, Chessell T, Mackay C, Gunn R, Sahakian B, et al. Deep and Frequent Phenotyping study protocol: an observational study in prodromal Alzheimer’s disease. BMJ Open. 2019 Mar 23;9(3):e024498.
30. Sano M, Egelko S, Zhu CW, Li C, Donohue MC, Ferris S, et al. Participant satisfaction with dementia prevention research: Results from Home-Based Assessment trial. Alzheimers Dement. 2018;14(11):1397–405.
31. Toga AW, Bhatt P, Ashish N. Global data sharing in Alzheimer’s disease research. Alzheimer Dis Assoc Disord. 2016;30(2):160–8.
32. Alzheimer’s Disease Data Initiative. AD Workbench [Internet]. ADDI. [cited 2021 Jan 11]. Available from:
33. Takeda C, Guyonnet S, Ousset PJ, Soto M, Vellas B. Toulouse Alzheimer’s Clinical Research Center Recovery after the COVID-19 Crisis: Telemedicine an Innovative Solution for Clinical Research during the Coronavirus Pandemic. J Prev Alzheimers Dis. 2020;7(4):301–4.
34. Alzheimer Disease Cooperative Study. PEACE-AD [Internet]. ADCS PEACE-AD. [cited 2021 Jan 11]. Available from:




M. Senda1, K. Ishii2, K. Ito3, T. Ikeuchi4, H. Matsuda5, T. Iwatsubo6, A. Iwata7, R. Ihara7, K. Suzuki8, K. Kasuga4, Y. Ikari1,9, Y. Niimi6, H. Arai10, A. Tamaoka11, Y. Arahata3, Y. Itoh12, H. Tachibana13, Y. Ichimiya14, S. Washizuka15, T. Odawara16, K. Ishii17, K. Ono18, T. Yokota19, A. Nakanishi20, E. Matsubara21, H. Mori12, H. Shimada12


1. Kobe City Medical Center General Hospital, Japan; 2. Tokyo Metropolitan Institute of Gerontology, Japan; 3. National Center for Geriatrics and Gerontology, Japan; 4. Niigata University, Japan; 5. National Center of Neurology and Psychiatry, Japan; (currently, Southern Tohoku Drug Development and Cyclotron Research Center, Japan); 6. The University of Tokyo, Japan; 7. The University of Tokyo, Japan; (currently, Tokyo Metropolitan Geriatric Hospital, Japan); 8. The University of Tokyo, Japan; (currently, National Defense Medical College, Japan); 9. Osaka University, Japan; 10. Tohoku University, Japan; 11. University of Tsukuba, Japan; 12. Osaka City University, Japan; 13. Kobe University, Japan; 14. Juntendo Tokyo Koto Geriatric Medical Center, Japan; 15. Shinshu University, Japan; 16. Yokohama City University, Japan; 17. Kindai University, Japan; 18. Showa University, Japan; 19. Tokyo Medical and Dental University, Japan; 20. Osaka City Kosaiin Hospital, Japan; 21. Oita University, Japan

Corresponding Author: Michio Senda, Division of Molecular Imaging Research Kobe City Medical Center General Hospital (KCGH), 2-1-1 Minatojima-Minamimachi, Chuo-ku, Kobe 650-0047 Japan, E-mail:, Phone: 81-78-304-5212, Fax: 81-78-304-5201.



BACKGROUND: PET (positron emission tomography) and CSF (cerebrospinal fluid) provide the “ATN” (Amyloid, Tau, Neurodegeneration) classification and play an essential role in early and differential diagnosis of Alzheimer’s disease (AD).
OBJECTIVE: Biomarkers were evaluated in a Japanese multicenter study on cognitively unimpaired subjects (CU) and early (E) and late (L) mild cognitive impairment (MCI) patients.
MEASUREMENTS: A total of 38 (26 CU, 7 EMCI, 5 LMCI) subjects with the age of 65-84 were enrolled. Amyloid-PET and FDG-PET as well as structural MRI were acquired on all of them, with an additional tau-PET with 18F-flortaucipir on 15 and CSF measurement of Aβ1-42, P-tau, and T-tau on 18 subjects. Positivity of amyloid and tau was determined based on the positive result of either PET or CSF.
RESULTS: The amyloid positivity was 13/38, with discordance between PET and CSF in 6/18. Cortical tau deposition quantified with PET was significantly correlated with CSF P-tau, in spite of discordance in the binary positivity between visual PET interpretation and CSF P-tau in 5/8 (PET-/CSF+). Tau was positive in 7/9 amyloid positive and 8/16 amyloid negative subjects who underwent tau measurement, respectively. Overall, a large number of subjects presented quantitative measures and/or visual read that are close to the borderline of binary positivity, which caused, at least partly, the discordance between PET and CSF in amyloid and/or tau. Nine subjects presented either tau or FDG-PET positive while amyloid was negative, suggesting the possibility of non-AD disorders.
CONCLUSION: Positivity rate of amyloid and tau, together with their relationship, was consistent with previous reports. Multicenter study on subjects with very mild or no cognitive impairment may need refining the positivity criteria and cutoff level as well as strict quality control of the measurements.

Key words: Alzheimer’s disease, PET, CSF biomarker, amyloid, tau.




Early and differential diagnosis of Alzheimer’s disease (AD) has been drawing more and more attention these days as the target population of the therapeutic trials has shifted toward the early phases of the AD continuum. Biomarkers including PET, MRI and cerebrospinal fluid(CSF)/plasma play an essential role in such early phases, where clinical manifestation and behavioral findings are limited. Jack et al (1) extracted three markers, i.e., amyloid (A), tau (T) and neurodegeneration (N), and proposed the “ATN” classification for differential diagnosis of AD continuum. PET provides imaging and quantification of amyloid and tau deposition as well as neurodegeneration evaluable with 18F-fluorodeoxyglucose(FDG)-PET. Amyloid and tau can also be evaluated with CSF sampling, and recently with plasma as well, and MRI volumetry has also been used as a marker of neurodegeneration.
In Japan, a large-scale prospective observational study called J-ADNI (Japanese Alzheimer’s Disease Neuroimaging Initiative) was completed (2), in which a total of 537 subjects were enrolled, comprising 154 cognitively unimpaired subjects (CU), 234 MCI and 149 AD patients.
Then, a new version of J-ADNI was designed by the same group, named “AMED Preclinical AD Study”, which focused on CU and MCI and acquired amyloid-PET and FDG-PET on all subjects. Part of the subjects also underwent a tau-PET scan and/or a CSF sampling. The objective of the study was to evaluate PET and MRI images and CSF biomarkers in CU and MCI subjects in Japan, compare those biomarkers between modalities to explore their reliability and usefulness in such early-phase subjects, and obtain a rough idea of the fractions of ATN-based classifications. This report summarizes the results of the study that was recently completed.




The study was a non-randomized prospective observational study, and was designed and conducted in accordance with the ethical principles as proclaimed in the Declaration of Helsinki. The study protocol was first approved by Ethical Committee of Osaka City University Graduate School of Medicine (site of leading PI) and registered as UMIN000019926, and was later re-approved by Osaka City University Hospital Certified Review Board when the Japanese Law on Clinical Research was enacted, and was registered as jRCTs051180239. The protocol was also approved by each participating site according to the Japanese regulations and ethics guidance. The tau-PET portion of the study was designed as a nominally separate add-on study when tau-PET became available, though limited, later in the course of the research project, and was approved and registered as jRCTs051190065.
The subjects were enrolled at a total of 14 clinical sites and consisted of 26 CU subjects and 12 MCI patients (7 early (E) MCI and 5 late (L) MCI as classified below) based on the neuropsychological tests.
Written informed consent was obtained from each subject and the study partner such as a family member of the subject.
The CU subjects were 65-82 years old without any memory problem and CDR-J=0. The MMSE-J score was 25-30 (higher than or equal to 24), and the delayed recall score of WMS-R logical memory (WMS-R LM II) ranged from 4 to 19 except for one subject (See footnote of Table 1).
The MCI subjects were 65-84 years old with objective persistent memory impairment reported by the study partner and CDR-J=0.5 with memory box score being 0.5 or higher. The MMSE-J score was 27-30 (higher than or equal to 24). The WMS-R LM II was used to classify the MCI subjects into EMCI (WMS-R LM II = 3-6, 5-9, 9-11) and LMCI (≤2, ≤4, ≤8) depending on the educational years (0-7, 8-15, ≥16 years, respectively), and was 6-18 for EMCI and 0-8 for LMCI.

CSF measurements and genotyping

CSF was collected from 18 subjects by lumbar puncture and stored in polypropylene tubes at -80℃ until biochemical analysis. CSF concentration of Aβ1-42 was analyzed using V-PLEX Aβ Peptide Panel 1 kit with MESO QuickPlex SQ120 (MesoScale Discovery, Rockville, MD). CSF phosphorylated tau (P-tau) and total tau (T-tau) were measured using commercially available ELISA kits, INNOTEST hTAU and PHOSPHO-TAU (181P) (Fujirebio Europe, Belgium), respectively, according to the manufacturer’s instructions. Stability of the results was monitored in the Alzheimer’s Association QC program. Cutoff values (Aβ42<378.7 pg/mL, P-tau>29.1 pg/mL, and T-tau>88.8 pg/mL) that best discriminated PiB-PET positive AD patients from PiB-PET negative CU subjects were determined using independent J-ADNI cohort (2). Because the CSF assays used in this study were different from those used in J-ADNI study, calibration between two assays were performed.
APOE genotyping (rs429358 and rs7412) was performed by Taq-Man based assay using blood samples.

MRI imaging

The brain MRI was acquired for each subject using a 3-Tesla or 1.5-Tesla scanner. The structural 3D-T1 images (MP-RAGE or IR-SPGR) were analyzed with FreeSurfer (Ver. 6.0) to measure the regional cerebral gray-matter volumes. Because the absolute volumetry depends on the version of the software and other conditions, the regional atrophy of the subject was derived as z-score using mean and SD of the baseline scan for the 26 CU subjects of this study. The volume of 8 regions in the temporal lobe (right and left entorhinal cortex, parahippocampal gyrus, hippocampus, and amygdala) were summed up and the z score was derived as a measure of the temporal lobe atrophy for each subject.

PET image acquisition

All subjects underwent an amyloid-PET and an FDG-PET. Each PET imaging site, together with the PET camera, was qualified, in which the reconstruction parameters were determined for each PET camera so that all the PET cameras satisfied the image quality criteria with the Hoffman 3D brain phantom and the uniform cylindrical phantom (3).
For amyloid PET, either 11C-PiB (PiB), 18F-florbetapir (FBP) or 18F-flutemetamol (FMM) was used for 23, 13, and 2 subjects, respectively. The injection activity was 555MBq, 370MBq, 185MBq, the uptake time (start of emission scan post injection) was 50min, 50min, 90min, and the scan duration was 20min, 20min, 30min, for PiB, FBP, and FMM, respectively.
For the FDG scan, after at least 4 hours of fasting, the subject was administered with 185 MBq of 18F-FDG in a quiet, dimly lit room while resting in a reclining chair or bed, and the subject remained in the condition until several minutes before the start of the scanning session. The PET emission data was acquired for 30 minutes starting at 30 minutes post injection.
Tau-PET was performed with 18F-flortaucipir (FTP) on 15 subjects. Because tau-PET was not ready until late in the course of the research project, the time span from amyloid to tau-PET ranged from 1.0 to 2.0 (mean 1.56) years. The subject was administered with 240.5 MBq of FTP and a 30 min emission scan started 75 min post injection.
No adverse effects were observed at the PET scans of this study.

PET image analysis

The amyloid PET images were binary interpreted visually in a blind manner by the readers who were qualified for this study, and the adjudicator (K.I.) confirmed them. The PiB images were interpreted visually using the criteria adopted in J-ADNI (4), and the FMM and FBP images were interpreted with each vendor’s criteria.
As a quantitative analysis of the amyloid PET, mean cortical standardized uptake value ratio (mcSUVR) of PiB images was computed using the cerebellar cortex as a reference based on the method of J-ADNI, and the cutoff value of 1.5 was used to determine the quantitative positivity (4). The FMM images were analyzed with CORTEX ID (GE Healthcare) to derive mcSUVR using the pons as a reference, for which the cutoff value of 0.58 was used for the quantitative positivity (5). The FBP images were analyzed with MIMneuro (MIM Software) to derive mcSUVR using the whole cerebellum as a reference, for which the cutoff value of 1.10 was used for the quantitative positivity (6).
The FDG images together with the semiquantitative 3D-SSP results were visually interpreted by three independent readers followed by a consensus read in the same way as J-ADNI (7), and the images were classified into N1 (normal), N2 (reflecting atrophy), N3, P1 (AD pattern), P2 (FTD pattern), P3, and P1+ (DLB pattern) (8). No one presented N3 or P3 in this study. The DLB pattern criteria was interpreted in a broader sense to include cases with occipital hypometabolism extending to neighboring areas even if typical temporoparietal hypometabolism was not observed. The FDG images were also quantified with AD t-sum (9) using the module PALZ in the PMOD software package (Ver. 3.2; PMOD Technologies, Zurich, Switzerland), which were then converted into PET score [10] that reflects the severity of temporoparietal hypometabolism (AD pattern).
The FTP images were interpreted and classified into AD negative, AD+ and AD++, according to the vendor’s criteria that regards cortical uptake except anterior temporal as AD-related ( The FTP-PET was also analyzed with MUBADA-PERSI method to derive SUVR over the area affected by AD process (posterior temporal, occipital, parietal and part of frontal cortex) with white matter as a reference (11, 12).

Follow up

Whenever possible, each subject was followed up every year with a general clinical interview with neuropsychological tests, an MRI scan, and an FDG-PET scan. As a result, 1-year follow-up data were acquired on 32 subjects, and 2-year follow-up on 5 subjects.

Statistical methods

Because the number of subjects was small, descriptive results were presented in general. Proportion of positivity was compared between groups using chi-square tests, in which EMCI and LMCI were combined to increase the number of observations. Statistical tests were also performed on the Pearson correlation coefficient between two variables.



Findings of each subject

Table 1 describes findings of each subject as classified according to the ATN concept. In this study, amyloid (A) was interpreted as positive (A+) when either PET or CSF Aβ was positive. Tau (T) was interpreted as positive (T+) when either PET or CSF P-tau was positive; negative (T-) when either of them was obtained and neither of them were positive; and was “na” (not available) (Tna) when neither of them were obtained. Neurodegeneration (N) was interpreted as positive (N+) when the consensus visual read of FDG-PET showed a progressive pattern (P1, P2, or P1+), and negative (N-) when it was a non-progressive pattern (N1 or N2).

Table 1. Findings for each subject and ATN classification


The amyloid positivity rate was 13/38 overall (6/26 CU, 4/7 EMCI, 3/5 LMCI, p>0.05 between CU and MCI), while it was 8/38 based on the PET alone (3/26 CU, 2/7 EMCI, 3/5 LMCI, p>0.05 between CU and MCI).
Tau was positive for 7, negative for 2 and not available for 4 out of the 13 A+ subjects, being 2, 2, 2 and 5, 0, 2 out of the 6 A+ CU and 7 A+ MCI subjects, respectively.
FDG-PET showed a progressive pattern in 6/13 A+ subjects (3/6 CU, 1/4 EMCI, 2/3 LMCI) as compared to 3/25 A- subjects (0/20 CU, 1/3 EMCI, 2/2 LMCI). Significant difference was observed in the FDG-PET positivity (N+) proportion between A+ and A- (p<0.05) as well as between CU (3/26) and MCI (6/12) (p<0.05).
Of interest, tau was positive for as many as 8 (negative for 8, not available for 9) out of the 25 amyloid negative subjects, indicating tau deposition without AD pathological process. It should be noted that all the 8 A-T+ subjects was tau positive due to CSF test, in spite of negative tau PET for two of them.
Association of APOE genotypes with amyloid PET (p>0.6) or CSF Aβ (p>0.5), or with any other biomarkers, was not observed for the presence of E4, probably due to the small number of subjects.

Representative cases

Figures 1 (#24, LMCI) and 2 (#22, CU) depict a case with prodromal AD (A+T+N+) and preclinical AD (A+T+N-), respectively. PET and CSF were discordant for “A” and/or “T” in both cases, which may be related to visually equivocal images and near-cutoff level quantified values. In the case of Figure 2, CSF P-tau was positive while tau PET was negative, consistent with the report of earlier and more sensitive positivity of CSF P-tau than tau-PET in the AD continuum (13).
Four cases (1 CU, 1 EMCI, 2 LMCIs) showed a mild/partial DLB pattern in FDG-PET marked with “P1+” in Table 1, featuring hypometabolism in the occipital cortex extending into surrounding areas but not showing a typical AD pattern of temporo-parietal hypometabolism. Amyloid was positive for 3/4 and tau was positive for 4/4. Figure 3 (#26, CU) depicts one of them.

Figure 1. PiB, FTP and FDG-PET of a female LMCI patient in her 70s (#24) interpreted as prodromal AD

Amyloid PET with PiB was visually negative, as the left parietal mild accumulation did not reach the cortical surface (arrow). However, the subject was classified as “A+” because quantitative analysis revealed SUVR (1.57) above cutoff. The CSF Aβ was negative (399.8 pg/mL). The FTP-PET showed abnormal tau accumulation in the left posterior temporal lobe (arrow), typical of AD process. Note off-target uptake of FTP in choroid plexus (arrowheads), substantia nigra, and striatum. The FDG-PET was read as temporo-parietal hypometabolism indicating AD pattern in the baseline that progressed in two years (arrows). PET score and MRI z-score also increased in two years: from 0.76 to 1.08 and from 2.7 to 3.2, respectively.

Figure 2. PiB, FTP and FDG-PET of a female CU subject of her 70s (#22) interpreted as preclinical AD

PiB-PET revealed positive amyloid accumulation in the left temporal and parietal areas (arrows). Tau PET with FTP acquired 1 year later was negative, because mild activity along the cortical rim was interpreted as off-target uptake by the meninges (short arrows) and that the left anterior temporal uptake was considered non-pathological within the AD continuum (long arrow). CSF P-tau was positive. FDG-PET showed a normal pattern.

Figure 3. FBP, FTP and FDG-PET of a female CU subject of her 70s (#26)

FBP-PET presented negative amyloid, and tau was negative in FTP-PET, although CSF showed positive Aβ (317.9pg/mL) and P-tau (38.2pg/mL). FDG-PET revealed a DLB pattern, presenting occipital hypometabolism (long arrows) extending to the right temporal and parietal cortex (short arrows), which progressed 1 year later. Note cingulate island sign denoting preserved metabolism in the posterior cingulate cortex (arrowheads).


Association between PET and CSF

For the 18 subjects, in which CSF data were obtained, amyloid positivity by CSF agreed with that by PET in 12 cases while 6 showed a discordance (Table 1). The rate of discordance was consistent with previous reports and may be caused by various factors (13).
Quantified tau uptake (SUVR) measured with FTP-PET using MUBADA-PERSI method was significantly correlated with CSF P-tau (r=0.92, p<0.001, n=8) (Figure 4). Although the cutoff for SUVR with MUBADA-PERSI SUVR is not established yet, the visual read of the FTP-PET was positive only for two of them. In CSF P-tau, however, 7 out of the 8 subjects showed P-tau above the cutoff level, indicating a discordance in the tau positivity between PET and CSF. This is consistent with recent investigations that reported earlier or more sensitive positivity of CSF P-tau than tau-PET in the AD continuum (i.e. in amyloid positive subjects), because secretion of soluble p-tau to CSF is increased by Aβ pathology before tau begins to accumulate in the brain (14).

Figure 4. Scatter plot of tau uptake (SUVR) quantified with FTP-PET and CSF P-tau

Red marks indicate PET-positive cases by visual read. Arrow indicates cutoff value for CSF P-tau.



PET/CSF discordance for amyloid and tau

This study suffers limitations such as the small number of subjects, poor follow-up records, and lack of tau-PET and CSF measurement for a large fraction of the subjects. However, some findings are notable.
The rate of amyloid positivity based on the combination of PET and CSF (6/26=23% for CU, 7/12=58% for MCI) was consistent with previous reports including J-ADNI. Discordance of positivity between amyloid-PET and CSF Aβ was observed in 6 subjects (5 PET-/CSF+, 1 PET+/CSF-), suggesting higher sensitivity of CSF, which was also consistent with the ADNI data on CU and MCI (13).
The rate of tau positivity was 4/6 for amyloid PET-/CSF Aβ+ or amyloid PET+/CSF Aβ-, and 2/2 for amyloid PET+/CSF Aβ+ in this study (Table 1). This was agreeable with the above ADNI data, in which the former two groups presented significantly lower CSF p-tau and PET-measured tau deposition than the latter and suggested earlier manifestations of AD process (13).
It is known that CSF p-tau is quantitatively associated with PET-measured tau deposition, especially in the AD continuum, and that CSF p-tau rises in the earlier stage than the pathological uptake of FTP-PET accrues (14). The result of the present study indicated a similar association in spite of the small number of subjects (Figure 4), in which FTP-PET was quantified with MUBADA, and that CSF p-tau was more sensitive than visual read of FTP-PET. In the early phase of AD continuum, CSF p-tau and FTP-PET could be considered to reflect different pathological changes, as the former may indicate excess secretion of p-tau to CSF and the latter represents tissue tau deposition. The present study adopted the criteria of visual FTP-PET interpretation to determine the positivity, by which the anterior temporal FTP uptake was considered insignificant. Because MUBADA VOI covers wide cortical areas, it may not be a sensitive measure of early tau deposition in the AD continuum that begins in the temporal cortex. Although a recent study suggests the earliest tau deposition at the rhinal cortex located in the anteromedial temporal lobe (15), it was hard to quantify the pathological FTP uptake in that region due to off-target binding to the choroid plexus in the present study (Figure 2). In that sense, use of FTP was another limitation of this study. Tau PET drugs with little off-target binding such as 18F-MK-6240 or 18F-PI-2620 may be more suitable for evaluation of the earliest stage of AD.

Neurodegeneration marker

In this study only visual assessment of FDG-PET was used to evaluate the “N” (neurodegeneration) marker, and quantitative PET score was not used so that one “N-” subject presented a high PET score (#4). Although CSF T-tau and MRI-volumetry are also regarded as an N-marker, their association with FDG-PET remains to be investigated as they represent different pathophysiology.
Hypometabolism depicted by FDG-PET reflects reduced neuronal activity in general, regardless of pathophysiology. No subjects showed A+T-N+ in this study, which agrees with the concept of tau deposition leading to neurodegeneration in AD continuum, although such manifestation, if occurred, might have suggested combined AD and non-AD processes. Outside the AD continuum (A-), however, FDG-PET neurodegeneration was positive in 3/5 MCI (2/2 LMCI) subjects and was not observed in CU subjects (0/20), which is consistent with the above notion and agrees with previous reports (16).

Binary criteria

A large number of subjects presented quantitative measures and/or visual read that are close to the borderline of binary positivity in the present study, which caused, at least partly, discordance between PET and CSF in amyloid and/or tau, such as the cases in Figures 1 and 2. This is understandable because most of the subjects in the present study were in the early phase of AD continuum or of a non-AD disease if any, and that the current criteria and cutoff level have been derived from differential diagnosis of AD patients from CU subjects. To deal with early-phase subjects having no or very mild cognitive impairment, the criteria and cutoff level of the biomarkers may need refining, and the data acquisition may need strict quality control.

Non-AD disorders

The present study revealed 8 A-T+ subjects. Because two of them (#37, #38) were LMCI patients and showed AD or DLB pattern in FDG-PET, they are considered SNAP and to have cognitive impairment due to non-AD disorders (17). The other six are CU subjects and had non-progressive FDG pattern, and may suggest a very early stage of various non-AD tauopathy such as PART or normal aging process (17). There is also a possibility of false-positive CSF P-tau as the value was close to the cutoff, ranging 31.8-38.2 pg/mL for 5 of the 6 subjects. Since biomarkers in non-AD tauopathies are not well understood, further investigations are needed.
Another EMCI subject (#36) presented “A-T-N+” with FTD pattern in FDG-PET, and was suspected of early stage of non-amyloid non-tau FTLD.
The present study also revealed 4 subjects presenting DLB pattern in FDG-PET, with occipital hypometabolism extending to surrounding areas (Figure 3). Three of them were amyloid positive and all were tau positive. It is not clear whether they were preclinical or prodromal stage of atypical AD, or DLB with or without amyloid deposition.
These findings suggest that a significant fraction of the subjects in this study might be related to non-AD disorders such as DLB, SNAP, PART, argyrophilic grain disease (AGD), and TDP-43 proteinopathy (like LATE) (17). Even if they are amyloid positive, there is a possibility of incidental amyloid deposition. Therefore, possibility of non-AD disorders should always be considered when clinical trials targeting preclinical or prodromal AD are designed and subjects are selected based on the biomarkers.



In conclusion, this study confirmed the known changes of PET and CSF biomarkers in preclinical and prodromal AD, and at the same time, suggested difficulties of determining the criteria and cutoff level of those biomarkers to evaluate such subjects as well as the possibility of unsolicited inclusion of early-phase non-AD disorders.


Acknowledgements: We are grateful for the materials and technical supports for the PET imaging by Fujifilm Toyama Chemical, Avid Radiopharmaceuticals/Eli Lilly Japan, and GE Healthcare. PET centers that imaged the subjects but did not belong to the clinical site that enrolled the subjects are also acknowledged, including Tohoku University Cyclotron and Radioisotope Center (CYRIC), Tsukuba Advanced Imaging Center (AIC), Tokyo Metropolitan Institute of Gerontology (TMIG), Aizawa Hospital, MI Clinic, and Kobe City Medical Center General Hospital (KCGH). We thank all the people who participated in this study in the clinical and imaging sites as well as in the Core sites.

Conflict of interest: The following conflicts of interest are disclosed by the authors. Senda reports provision of devices, cassettes, and precursors from Avid/Eli Lilly Japan and GE, funding as PI of clinical trials sponsored by Eli Lilly, Eisai, Biogen, Cerveau and Merck, as well as leadership role in the Japanese Society of Nuclear Medicine as board member, congress chair and committee chair. Ikeuchi reports grants from AMED (JP19dk0207020, JP20dk0202028, JP20dm0207073). Matsuda reports a grant from AMED (19dk0207020h0005), intramural grants from National Center of Neurology and Psychiatry, and an entrusted research fund from Nihon Medi-Physics Co. Ltd. Iwatsubo reports a grant from an anonymous Foundation. Iwata reports grants from AMED (19dk0207020h0005, 16dk0207028h0001). Ikari is a full time employee of CMIC Inc. as well as graduate student of Osaka University. Washizuka reports research funding from AMED and pharmaceutical companies including Otsuka, Eisai, Pfizer, Daiichi-Sankyo, Tsumura, Mochida, Astellas, Shionogi, Takeda, Sumitomo-Dainippon, as well as honoraria from such pharmaceutical companies. Kazunari Ishii reports honoraria from Nihon Medi-Physics. Yokota reports licensing and collaboration research with Takeda Pharmaceutical Company. Nakanishi reports research funding from Eisai and Elli Lilly Japan as well as leadership role as a director in the Japan Society for Dementia Research. Shimada reports grants from AMED (19dk0207020h0005, 20dk0207028h0005). The other authors have nothing to disclose.

Funding: This study was financially supported by grants from Japan Agency for Medical Research and Development (AMED) 19dk0207020h0005, 20dk0202028h0005 and 20dm0207073h003, as well as by an anonymous Foundation.



1. Jack CR, Jr., Bennett DA, Blennow K, et al. A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 2016;87:539-547.
2. Iwatsubo T, Iwata A, Suzuki K, et al. Japanese and North American Alzheimer’s Disease Neuroimaging Initiative studies: Harmonization for international trials. Alzheimers Dement 2018;14:1077-1087.
3. Ikari Y, Akamatsu G, Nishio T, et al. Phantom criteria for qualification of brain FDG and amyloid PET across different cameras. EJNMMI physics 2016;3:23.
4. Yamane T, Ishii K, Sakata M, et al. Inter-rater variability of visual interpretation and comparison with quantitative evaluation of (11)C-PiB PET amyloid images of the Japanese Alzheimer’s Disease Neuroimaging Initiative (J-ADNI) multicenter study. Eur J Nucl Med Mol Imaging 2017;44:850-857.
5. Thurfjell L, Lilja J, Lundqvist R, et al. Automated quantification of 18F-flutemetamol PET activity for categorizing scans as negative or positive for brain amyloid: concordance with visual image reads. J Nucl Med 2014;55:1623-1628.
6. Clark CM, Pontecorvo MJ, Beach TG, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-β plaques: a prospective cohort study. The Lancet Neurology 2012;11:669-678.
7. Yamane T, Ikari Y, Nishio T, et al. Visual-statistical interpretation of (18)F-FDG-PET images for characteristic Alzheimer patterns in a multicenter study: inter-rater concordance and relationship to automated quantitative evaluation. AJNR Am J Neuroradiol 2014;35:244-249.
8. Silverman DHS, Small GW, Chang CY, et al. Positron Emission Tomography in Evaluation of DementiaRegional Brain Metabolism and Long-term Outcome. JAMA 2001;286:2120-2127.
9. Herholz K, Salmon E, Perani D, et al. Discrimination between Alzheimer dementia and controls by automated analysis of multicenter FDG PET. NeuroImage 2002;17:302-316.
10. Herholz K, Westwood S, Haense C, Dunn G. Evaluation of a calibrated (18)F-FDG PET score as a biomarker for progression in Alzheimer disease and mild cognitive impairment. J Nucl Med 2011;52:1218-1226.
11. Southekal S, Devous MD, Sr., Kennedy I, et al. Flortaucipir F 18 Quantitation Using Parametric Estimation of Reference Signal Intensity. J Nucl Med 2018;59:944-951.
12. Devous MD, Sr., Joshi AD, Navitsky M, et al. Test-Retest Reproducibility for the Tau PET Imaging Agent Flortaucipir F 18. J Nucl Med 2018;59:937-943.
13. Reimand J, Collij L, Scheltens P, Bouwman F, Ossenkoppele R, Alzheimer’s Disease Neuroimaging I. Association of amyloid-beta CSF/PET discordance and tau load 5 years later. Neurology 2020;95:e2648-e2657.
14. Mattsson-Carlgren N, Andersson E, Janelidze S, et al. Aβ deposition is associated with increases in soluble and phosphorylated tau that precede a positive Tau PET in Alzheimer’s disease. Science advances 2020;6:eaaz2387.
15. Sanchez JS, Becker JA, Jacobs HIL, et al. The cortical origin and initial spread of medial temporal tauopathy in Alzheimer’s disease assessed with positron emission tomography. Science translational medicine 2021;13:eabc0655.
16. Coutinho AM, Busatto GF, de Gobbi Porto FH, et al. Brain PET amyloid and neurodegeneration biomarkers in the context of the 2018 NIA-AA research framework: an individual approach exploring clinical-biomarker mismatches and sociodemographic parameters. Eur J Nucl Med Mol Imaging 2020;47:2666-2680.
17. Jack CR, Jr., Knopman DS, Chételat G, et al. Suspected non-Alzheimer disease pathophysiology–concept and controversy. Nature reviews Neurology 2016;12:117-124.



I. Choi1, H. La Monica1, S.L. Naismith2, A. Rahmanovic2, L. Mowszowski2, N. Glozier1


1. Central Clinical School, Brain and Mind Centre, Faculty of Medicine and Health, University of Sydney, Australia; 2. Charles Perkins Centre, Faculty of Science, School of Psychology and the Brain and Mind Centre, University of Sydney, Australia

Corresponding Author: Dr Isabella Choi, 94 Mallett Street, Camperdown, NSW 2050, Australia,, +612 8627 7240.

J Prev Alz Dis 2021;
Published online June 23, 2021,



Communicating personal Alzheimer’s disease risk profiles based on validated risk algorithms may improve public knowledge about risk reduction, and initiate action. This proof of concept pilot trial aimed to test whether this is feasible and potentially effective and/or harmful. Older at-risk adults (N=24) were provided with their personal Alzheimer’s disease risk profile online, which contained information on their personal risk level, scores and tailored recommendations to manage modifiable risk factors. After receiving the risk profile, participants were significantly more accurate in identifying risk and protective factors, and revised their perceived risk to be lower than their initial estimate. There was no apparent harm seen in psychological distress or dementia-related worry. This shows preliminary support for the feasibility of delivering personal dementia risk profiles to low risk, help-seeking older adults in an online format. A definitive trial examining behavioural outcomes and testing in groups with higher risk profiles is now warranted.

Key words: Risk communication, health literacy, psychological distress, prevention, Alzheimer’s disease.



Up to a third of Alzheimer’s disease cases can be prevented through improved education and reduction of modifiable risk factors (1). Growing evidence from multi-domain interventions shows that targeting modifiable risk factors can reduce risk of Alzheimer’s disease (AD) and improve cognition (2, 3). However, lack of knowledge about dementia and its risk factors among the public is a major barrier to individuals implementing behavioural and lifestyle change and, in turn, to dementia risk reduction (4).
Having an accurate understanding of one’s personal risk of future disease is considered essential for engaging in behaviours for risk reduction. Most health behaviour change models, including the Health Belief Model, identify four major constructs that surround health behaviour: health literacy, perceived susceptibility, motivation to change, and perceived barriers to change (5). However, there is poor health literacy of dementia risk, risk factors, and prevention strategies among the public. A systematic review found that almost half of the general public agreed that dementia was a normal and non-preventable part of ageing (6). Mental activity, healthy diet, physical exercise and social engagement were the most commonly nominated ways to reduce risk, but other well established risk factors such as vascular risk (including smoking, high blood pressure, high cholesterol, obesity), low education, poor mental health, brain trauma, and environmental toxins were rarely mentioned (6-8). Only around 25% of Australians were confident they could reduce risk (8). However, there is support that improving dementia health literacy has a positive impact on risk reduction. People who had a strong belief that dementia risk could be reduced, had moderate to high knowledge of risk-reduction behaviours, or had high confidence that risk reduction can be achieved were almost twice as likely to take action to reduce risk (9).
Communicating personal dementia risk level and risk factors to at-risk adults, based on validated AD risk algorithms, may be one way to improve dementia risk knowledge and engage people in risk reduction behaviours. A number of dementia risk algorithms have been validated for the general population and can identify those with high risk with acceptable predictive ability (10). For instance, the Australian National University Alzheimer’s Disease Risk Index (ANU-ADRI) includes various modifiable risk factors that have been validated for middle-age and older adults that are easily assessed via self-report (11, 12). These evidence-based algorithms, along with personalised risk factor feedback and recommendations to reduce risk, can be delivered online for wide access, allowing people to screen for their risk without having had to first consult with a physician. Older adults are able to use technology proficiently and over 90% of help-seeking older adults with some degree of cognitive impairment use the Internet at home, most commonly for emails (13). Such online dementia risk algorithms already exist and are made freely available to the public, for instance, the CAIDE Risk Score App allows users to detect their dementia risk, obtain information on modifiable risk factors, and receive suggestions on how to modify their risk (14).
Surveys have found that there is high interest among older adults in knowing their risk of AD (15, 16) While there may be potential benefits to disclosing risk, there are also concerns that this could cause negative psychological outcomes. For instance, about 30% of older adults who were interested in knowing their risk also actively worried about developing AD (17). Evidence from a systematic review suggests that disclosure of increased AD risk was not associated with anxiety or depression, but did lead to heightened test-related distress, long-term care insurance uptake and health-related behaviour changes (18).
This pilot trial explores the feasibility and acceptability of communicating online personal dementia risk profiles to at-risk older adults and the impact on dementia health literacy, motivation to engage in behaviour change, and potential harmful psychological effects.




This is an uncontrolled proof of concept pilot trial. The study was approved by the University of Sydney Human Research Ethics Committee (protocol number: 2019/669) and registered with the Australian New Zealand Clinical Trials Registry (ANZCTRN12619001242112p).


People attending the Healthy Brain Ageing Clinic, a specialised memory research clinic for people aged 50 years or older in Sydney, Australia, who were clinically diagnosed with mild cognitive impairment (MCI) or subjective cognitive decline (SCD) between October 2019 to May 2020 were recruited. Those who had dementia or pre-existing severe cognitive impairment due to neurological conditions, psychosis, intellectual disability, substance misuse, stroke, or acquired brain injury were excluded.
As part of standard clinic procedure, all clinic attendees completed self-report measures and were assessed by a geriatrician or neurologist, a psychologist, and a neuropsychologist for a review of medical and psychiatric history, mood, and cognitive functioning. Diagnoses were rated according to consensus, including at least two neuropsychologists and one specialist, and were used to exclude those with a dementia diagnosis and other exclusion criteria. Within two weeks of the clinic assessment, attendees received neuropsychological test result feedback over the phone from a neuropsychologist. The clinic does not provide treatment but refers attendees to suitable clinical investigations, e.g. sleep studies, if warranted, as well as clinical trials for which they are deemed eligible.
After receiving neuropsychological test feedback from a neuropsychologist, eligible attendees were invited via a telephone call to participate in the current study. Interested people had to have an email account and were emailed a survey link via REDCap (Research Electronic Data Capture), a web-based research management platform, with the Participant Information Statement and Consent Form. Participants gave consent to extract relevant data collected from their recent clinic assessment to populate their personal dementia risk profile.


Participants completed self-report baseline measures online via REDCap. Participants’ demographic and risk information were extracted from their standard clinic assessment to compile their personal dementia risk profile using the ANU-ADRI (11). Risk factors in the model included: age, gender, highest level and total number of years of education completed, body mass index, diabetes, depression, high cholesterol, traumatic brain injury, smoking status, alcohol intake, social engagement, physical activity, cognitive activity, and diet.
Within two weeks of completing the baseline self-report measures, participants were emailed a pdf document with their personal dementia risk profile. The risk profile contained standard information about dementia, an explanation of their personal dementia risk profile, and information about the ANU-ADRI risk model. Participants were presented with a visual representation of their risk level in the form of a thermometer showing their risk from 0 to 100, along with an explainer “Your risk of developing dementia is low/ moderate/ high. It is estimated that XX out of 100 people with your risk factors will develop dementia in their lifetime” (Figure 1). They were reminded that this is an estimate based on their risk factors rather than a definitive guarantee, and that there are some risk factors they cannot change but some they could potentially work on to reduce their risk. Participants also received a summary of the dementia risk factors included in the risk model and their scores on each risk factor. They were told why the risk factor was important for brain functioning and were given tailored recommendations to manage each risk factor based on their risk factor score, as well as links for more information.
One week after receiving their risk profile, participants received an email asking them to complete the online post-intervention measures in REDCap. After completing all study measures, participants were reimbursed with a $20 gift card in return for their time.

Figure 1. Example of the risk level and risk factor feedback provided in the personal risk profile




Primary outcome: Dementia health literacy

Participants were asked “How likely do you think that you will get Alzheimer’s disease in your lifetime?” to assess perceived risk on a scale, where 0%=certain not to happen and 100%=certain to happen. To examine accuracy of perceived risk, the participant’s perceived risk was subtracted from their ANU-ADRI risk estimate. We recoded the difference (D) into a categorical variable, with <−10% indicating overestimation, >10% indicating underestimation, and accurate if −10% ≤ D ≤ 10%, in accordance with previous studies (19). Similarly, to examine change in perceived risk, participants’ perceived risk at post-intervention was subtracted from their perceived risk at baseline (d), and <-10% indicates increased perceived risk, >10% indicates reduced perceived risk, and −10% ≤ d ≤ 10% indicates no change.
We adapted the dementia risk and protective factors questionnaire in the MijnBreincoach survey (20) to assess for knowledge of dementia risk factors. We included five additional modifiable risk and protective factors that were identified in the ANU-ADRI (11) and the Lancet Commission for dementia prevention (21) (i.e. traumatic brain injury, social activity, sleep, education, and age), totalling 19 risk factors. Additional questions asked about barriers to improving brain health, confidence in risk reduction (8), and worry about getting dementia (7).

Secondary outcome: Motivation to Change Lifestyle and Health Behaviours for Dementia

The Motivation to Change Lifestyle and Health Behaviours for Dementia Risk Reduction (MCLHB-DRR) Scale is designed to assess beliefs and attitudes about lifestyle and health behavioural changes for dementia risk reduction among middle-aged and older adults (22). The scale includes (27) items matched onto seven subscales that reflect the seven concepts of the Health Behaviour Model. All items are rated on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). The scale has moderate to high internal consistencies for the seven subscales, and moderate test-retest reliability (18). Cronbach’s alpha for each of the subscales in this study were: perceived susceptibility (.916), perceived severity (.331), perceived benefits (.715), perceived barriers (.878), cues to action (.656), general health motivation (.638), and self-efficacy (.615).

Secondary outcome: Psychological distress

The K10 is a commonly used screening scale for non-specific psychological distress validated for use in Australia (23). The K10 has also been demonstrated as having moderate sensitivity to symptom change in an Australian sample (24). Scores on the K10 range from 10 to 50, and a score of 30 or more indicates a severe level of distress. Cronbach’s alpha in this study was 0.841.

Secondary outcome: Dementia-related worry

The Dementia Worry Scale was used to assess dementia-related worry (25). It has strong internal consistency and test-retest reliability. It consists of 12 items with scores ranging from 15 to 60. Cronbach’s alpha in this study was 0.908.

User evaluation

We adapted a five-point scale (from 1= not at all to 5= completely) previously used to assess user experience of a dementia information website (26). Participants were asked whether the information provided was engaging and easy to understand as well as how helpful they found the risk profile and how much they felt they had learned (from 1=nothing at all to 5=a great deal). Additionally, participants were asked if they required more information about their dementia risk profile and were given the option to discuss their experience of using the risk profile with a researcher in a telephone interview.

Data analysis

Data was analysed using SPSS version 23.0. Descriptive statistics regarding participant and baseline characteristics were analysed. Fischer’s exact tests and paired samples t-tests were used to test for differences between outcome measures pre- and post- receiving the dementia risk profile. All p-values were two-sided with an alpha of 0.05 to test for significance.



Demographics and baseline characteristics

Overall, 24 eligible participants participated in the trial (Figure 2). Participants’ ages ranged from 53 to 87, with a mean age of 69.54 years (SD 7.69). Over half of the participants (54%) were female and majority spoke English as a first language (83%). Majority were tertiary educated (75%), 21% had completed a trade certificate, and 4% had completed high school. Majority were retired (58%), 29% were employed, and 13% were unemployed. Over half were married or in a de facto relationship (58%), 25% were widowed or divorced, and 17% had never married. The majority (75%) of participants had MCI and 25% had SCD.
Almost all participants (96%; 23/24) were considered to have Low Risk of developing AD (ANU-ADRI score of less than 17), and one participant (4%) was considered as having High Risk (ANU-ADRI score of greater than 27). Participants’ perceived risk of developing AD ranged from 10-99 (M=51.63, SD=23.85), with the majority of participants overestimating their personal risk (87.5%; 21/24).

Figure 2. Participant flow

Pre-post change on dementia health literacy

All participants completed the post-intervention questionnaires. After receiving their personal dementia risk profile, participants’ perceived risk of developing AD ranged from 3-81 (M=38.83, SD=25.02). There was a significant decrease in perceived risk among the group (p = .010). A total of 18 participants (75%) still overestimated their level of risk whereas the remaining 6 correctly identified their level of risk (25%). Eleven participants (45.8%) reported a reduction in their perceived risk, eleven (45.8%) reported no change, and two participants’ (8.3%) perceived risk had increased.
The average number of correctly identified risk and protective factors increased from a mean of 11.42 items (SD= 4.50) at baseline to 13.96 items (SD= 3.98) (t1,23= -3.839, p=.001) at follow-up.

Pre-post change on motivation to engage in behaviour change and psychological effects

There was a significant reduction on the perceived susceptibility subscale of the MCLHB (t1,23=4.416, p<.001) from baseline (M=12.86; SD=3.30) to 1-week follow up (M=10.29; SD=4.21), but no change on the other subscales. There was no change on the K10 or Dementia Worry Scale.
Participants’ self-reported worry about getting dementia was significantly reduced at follow-up, from 2.75 (SD=.85) to 2.29 (SD=.91) (t1,23= 3.412, p= .002). The most common barriers to reducing risk factors at baseline were lack of knowledge (45.8%; 11/24), followed by health problems (25%; 6/24). At 1-week follow up, the most common barriers were lack of motivation (29.2%; 7/24), health problems (29.2%; 7/24), and difficulty with organisation (25%; 6/24). There was no change in confidence to take action to change risk.

User evaluation

Overall, 70.9% (17/24) of participants agreed that the information in the personal risk profile was engaging, 79.2% (19/24) agreed the information was easy to understand, 79.2% (19/24) agreed it was helpful, and 79.2% (19/24) reported they learned a good deal from their personal risk profile.
Two participants participated in the optional telephone interview with a researcher. Both expressed surprise at their lower than expected AD risk feedback, and identified difficulty addressing some of their risk factors (e.g. getting motivated to exercise). One participant agreed that seeking guidance from a health professional may support them to work on their risk factors.



This pilot study aimed to explore the feasibility, acceptability and potential impact of providing an online personal dementia risk profile to help-seeking older adults at risk of developing AD on risk knowledge, motivation to change health-related behaviours, and psychological effects. To our knowledge, this is the first study focusing on the effects of communicating personal risk profile using risk algorithms based on epidemiological risk factors. Communication of the personal dementia risk profile led to more accurate knowledge of AD risk factors and improved understanding of perceived susceptibility among patients with MCI and SCD. Importantly, there was no negative effect of communicating the personal risk profile online on psychological distress or dementia-related worry among our participants. Participants mostly had a low risk of developing AD, but still reported reduced worry about getting dementia after receiving their risk profile. These findings support the feasibility and acceptability of using dementia risk algorithms to deliver personal risk profiles to low risk older adults in an online format, and indicate that providing this information can improve AD health literacy without a negative impact on psychological wellbeing.
However, there was little evidence in this study that providing personal risk profiles as a standalone intervention was sufficient to motivate change in behaviours to address AD risk factors. Although providing the dementia risk profile addressed one main barrier for risk reduction at baseline (i.e. lack of knowledge of dementia risk factors), participants reported that lack of motivation, health problems, and difficulty with organisation became the main barriers after receiving their risk profile. This suggests that older adults need extra support to effect behavioural change. The personal dementia risk profile could potentially be used as part of a collaborative, shared decision-making approach to address these barriers by guiding and engaging users, carers and clinicians to choose several high impact or easy-to-change risk factors to focus on, and by providing feedback on the change in risk level if risk factors are modified. Trials are underway to test the impact of a tailor-made online lifestyle programme targeting modifiable risk factors on risk score and health behaviours and compliance to health advice (27). There may also be a role for clinicians to follow up with specific guidance on addressing health problems and to assist the older adult to develop a personalised risk reduction plan. A recent rapid review on approaches to healthy ageing interventions for older adults demonstrated that optimal interventions are those that incorporate collaborative approaches with shared decision-making and behavioural change techniques (28). In this regard, the personal dementia risk profile represents a useful tool that clinicians can draw on to present evidence-based, tailored, health and risk information, which in turn can stimulate a collaborative decision-making process around which health/lifestyle factors to target and how best to achieve long-lasting behaviour change.
An interesting finding was that most participants overestimated their risk of developing AD, even after receiving their personal risk profile. This is possibly reflective of our cohort which was composed of help-seeking older adults seeking an evaluation at a memory clinic and were concerned with developing dementia. The continuing high levels of perceived risk at follow up is unsurprising given that previous research has found that even among individuals who accurately recalled their communicated AD risk, over 50% did not fully adjust their perceived risk to match the communicated risk, and that high baseline AD risk perception was the strongest predictor of overestimation of risk (29). It is also possible that participants may not have readily accepted the communicated risk after receiving lower than expected risk feedback, as seen by interview participants expressing surprise at their communicated risk.
Nonetheless, this has implications for supporting clinicians to communicate AD risk information to people with MCI. A survey found that 90% of neurologists said they counselled MCI patients on their risk of AD in general terms but only 60% communicated AD risk in numeric terms (30). Our findings provide preliminary support that patients with MCI or SCD understood numeric AD risk information and risk factors even when it is communicated online without support, and that the multifaceted approach with a clear visual representation and accompanying explanatory text may have facilitated understanding. Clinicians are encouraged to discuss numeric risk estimates with patients with visual aids, explain how these are estimated from risk models, and explore reasons for discord to improve risk appraisal.
There are several important considerations in interpreting the findings. The study sample was a well-educated inner-city cohort who have expressed concern about their memory and were highly motivated to seek help. Participants already knew that they did not have dementia. Their relatively positive reactions to their personal risk profile may reflect their personal interest in managing their brain health or because they were reassured of having low risk. It is unclear whether a general population or primary care sample, who had subjective memory concerns and not assessed for AD, would have similar reactions. It should be noted that a number of older adults approached to take part in the trial declined due to not wanting to know their risk or because they felt overwhelmed. Further research is needed to explore these concerns about knowing one’s personal risk.
The majority of our sample had low risk of developing AD, and it is unknown how moderate or high risk adults would respond to their personal dementia risk profile. There is some indication that high risk individuals, such as those who screen positive to genetic biomarkers, have heightened test-related distress (18). In order for dementia risk profiles to be widely and safely distributed to older adults in public health programs, particularly if they are to be delivered online in the absence of immediate clinical support, it is important to understand how moderate or high risk older adults react to their personal dementia risk profile and to monitor any potential adverse reactions. Finally, this was a pilot trial with a small sample size and short-term follow up. Longer-term follow up and randomised controlled trials to examine effects of communication of personal risk of developing AD are required.
The application of dementia risk algorithms to identify those at risk and to promote and encourage risk reduction behaviour is still in its early stages. This study provides preliminary support for the utility of using risk models that incorporate accessible and potentially modifiable risk factors to communicate personal dementia risk profiles to at-risk older adults.


Acknowledgements: The authors would like to acknowledge Professor Kaarin J. Anstey and Dr Sarang Kim for permission to use the ANU Alzheimer’s Disease Risk Index and their advice on adapting it to the Healthy Brain Ageing clinic measures. We thank the participants who have helped make this research possible.

Funding: This study was supported by a Dementia Australia Research Foundation project grant award. The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.

Ethical standards: The study was approved by the University of Sydney Human Research Ethics Committee (protocol number: 2019/669) and registered with the Australian New Zealand Clinical Trials Registry (ANZCTRN12619001242112p).

Conflict of interest: The authors confirm that there are no known conflicts of interest associated with this publication.



1. Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential for primary prevention of Alzheimer’s disease: an analysis of population-based data. The Lancet Neurology. 2014;13(8):788-794
2. Isaacson RS, Hristov H, Saif N, et al. Individualized clinical management of patients at risk for Alzheimer’s dementia. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2019;15(12):1588-1602
3. Ngandu T, Lehtisalo J, Solomon A, et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. The Lancet. 2015;385(9984):2255-2263 S0140-6736(15)60461-5.
4. Kim S, Sargent-Cox KA, Anstey KJ. A qualitative study of older and middle-aged adults’ perception and attitudes towards dementia and dementia risk reduction. J Adv Nurs. 2015;71(7):1694-1703
5. Janz NK, Becker MH. The Health Belief Model: A Decade Later. Health Educ Q. 1984;11:1-47
6. Cations M, Radisic G, Crotty M, Laver KE. What does the general public understand about prevention and treatment of dementia? A systematic review of population-based surveys. PLoS One. 2018;13(4):e0196085
7. Low L-F, Anstey KJ. Dementia literacy: Recognition and beliefs on dementia of the Australian public. Alzheimer’s & Dementia. 2009;5(1):43-49
8. Smith BJ, Ali S, Quach H. Public knowledge and beliefs about dementia risk reduction: a national survey of Australians. BMC Public Health. 2014;14(1):661
9. Smith BJ, Ali S, Quach H. The motivation and actions of Australians concerning brain health and dementia risk reduction. Health Promot J Austr. 2015;26(2):115-121
10. Hou X-H, Feng L, Zhang C, Cao X-P, Tan L, Yu J-T. Models for predicting risk of dementia: a systematic review. J Neurol Neurosurg Psychiatry. 2018:jnnp-2018-318212
11. Anstey KJ, Cherbuin N, Herath PM. Development of a New Method for Assessing Global Risk of Alzheimer’s Disease for Use in Population Health Approaches to Prevention. Prevention Science. 2013;14(4):411-421
12. Anstey KJ, Cherbuin N, Herath PM, et al. A Self-Report Risk Index to Predict Occurrence of Dementia in Three Independent Cohorts of Older Adults: The ANU-ADRI. PLoS One. 2014;9(1):e86141
13. LaMonica HM, English A, Hickie IB, et al. Examining Internet and eHealth Practices and Preferences: Survey Study of Australian Older Adults With Subjective Memory Complaints, Mild Cognitive Impairment, or Dementia. J Med Internet Res. 2017;19(10):e358
14. Sindi S, Calov E, Fokkens J, et al. The CAIDE Dementia Risk Score App: The development of an evidence-based mobile application to predict the risk of dementia. Alzheimer’s & Dementia : Diagnosis, Assessment & Disease Monitoring. 2015;1(3):328-333
15. Sheffrin M, Stijacic Cenzer I, Steinman MA. Desire for predictive testing for Alzheimer’s disease and impact on advance care planning: a cross-sectional study. Alzheimers Res Ther. 2016;8(1):55.
16. Wikler EM, Blendon RJ, Benson JM. Would you want to know? Public attitudes on early diagnostic testing for Alzheimer’s disease. Alzheimers Res Ther. 2013;5(5):43
17. Roberts JS, McLaughlin SJ, Connell CM. Public beliefs and knowledge about risk and protective factors for Alzheimer’s disease. Alzheimer’s & Dementia. 2014;10(5):S381-S389
18. Bemelmans SASA, Tromp K, Bunnik EM, et al. Psychological, behavioral and social effects of disclosing Alzheimer’s disease biomarkers to research participants: a systematic review. Alzheimers Res Ther. 2016;8(1):46
19. Rimer BK, Halabi S, Sugg Skinner C, et al. Effects of a mammography decision-making intervention at 12 and 24 months. Am J Prev Med. 2002;22:247-257
20. Heger I, Deckers K, van Boxtel M, et al. Dementia awareness and risk perception in middle-aged and older individuals: baseline results of the MijnBreincoach survey on the association between lifestyle and brain health. BMC Public Health. 2019;19(1):678
21. Livingston G, Sommerlad A, Orgeta V, et al. Dementia prevention, intervention, and care. The Lancet. 2017;390(10113):2673-2734
22. Kim S, Sargent-Cox K, Cherbuin N, Anstey KJ. Development of the motivation to change lifestyle and health Behaviours for Dementia Risk Reduction Scale. Dement Geriatr Cogn Dis Extra. 2014;4(2):172-183
23. Kessler R, Andrews G, Colpe L, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med. 2002;32(06):959-976
24. Perini SJ, Slade T, Andrews G. Generic effectiveness measures: Sensitivity to symptom change in anxiety disorders. J Affect Disord. 2006;90:123-130
25. Kinzer A, Suhr JA. Dementia worry and its relationship to dementia exposure, psychological factors, and subjective memory concerns. Applied Neuropsychology: Adult. 2016;23(3):196-204
26. Farrow M. User perceptions of a dementia risk reduction website and its promotion of behavior change. JMIR research protocols. 2013;2(1):e15-e15
27. Vrijsen J, Abu-Hanna A, Maeckelberghe ELM, et al. Uptake and effectiveness of a tailor-made online lifestyle programme targeting modifiable risk factors for dementia among middle-aged descendants of people with recently diagnosed dementia: study protocol of a cluster randomised controlled trial (Demin study). BMJ Open. 2020;10(10):e039439
28. Owusu-Addo E, Ofori-Asenso R, Batchelor F, Mahtani K, Brijnath B. Effective implementation approaches for healthy ageing interventions for older people: A rapid review. Arch Gerontol Geriatr. 2021;92:104263
29. Linnenbringer E, Roberts JS, Hiraki S, Cupples LA, Green RC. “I know what you told me, but this is what I think:” Perceived risk of Alzheimer disease among individuals who accurately recall their genetics-based risk estimate. Genet Med. 2010;12(4):219-227
30. Roberts JS, Karlawish JH, Uhlmann WR, Petersen RC, Green RC. Mild cognitive impairment in clinical care: A survey of American Academy of Neurology members. Neurology. 2010;75(5):425-431



T. Ochiai1,2, T. Nagayama1, K. Matsui1, K. Amano1, T. Sano1, T. Wakabayashi1,3, T. Iwatsubo1


1. Department of Neuropathology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; 2. Pharmacology Department, Drug Research Center, Kaken Pharmaceutical Co., LTD., Kyoto, Japan; 3. Department of Innovative Dementia Prevention, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan

Corresponding Author: Tomoko Wakabayashi, Takeshi Iwatsubo, Department of Neuropathology, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan, Tel: +81-3-5841-3541, Fax: +81-3-5841-3613,

J Prev Alz Dis 2021;
Published online June 17, 2021,



BACKGROUND: Obesity and diabetes are well-established risk factors of Alzheimer’s disease (AD). In the brains of patients with AD and model mice, diabetes-related factors have been implicated in the pathological changes of AD. However, the molecular mechanistic link between the peripheral metabolic state and AD pathophysiology have remained elusive. Endoplasmic reticulum (ER) stress is known as one of the major contributors to the metabolic abnormalities in obesity and diabetes. Interventions aimed at reducing ER stress have been shown to improve the systemic metabolic abnormalities, although their effects on the AD pathology have not been extensively studied.
OBJECTIVES: We examined whether interventions targeting ER stress attenuate the obesity/diabetes-induced Aβ accumulation in brains. We also aimed to determine whether ER stress that took place in the peripheral tissues or central nervous system was more important in the Aβ neuropathology. Furthermore, we explored if age-related metabolic abnormalities and Aβ accumulation could be suppressed by reducing ER stress.
METHODS: APP transgenic mice (A7-Tg), which exhibit Aβ accumulation in the brain, were used as a model of AD to analyze parameters of peripheral metabolic state, ER stress, and Aβ pathology in the brain. Intraperitoneal or intracerebroventricular administration of taurodeoxycholic acid (TUDCA), a chemical chaperone, was performed in high-fat diet (HFD)-fed A7-Tg mice for ~1 month, followed by analyses at 9 months of age. Mice fed a normal diet were treated with TUDCA by drinking water for 4 months and intraperitoneally for 1 month in parallel, and analyzed at 15 months of age.
RESULTS: Intraperitoneal administration of TUDCA suppressed ER stress in the peripheral tissues and ameliorated the HFD-induced obesity and insulin resistance. Concomitantly, Aβ levels in the brain were significantly reduced. In contrast, intracerebroventricular administration of TUDCA had no effect on the Aβ levels. Peripheral administration of TUDCA was also effective against the age-related obesity and insulin resistance, and markedly reduced amyloid accumulation.
CONCLUSIONS: Interventions that target peripheral ER stress might be beneficial therapeutic and prevention strategies against brain Aβ pathology associated with metabolic overload and aging.

Key words: ER stress, Aβ, obesity, diabetes, Alzheimer’s disease.



A growing body of evidence suggests that a certain proportion of the causative factors of dementia (1), and also those of Alzheimer’s disease (AD) (2), may be attributable to lifestyle-related risks. Therefore, it may be possible to prevent AD by identifying and reducing those risk factors (3). Epidemiological and biological evidence that support the association of obesity and type 2 diabetes mellitus (T2DM) with AD is increasingly accumulating. In particular, meta-analyses of prospective cohort studies showed that the presence of diabetes increases the risk of AD by ~1.5 times (4, 5). Various biological mechanisms have been suggested to explain the link between these lifestyle-related risks and AD: pathological conditions such as chronic inflammation, hyperglycemia, insulin resistance, and vascular complications underlie cognitive dysfunctions (6, 7). Notably, it has been established that peripheral insulin resistance is associated with higher amyloid plaque loads in the brains of patients with AD (8). This has been recapitulated in the mouse models of AD, in which the degrees of diet-induced insulin resistance and metabolic abnormalities correlated well with brain Aβ accumulation, and moreover, the effects were reversible (9–13). Because Aβ accumulation is considered to play a causative role in the pathophysiology of AD (14), deterioration of the peripheral metabolic state may adversely affect both the pathological and symptomatic manifestations of AD. However, it remains unresolved what kind of molecular abnormalities observed in the pathophysiology of diabetes are causative to the AD pathogenesis in the brain.
One of the key factors that plays a central role in the pathogenesis of obesity and diabetes is the endoplasmic reticulum (ER) stress. Metabolic overload, e.g., high-fat diet (HFD) feeding and overnutrition, is a primary cause of obesity and T2DM, in which the action of insulin on peripheral tissues is attenuated. ER stress due to metabolic overload is considered to be a major culprit in the insulin resistance in metabolic organs, e.g., liver and adipose tissue (15). Under the ER stress conditions, accumulation of misfolded proteins and loss of homeostasis in the ER are initially detected by sensor proteins. This results in the activation of three major unfolded protein response (UPR) signaling pathways, i.e., IRE1-XBP1, PERK-eIF2α-ATF4, and ATF6, which in turn induce the expression of chaperone proteins and repress overall protein translations (16). Several mechanisms that link UPR to insulin resistance through direct inhibition of the insulin signaling pathway have been postulated. It has been shown that IRE1-dependent activation of JNK1 phosphorylates serine residues of insulin receptor substrate 1 (IRS-1) and attenuates insulin receptor signaling (17, 18).
Molecular signatures of ER stress and UPR activation have also been documented in postmortem AD brains (19–22). Accumulation of Aβ and tau, pathogenic proteins in AD, has been proposed to cause ER stress, leading to synaptic dysfunction and neurodegeneration (22–24). These observations have highlighted ER stress as a common pathophysiological feature of the metabolic disorders and AD. However, the question as to whether ER stress elevated in the peripheral tissues or in the central nervous system is causative to the AD changes, remains elusive.
ER stress can be suppressed in vivo by administration of chemical chaperones, e.g. tauroursodeoxycholic acid (TUDCA) and 4-phenyl butyric acid (4-PBA). Treatment with TUDCA and 4-PBA has been shown to improve the systemic metabolic abnormalities in obesity/diabetes models, e.g., ob/ob mice and HFD-fed mice (25–28), supporting the notion that ER stress is causally involved in the pathogenesis of metabolic disorders. Given the correlation between metabolic status and AD pathology, interventions targeting ER stress induced by metabolic overload may also be a potential preventive target for the concomitant worsening of AD pathology.
In this study, we investigated the effects of metabolic improvement by administration of TUDCA on the HFD-induced increase in ER stress and exacerbation of AD pathophysiology in mice overexpressing the amyloid precursor protein (APP) in the brain (A7-Tg mice). Peripheral administration of TUDCA counteracted HFD-induced ER stress and metabolic abnormalities in the periphery, and decreased Aβ levels in the brain, whereas direct brain administration of TUDCA had no effect. Furthermore, age-related exacerbation of ER stress and metabolic abnormalities also were improved by TUDCA, resulting in a marked suppression of brain amyloid accumulation. These results support the view that interventions targeting peripheral ER stress might be effective in reducing Aβ accumulation in the brain caused by obesity and diabetes that are associated with metabolic overload or by aging.


Materials and Methods


A7 transgenic mice (A7-Tg) overexpressing human APP695 harboring KM670/671NL and T714I familial AD mutations under the control of the Thy-1.2 promoter were backcrossed and maintained on a C57BL/6J background (11). Mice were maintained on a 12 h light/dark cycle and provided ad libitum access to water. In the experiments with a high-fat diet, mice were fed standard chow diet (CRF-1, Oriental Yeast Co., Ltd.) until 3 months of age. Thereafter, they were either maintained on the standard chow or switched to a high-fat diet (HFD32, CLEA Japan Inc.) containing 32% fat. The animal care and experimental procedures were approved by the animal experiment committee of The University of Tokyo Graduate School of Medicine.

TUDCA treatment

We performed three TUDCA administration experiments. In the first experiment, 8-month-old HFD-fed A7-Tg mice were intraperitoneally administered TUDCA (250 mg/kg, Merck) or saline twice per day for 30 days, as previously reported (25). In the second experiment, 7- to 8-month-old HFD-fed A7-Tg mice received intracerebroventricular administration of TUDCA (10 µg/day) or phosphate buffered saline (PBS) for 28 days, in accordance with previous reports (29, 30). For intracerebroventricular administration of TUDCA, cannulas (ALZET Brain Infusion Kit 3, ALZET Osmotic Pumps) were stereotaxically implanted into the lateral ventricle (from bregma: anteroposterior -0.5 mm, mediolateral -1.1 mm, dorsoventral -2.5 mm) under anesthesia (0.3 mg/kg of medetomidine, 4.0 mg/kg of midazolam, and 5.0 mg/kg of butorphanol). Then a catheter tube was connected to an osmotic minipump flow moderator (model 2004, ALZET Osmotic Pumps). The minipump was inserted in a subcutaneous pocket on the dorsal surface of the mouse, and the incision was closed. In the third experiment, TUDCA was administered by drinking water to 11-month-old A7-Tg mice for 120 days. The concentration of TUDCA was increased gradually to avoid taste aversion (1 mg/mL for 60 days, 3 mg/mL for 7 days, 5 mg/mL for 14 days, 6.5 mg/mL for 39 days). For the last 31 days, TUDCA was administered intraperitoneally (250 mg/kg) twice per day, 6 days per week, in parallel.


The following antibodies were used: anti-phospho-eIF2α (3398, Cell Signaling Technology (CST)), anti- eIF2α (5324, CST), anti-phospho-JNK (4668, CST), anti-JNK (9258, CST), anti-phospho-PERK (3179, CST), anti-PERK (3192, CST), anti-BiP (610978, BD biosciences), anti-CHOP (ab11419, Abcam), anti-Human APP (C) (18961, IBL), anti-ADAM10 (ab124695, Abcam), anti-BACE1 (5606, CST), anti-α-tubulin (DM1A, Merck), anti-LC3B (3868, CST), anti-Beclin1 (612113, BD biosciences), anti-Sirt1 (2028, CST), anti-PGC-1α (SC-518025, Santa Cruz Biotechnology), anti-Aβ (82E1, IBL), peroxidase-conjugated AffiniPure anti-rabbit IgG (111-035-003, Jackson ImmunoResearch), and peroxidase-conjugated AffiniPure anti-Mouse IgG (515-035-003, Jackson ImmunoResearch).

Metabolic measurements

Blood glucose was measured using Glutest sensor (Sanwa Kagaku Kenkyusho Co., LTD.). For an insulin tolerance test (ITT), 3-hour-fasted mice were intraperitoneally injected with human insulin (Humulin R, Eli Lilly) at 0.75 U/Kg body weight, and blood glucose levels were measured every 20 min for 120 min.

Western blot analysis

Epididymal adipose tissues or periovarian adipose tissues were obtained and weighed. Brains were harvested, dissected into the hypothalamus, hippocampus, and cerebral cortex. These tissue samples were snap frozen in liquid nitrogen and stored at -80 °C until use. Tris-buffered saline (TBS)-soluble fractions were obtained as the supernatant from homogenizing tissues in a 1:10 (w/v) volume of TBS and centrifuging at 347,600 x g for 20 min at 4 °C. TBS-insoluble pellets were homogenized in a 1:10 (w/v) volume of 2% Triton X-100 (TX) in TBS and centrifuged at 347,600 x g for 20 min at 4 °C and supernatants were saved as TX-soluble fractions. RIPA-soluble fcations were obtained as the supernatant from homogenizing tissues in a 1:10 (w/v) volume of RIPA buffer (1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS) in TBS) and centrifuging at 17,800 x g for 30 min at 4 °C. Protein concentration was determined with BCA protein assay kit (Takara Bio Inc.). All the buffers were supplemented with cOmplete protease inhibitor and PhosSTOP phosphatase inhibitor cocktails (Merck). TBS-, TX-, and RIPA-soluble fractions were used for immunoblotting.
For immunoblotting, samples were separated by SDS-polyacrylamide gel electrophoresis under a reducing condition using a Tris-Glycine gel system, transferred to polyvinylidene difluoride membranes (Merck), and incubated with antibodies. The immunoblots were developed using ImmunoStar reagents (Wako) and SuperSignal (Thermo Fisher), and visualized by LAS-4000 mini (Fujifilm).

ELISA quantitation of Aβ

For the measurement of soluble and insoluble Aβ, the TBS-soluble and SDS-insoluble/formic acid-soluble fractions were used, respectively. For extraction of the insoluble fraction, brains were homogenized in a 1:10 (w/v) volume of RIPA buffer, centrifuged at 347,600 x g for 20 min at 4 °C. Resulting pellets were homogenized in a 1:10 (w/v) volume of 2% SDS in TBS, incubated for 30 min at 37 °C and centrifuged at 347,600 x g for 20 min at 20 °C. SDS-insoluble pellets were dissolved in 70% formic acid using a sonicator (Branson), centrifuged at 347,600 x g for 20 min at 4 °C and supernatants were desiccated by Speed-Vac followed by resuspension in dimethyl sulfoxide (DMSO). Levels of Aβ were quantitated by BNT77/BA27 or BNT77/BC05 Human/Rat β Amyloid ELISA kit (Wako). Prior to the measurement of soluble Aβ, an equal volume of 1 M guanidine hydrochloride was added and incubated for 30 min at room temperature.

Immunohistochemical analysis and morphometry

Mouse brains were fixed with 4% paraformaldehyde in PBS for 24 h, dehydrated, and embedded in paraffin. Serial sections were cut at 4-µm thickness. Deparaffinized sections were treated with microwave (700 W) in citrate buffer pH 6.0 for 20 min, followed by digestion with 100 µg/ml proteinase K (Worthington) in TBS for 6 min at 37 °C. After blocking by incubation with 10% calf serum in TBS, the sections were incubated with an anti-Aβ antibody 82E1 and then a biotinylated anti-mouse IgG antibody (Vector Laboratories), followed by visualization by avidin-biotin complex method (ABC elite, Vector Laboratories) using diaminobenzidine as chromogen. The percentage area covered by Aβ immunoreactivity in the parietal cortex/cingulate gyrus, hippocampus, and piriform cortex was measured using Image J software (NIH) as previously described (31).

Quantitative reverse transcription PCR

Total RNA was isolated using TRIzol Plus RNA Purification Kit and PureLink RNA Mini kit (Thermo Fisher). RNA purity and concentration were measured by NanoDrop (ThermoFisher). Total RNA was reverse-transcribed into cDNA using ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO). Real-time PCR was performed with LightCycler 480 system (Roche) using THUNDERBIRD SYBR qPCR Mix (TOYOBO). Threshold cycle values were normalized to Gapdh. The primer pairs used in this study are as follows: 5′- AACGACCCCTTCATTGAC -3′ and 5′- GAAGACACCAGTAGACTCCAC -3′ for Gapdh; 5′- AAGCTATTTCAGTCCCCAGTGG -3′ and 5′- AAGAGCAACCCGAACATGAC -3′ for Map1lc3a; 5′- GACGTGGAGAAAGGCAAGATTG -3′ and 5′- TTGAGCGCTTTTGTCCACTG -3′ for Becn1; 5′- TTGACCGATGGACTCCTCAC -3′ and 5′- AACAAAAGTATATGGACCTATCCGC -3′ for Sirt1.

Statistical analysis

Quantitative data were analyzed statistically by unpaired t test for two-group data, or one-way ANOVA followed by Tukey’s multiple comparisons test for three-group data using GraphPad Prism 7. In figures, all data are represented by mean ± SEM. Statistical significance is indicated by *p < 0.05, ** p < 0.01, and *** p < 0.001.



Intraperitoneal administration of TUDCA attenuated diet-induced peripheral ER stress and improved systemic metabolic abnormalities in A7-Tg mice

To study the link between the peripheral metabolic state and brain AD pathology, we have used A7-Tg mice expressing human APP harboring the Swedish and Austrian mutations in neurons. A7-Tg mice develop progressive Aβ deposition in the brain starting at ~12 months of age, and we have previously shown that inducing obesity/diabetes by feeding HFD accelerates amyloid pathology in the brains of A7-Tg mice (11). Furthermore, HFD-induced acceleration of Aβ pathology was reversibly suppressed in dietary intervention experiments that switched from HFD to normal diet (ND) either starting at the age of 9 months (11) or 15 months (Figure S1), as the systemic metabolic abnormalities were improved. These results indicate that metabolic improvement is effective in reversing the HFD-induced Aβ deposition.
We investigated the effects of mitigating ER stress on metabolic state and consequently amyloid pathology using a chemical chaperone TUDCA, which has been shown to be effective in improving HFD-induced metabolic abnormalities (25). TUDCA was intraperitoneally administered to HFD-fed A7-Tg mice for 30 days starting at 8 months of age (Figure 1a). HFD feeding increased body weights and wet weights of adipose tissues compared to normal diet (ND) feeding, whereas TUDCA treatment significantly alleviated HFD-induced obesity in A7-Tg mice (Figure 1b-d). HFD-fed A7-Tg mice exhibited significant hyperglycemia, which was associated with a decreased insulin sensitivity as shown by the insulin tolerance test (ITT) (Figure 1e-g). Administration of TUDCA also improved the impaired glucose metabolism to a level comparable to that of ND-fed mice (Figure 1e-g). These data suggest that TUDCA could ameliorate diet-induced obesity and diabetes-like metabolic abnormalities in A7-Tg mice.

Figure 1. Intraperitoneal administration of TUDCA improved diet-induced metabolic abnormalities and reduced peripheral ER stress

(a) Schematic diagram of the study design. Male A7-Tg mice were fed with normal diet (ND) or high-fat diet (HFD) from 3 months of age, and TUDCA (250 mg/kg) or saline (vehicle) was administered intraperitoneally twice per day for 30 days from 8 months of age. (b) The time course of body weight changes during TUDCA treatment. (c-d) Body weight (c) and adipose tissue wet weight (d) on day 31 after starting TUDCA treatment. (e-g) Changes in the blood glucose levels (e) and area under the curve (AUC) (f) during the insulin tolerance test (ITT) on day 15 or 16, and blood glucose levels before the test (g). (h-i) Immunoblot analyses of phosphorylation levels of eIF2α and JNK in the RIPA-soluble fractions of liver (h) and adipose tissue (i) (upper panels). The ratios of phosphorylation to total protein content were measured by densitometry (lower panels). Data are mean ± SEM (ND: n = 8, HFD: n = 5, HFD-TUDCA: n = 6). *, p < 0.05; **, p < 0.01; ***, p < 0.001, one-way ANOVA with Tukey’s post-hoc test.


Previous studies have shown that TUDCA treatment ameliorates insulin resistance and abnormal glucose metabolism via the reduction of ER stress in the liver and adipose tissue of obese mice (25, 28). We therefore examined the effects of HFD feeding and TUDCA treatment on UPR in the liver and adipose tissues in A7-Tg mice. HFD feeding induced phosphorylation of eIF2α and JNK in the liver, which was significantly decreased by TUDCA treatment, in HFD-fed A7-Tg mice (Figure 1h). In the adipose tissue, phospho-eIF2α showed an increasing trend with HFD feeding, which was not inhibited by TUDCA treatment (Figure 1i). On the other hand, phospho-JNK was markedly increased by HFD and decreased significantly by TUDCA treatment, as in the liver (Figure 1i).

Intraperitoneal administration of TUDCA prevented the HFD-induced exacerbation of Aβ accumulation in brains

We next examined the effects of TUDCA on HFD-induced exacerbation of Aβ pathology in the cerebral cortices of A7-Tg mice. As we previously reported (11), HFD-feeding significantly increased the levels of both Aβ40 and Aβ42 at 9 months of age (Figure 2a). Administration of TUDCA reduced the Aβ levels in the HFD group to a similar level to that of control ND group, indicating that the effect of metabolic overload on Aβ pathology was totally reversed by TUDCA (Figure 2a). The levels of APP fragments were not altered by HFD feeding, whereas the levels of CTFβ, a carboxy-terminal fragment produced upon β-cleavage, were decreased by TUDCA treatment (Figure 2b). Although the protein levels of ADAM10 and BACE1, corresponding to α- and β-secretases, respectively, did not change in any of the experimental groups (Figure 2c), the reduced levels of CTFβ might be related to the inhibitory effect of TUDCA on Aβ production.

Figure 2. Intraperitoneal administration of TUDCA prevented the HFD-induced increase of brain Aβ accumulation

(a) The levels of TBS-soluble Aβ40 and Aβ42 in the cerebral cortices of 9-month-old A7-Tg mice in the three experimental groups indicated in Figure 1a (ND, HFD, and HFD-TUDCA) were analyzed by ELISA. (b) Immunoblot analyses of full-length APP (APP FL), APP-CTFα, APP-CTFβ, and α-tubulin in the TX-soluble fractions of cerebral cortex. The lower graph shows the results of densitometry. The amount of protein was expressed as a relative value to the ND group. (c) Immunoblot and densitometric analyses of ADAM10 and BACE1 in the TX-soluble fraction of cerebral cortex. (d) Immunoblot analyses of peIF2α, pJNK, total eIF2α, total JNK, Grp78/BiP, CHOP, and α-tubulin in the TBS-soluble fractions of cerebral cortex (left panels). The right panel shows the results of densitometry. The levels of peIF2α and pJNK were normalized to total eIF2α and JNK, respectively; those of Grp78/BiP and CHOP were normalized to α-tubulin. Data are mean ± SEM (ND: n = 8, HFD: n = 5, HFD-TUDCA: n = 6). *, p < 0.05; **, p < 0.01; ***, p < 0.001, one-way ANOVA with Tukey’s post-hoc test.


To investigate the relationship between changes in Aβ levels and ER stress in the brain, we analyzed the expression of ER stress marker proteins phospho-eIF2α, phospho-JNK, Grp78/Bip, and CHOP in the cerebral cortex. In contrast to the peripheral tissues, no increase in the expression of ER stress marker proteins was observed upon HFD feeding in the cerebral cortices (Figure 2d), and TUDCA treatment did not alter the levels of any of these proteins (Figure 2d). We also examined the mRNA expression of TNFα in the hippocampus to investigate the possibility that TUDCA treatment affects inflammatory signals in the brain. The results showed that there was no significant difference in relative expression levels among the ND (1.000 ± 0.151), HFD (0.866 ± 0.115), and HFD-TUDCA (1.070 ± 0.176) groups. These results suggest that metabolic improvement through reduction of the peripheral ER stress by TUDCA may prevent HFD-induced exacerbation of Aβ pathology in brains.

Intracerebroventricular administration of TUDCA had no effect on the HFD-induced increase in Aβ levels in the cerebral cortex of A7-Tg mice

Previous studies have shown that TUDCA is able to cross the blood-brain barrier and exert the effects on the central nervous system tissues (32, 33). This suggests that the effects of intraperitoneally administered TUDCA on Aβ pathogenesis may be due to either an indirect effect of TUDCA in the periphery or a direct effect of TUDCA translocated to the brain. To examine the latter possibility, we directly administered TUDCA into the cerebral ventricules (i.c.v.) of HFD-fed A7-Tg mice for 28 days (Figure 3a).
Intraventricular administration of TUDCA decreased the blood glucose levels, but did not improve the HFD-induced obesity (Figure 3b-d). No change in UPR of the liver or adipose tissue of HFD-fed A7-Tg mice were observed (Figure 3e-f). In addition, UPR in the cerebral cortex also was not altered by intraventricular administration of TUDCA (Figure 3g). To confirm whether TUDCA reached effective concentrations in the brain, we evaluated the ER stress markers in the hypothalamus and found that the levels of phospho-JNK and Grp78/BiP were significantly reduced (Figure S2). Given that HFD-induced hypothalamic ER stress is known to affect the peripheral metabolic state (26, 34), this reduction of ER stress by intraventricular administration of TUDCA might have resulted in a decrease in blood glucose levels. Under these conditions, intraventricular administration of TUDCA did not reduce the levels of Aβ40 and Aβ42 in HFD-fed A7-Tg mice (Figure 3h). Taken together with the fact that HFD feeding did not enhance ER stress in the brain (Figure 2d), we reasoned that the central effect of TUDCA did not contribute much to the suppression of Aβ levels in the cerebral cortex of HFD-fed A7-Tg mice, and speculated that the effect on ER stress in the periphery was more important.

Figure 3. Intracerebroventricular administration of TUDCA did not affect ER stress and Aβ levels in HFD-fed A7-Tg mice

(a) Schematic diagram of the study design. Female A7-Tg mice were fed with HFD from 3 months of age, and TUDCA (10 µg/day) or PBS (vehicle) was intracerebroventricularly administered using ALZET Osmotic Pumps for 28 days from 7-8 months of age. (b-d) Body weight (b), adipose tissue wet weight (c), and blood glucose levels (d) on day 28 after starting TUDCA treatment. (e-f) Immunoblot and densitomeric analyses of phosphorylation levels of eIF2α and JNK in the RIPA-soluble fractions of liver (e) and adipose tissue (f). (g) Immunoblot and densitomeric analyses of phosphorylation levels of eIF2α and JNK, and the levels of Grp78/BiP, CHOP in the TBS-soluble fractions of cerebral cortex. The results were analyzed as in Figure 2d. (h) The levels of TBS-soluble Aβ40 and Aβ42 in the cerebral cortices at 9 months of age were analyzed by ELISA. Data are mean ± SEM (HFD: n = 8, HFD-TUDCA: n = 9). *, p < 0.05, unpaired t-test.


Peripheral administration of TUDCA improved age-related ER stress and metabolic abnormalities in A7-Tg mice

Both in humans and animal models, increasing adiposity and insulin resistance have been documented as the characteristics of aging. Furthermore, age-related decline in the UPR has also been suggested (35). We therefore wondered whether aging-related ER stress and the associated metabolic abnormalities might have contributed to the aggravation of amyloid pathology in the brain. To test this hypothesis, we examined the effects of TUDCA in A7-Tg mice raised on normal diet. Because A7-Tg mice require a long period of time, i.e., >12 months, by the accumulation of amyloid plaques, a long-term administration method was adopted. We administered TUDCA orally (1-6.5 mg/ml in drinking water) starting at 11 months of age and intraperitoneally in parallel starting at 14 months of age, and analyzed the mice at 15 months (Figure 4a).
TUDCA treatment significantly decreased body weight and wet weight of adipose tissue in these aged mice (Figure 4b-c). Insulin sensitivity also was improved as demonstrated by the ITT, but blood glucose levels were not significantly decreased by TUDCA treatment (Figure 4d-f). Notably, the metabolic status of TUDCA-treated 15-month-old A7-Tg mice (body weight: 33.6 ± 1.0 g, adipose tissue wet weight: 0.73 ± 0.06 g, area under the curve (AUC) for ITT: 8691 ± 374 mg/dl) was improved to levels equivalent to those in 9-month-old ND-fed A7-Tg mice (body weight: 31.8 ± 1.1 g, adipose tissue wet weight: 0.76 ± 0.10 g, AUC for ITT: 9755 ± 599 mg/dl, see Figure 1).
We next examined the ER stress in the peripheral tissues of these mice. In contrast to the results of HFD-fed A7-Tg mice (Figure 1), TUDCA treatment on 15-month-old A7-Tg mice did not alter the expression of UPR markers in the liver (Figure 4g). In contrast, the levels of phospho-JNK were significantly decreased, and phospho-eIF2α also showed a decreasing trend in the adipose tissue (Figure 4h). These results suggest that ER stress in adipose tissue, rather than that in liver, contributed to the age-related metabolic abnormalities, which can be ameliorated by TUDCA treatment.

Figure 4. Peripheral administration of TUDCA improved age-related metabolic abnormalities and reduced peripheral ER stress

(a) Schematic diagram of the study design. TUDCA was administered to 11-month-old male A7-Tg mice by drinking water for 120 days with a gradual increase in concentration to avoid taste aversion (1 mg/ml for 60 days, 3 mg/ml for 7 days, 5 mg/ml for 14 days, 6.5 mg/ml for 39 days). For the last 31 days, TUDCA (250 mg/kg) or saline was administered intraperitoneally twice per day, 6 days per week, in parallel. (b-c) Body weight (b) and adipose tissue wet weight (c) on day 120 after starting TUDCA treatment. (d-f) Changes in the blood glucose levels (d) and area under the curve (AUC) (e) during ITT on day 97 or 100, and blood glucose levels before the test (f). (e-f) Immunoblot and densitometric analyses of phosphorylation levels of eIF2α and JNK in the RIPA-soluble fractions of liver (g) and adipose tissue (h). Data are mean ± SEM (ND: n = 8, ND-TUDCA: n = 7). *, p < 0.05; **, p < 0.01; ***, p < 0.001, unpaired t-test.


Peripheral administration of TUDCA reduced amyloid deposition in the brain of 15-month-old A7-Tg mice

We then evaluated the effects of TUDCA on age-dependent amyloid accumulation in the brains of 15-month-old A7-Tg mice. Biochemical analyses revealed that the levels of insoluble Aβ40 and Aβ42 in the cerebral cortex were significantly decreased by TUDCA treatment (Figure 5a). Immunohistochemical analyses showed that Aβ plaques were significantly reduced in the piriform and cerebral cortices (Figure 5b-c). Aβ deposition in the hippocampus also showed a tendency to decrease (Figure 5d). Analyses of proteins related to Aβ production in the brains of this experimental group showed that TUDCA treatment caused a downward trend in the level of CTFβ, but no significant change was observed (Figure S3a). The UPR activities of the cerebral cortex was not changed by TUDCA treatment (Figure 5e). Overall, the reduction of peripheral ER stress and improvement of metabolism during aging by TUDCA treatment also attenuated the formation of amyloid pathology in the brain.

Figure 5. Peripheral administration of TUDCA reduced amyloid deposition in 15-month-old A7-Tg mice

(a) The levels of SDS-insoluble/formic acid-soluble Aβ40 and Aβ42 in the cerebral cortices of 15-month-old A7-Tg mice. (b-d) Immunohistochemical analyses of 15-month-old A7-Tg mouse brains using an anti-Aβ (82E1) antibody. Representative images of brain regions including the piriform cortex (b), parietal cortex and cingulate gyrus (c), and hippocampus (d) are shown. The accompanying graphs represent the quantitative results of amyloid deposition in percentage of the area covered by Aβ immunoreactivity. (e) Immunoblot and densitomeric analyses of phosphorylation levels of eIF2α and JNK, and the levels of Grp78/BiP, CHOP in the RIPA-soluble fractions of cerebral cortex. Data are mean ± SEM (ND: n = 8, ND-TUDCA: n = 7). *, p < 0.05; ***, p < 0.001, unpaired t-test. Scale bars: 1.0 mm.


TUDCA treatment caused dietary restriction-like expression changes both in the peripheral tissues and brain

The effects of TUDCA observed in this present study, i.e., inhibition of metabolic abnormalities and amyloid pathology associated with obesity/T2DM and aging, were similar to those caused by dietary restriction (11–13). Because dietary restriction has been documented to ameliorate the age-related abnormalities in several species through increased sirtuin expression and authophagy (36, 37), we investigated the expression of Sirt1 and genes related to autophagy, e.g. Map1lc3a and Becn1, in 15-months-old A7-Tg mice.
In the liver, peripheral TUDCA treatment showed a tendency to increase Sirt1 mRNA (Figure S3b). In adipose tissue, mRNA expression of Becn1 and Sirt1 was increased, and Map1lc3a showed an increasing trend (Figure S3c). In the hippocampus, TUDCA treatment increased the mRNA expression levels of Map1lc3a, Becn1, and Sirt1 (Figure S3d). Furthermore, protein expression analysis showed that the LC3B-II/LC3B-I ratio, which indicates activation of autophagy, and the levels of Sirt1 and its substrate, PGC-1α (Figure S3e) were increased in the cerebral cortex of 15-month-old A7-Tg mice. Taken together, we reasoned that reducing ER stress in the peripheral tissues by systemic administration of TUDCA has effects that mimic dietary restriction both on the peripheral tissues and the brain, which in turn might reduce brain Aβ accumulation.



Besides causing metabolic disturbances systemically, HFD feeding has been shown to promote Aβ accumulation in the brain in mouse models of AD (9, 11, 12). Here, we showed for the first time that intraperitoneal administration of TUDCA ameliorates not only the metabolic abnormalities caused by HFD, but also the concomitant increase in brain Aβ pathology, in A7-Tg mice. The most probable mechanism of TUDCA action would be that its chaperone activity reduced ER stress in the peripheral tissues, thereby ameliorating metabolic abnormalities such as obesity and insulin resistance, and consequently alleviating brain amyloid pathology. This hypothesis is consistent with the previous observations that Aβ accumulation is reversibly suppressed when HFD-dependent peripheral metabolic abnormalities are improved by dietary switching (11–13). Importantly, we showed that even under continuous HFD feeding, TUDCA treatment improved metabolism and lowered Aβ to a level similar to that in the ND diet group.
Since TUDCA is a brain-penetrating compound that has been shown to be neuroprotective through its anti-apoptotic effects (33, 38), another possibility for the mechanism of action is that central ER stress may have been targeted to reduce Aβ levels. Some studies in human AD postmortem brains have reported findings suggestive of increased ER stress, and the link between Aβ-induced toxicity and ER stress has been reported in various experimental models (39, 40). Thus, central ER stress may be augmented as a feedback mechanism for the pathological progression of AD, which may further contribute to the exacerbation of AD pathology. Furthermore, metabolic overload has been suggested to increase brain ER stress, especially in the hypothalamus, the latter being exposed to peripheral circulation (26, 34). Several studies have reported HFD-dependent UPR enhancement in brain regions including the hippocampus and cerebral cortex (41, 42). In our experiment, HFD feeding increased ER stress in the liver and adipose tissue, which was lowered by TUDCA. However, neither HFD feeding nor TUDCA treatment altered the UPR signaling activities in the cerebral cortex. Our results may have differed from those previously reported due to multiple factors, e.g., differences in the nutritional composition of the diet, the rearing environment, and the rate of pathological progression that varies among model animals. Nevertheless, our data may suggest that the cerebral cortex is less prone to elevated ER stress due to metabolic overload. Furthermore, even in conditions where the UPR activity was not increased, the Aβ pathology was exacerbated by HFD feeding. Thus, it may be reasonable to speculate that cortical ER stress is not the major culprit for the HFD-induced Aβ accumulation. Moreover, the finding that central administration of TUDCA did not alter brain Aβ levels supports the lack of interdependence between ER stress in the brain and HFD-induced Aβ accumulation. Although the UPR in the hypothalamus was significantly reduced, it cannot be ruled out that TUDCA diffused into the cerebral cortex in this administration paradigm may not have reached a sufficient concentration to exert an effect on AD pathology.
Aging is the greatest risk factor of AD. In our study, TUDCA ameliorated not only the diet-dependent but also the age-related metabolic deterioration, and suppressed amyloid accumulation. In 15-month-old A7-Tg mice, unlike HFD-fed A7-Tg mice, TUDCA did not alter the expression of UPR marker proteins in the liver, but reduced the levels of phospho-JNK and phspho-eIF2α in adipose tissues. It has been documented that diet-induced obese mice had a greater infiltration of macrophages and more pro-inflammatory immune cells in the liver than in aged obese mice, whereas adipose tissues showed similar levels of cytokine changes (43), which is consistent with our findings, prompting us to speculate that suppression of stress in the adipose tissue may have ameliorated the age-related, systemic metabolic deterioration.
In a series of studies using APP/PS1 mice, Rodrigues et al. have shown that treatment with TUDCA reduced Aβ production, suppressed amyloid accumulation, and prevented cognitive impairments (44–46). However, a reduction in amyloid deposits by TUDCA has been observed in APP/PS1 mice at a young age, without significant changes in body weight (45). Furthermore, it has previously been reported that intraperitoneal administration of TUDCA to young lean mice did not affect peripheral metabolic parameters (25). These results do not support the hypothesis that the inhibitory effect of TUDCA on Aβ accumulation is due to amelioration of the systemic metabolic abnormalities. However, it should be noted that ER stress might have been upregulated only in the brains of AD model mice that overexpress presenilin 1 together with APP, including APP/PS1 mice, raising serious concerns about its use in the study of ER stress (47, 48). Accordingly, upregulation of ER stress has not been described in other AD model mice, e.g. Tg2576, APP23, and AppNL-G-F mice (48–51). Thus, the effect of TUDCA on amyloid reduction in APP/PS1 mice might be attributable, at least in part, to the reduction of central ER stress. Future studies on the effects of TUDCA in young A7-Tg mice will better address these issues.
We found that the ratio of LC3BII/LC3BI and the expression of Sirt1 and PGC-1α are factors that are altered both in the periphery and brain upon TUDCA-treatment in A7-Tg mice. These molecules are involved in the molecular pathways that play an important role in the dietary restriction that delays the onset of many chronic diseases such as obesity, diabetes, and AD, as well as the anti-aging effects of its mimetic, resveratrol (52, 53). Sirt1 activation, or acetylation of PGC-1α by Sirt1, has been suggested to alter Aβ levels in the brain (54–56). In addition, increased autophagy has been suggested to reduce Aβ levels (57). This suggests that TUDCA may regulate Aβ pathology by suppressing ER stress in the periphery, thereby exerting a dietary restriction-like effect. Which of the processes underlying Aβ accumulation, i.e., production, clearance or aggregation, were altered by TUDCA is yet to be clarified. The levels of CTFβ, an indicator of β-secretase activity, tended to be decreased by TUDCA in HFD-fed A7-Tg mice, suggesting that reduced Aβ production may contribute at least in part to the anti-amyloid effect. On the other hand, we previously showed that HFD feeding decreases the clearance of Aβ in the brain interstitial fluid (11). Since TUDCA ameliorated the adverse effects of HFD on the systemic metabolism, it is also possible that Aβ clearance was improved accordingly. Further studies are needed to elucidate the molecular link between the peripheral metabolism and signal changes in the brain, and brain Aβ dynamics.
Overall, our results suggest a therapeutic potential of TUDCA in suppressing obesity/diabetes-induced Aβ accumulation. Recent progress in imaging biomarker research has revealed that amyloid accumulation occurs in the brains of AD patients decades before the onset of dementia (59). Thus, preclinical stage of pathological progression may be a critical period for disease prevention. Our results showed that interventions in the periphery at the early stage of AD pathology formation may be effective in counteracting Aβ accumulation in the brain. Considering the biosafety and its ability to inhibit apoptosis and neuroinflammation (38), TUDCA is expected to be a multifaceted prevention strategy of AD.


Funding: This work was supported by AMED under Grant Number JP20dm0107056, and JSPS KAKENHI Grant Number JP20H00525.

Declaration of Competing Interest: T.O. is an employee of Kaken Pharmaceutical Co., LTD. All other authors declare no conflict of interests.





1. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396, 413–446
2. Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential for primary prevention of Alzheimer’s disease: an analysis of population-based data. Lancet Neurol. 2014;13, 788–794
3. Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol. 2011;10, 819–828
4. Gudala K, Bansal D, Schifano F, Bhansali A. (2013) Diabetes mellitus and risk of dementia: A meta-analysis of prospective observational studies. J Diabetes Investig. 2013;4, 640–650
5. Zhang J, Chen C, Hua S, Liao H, Wang M, Xiong Y, Cao F. (2017) An updated meta-analysis of cohort studies: Diabetes and risk of Alzheimer’s disease. Diabetes Res Clin Pract. 124, 41-47
6. Kandimalla R, Thirumala V, Reddy PH. Is Alzheimer’s disease a Type 3 Diabetes? A critical appraisal. Biochim Biophys Acta Mol Basis Dis. 2017;1863, 1078–1089
7. Biessels GJ, Despa F. Cognitive decline and dementia in diabetes mellitus: mechanisms and clinical implications. Nat Rev Endocrinol. 2018;14, 591–604
8. Matsuzaki T, Sasaki K, Tanizaki Y, et al. Insulin resistance is associated with the pathology of Alzheimer disease: The Hisayama Study. Neurology. 2010;75, 764–770
9. Ho L, Qin W, Pompl PN, et al. Diet-induced insulin resistance promotes amyloidosis in a transgenic mouse model of Alzheimer’s disease. FASEB J. 2004;18, 902–904
10. Cao D, Lu H, Lewis TL, Li L. Intake of sucrose-sweetened water induces insulin resistance and exacerbates memory deficits and amyloidosis in a transgenic mouse model of Alzheimer disease. J Biol Chem. 2007;282, 36275–36282
11. Wakabayashi T, Yamaguchi K, Matsui K, et al. Differential effects of diet- and genetically-induced brain insulin resistance on amyloid pathology in a mouse model of Alzheimer’s disease. Mol Neurodegener. 2019;14, 15
12. Maesako M, Uemura K, Kubota M, et al. Exercise is more effective than diet control in preventing high fat diet-induced β-amyloid deposition and memory deficit in amyloid precursor protein transgenic mice. J Biol Chem. 2012;287, 23024–23033
13. Walker JM, Dixit S, Saulsberry AC, May JM, Harrison FE. Reversal of high fat diet-induced obesity improves glucose tolerance, inflammatory response, β-amyloid accumulation and cognitive decline in the APP/PSEN1 mouse model of Alzheimer’s disease. Neurobiol Dis. 2017;100, 87–98
14. Selkoe DJ, Hardy J. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol Med. 2016;8, 595–608
15. Villalobos-Labra R, Subiabre M, Toledo F, Pardo F, Sobrevia L. Endoplasmic reticulum stress and development of insulin resistance in adipose, skeletal, liver, and foetoplacental tissue in diabesity. Mol Aspects of Med. 2019;66, 49–61
16. Walter P, Ron D. The unfolded protein response: from stress pathway to homeostatic regulation. Science. 2011;334, 1081–1086
17. Hirosumi J, Tuncman G, Chang L, et al. (2002) A central role for JNK in obesity and insulin resistance. Nature. 2002;420, 333–336
18. Özcan U, Cao Q, Yilmaz E, et al. Endoplasmic reticulum stress links obesity, insulin action, and type 2 diabetes. Science. 2004;306, 457–461
19. Hoozemans JJM, Veerhuis R, Van Haastert ES, et al. The unfolded protein response is activated in Alzheimer’s disease. Acta Neuropathol. 2005;110, 165–172
20. Hoozemans JJM, van Haastert ES, Nijholt DAT, et al. The unfolded protein response is activated in pretangle neurons in Alzheimer’s disease hippocampus. Am J Pathol. 2009;174, 1241–1251
21. Stutzbach LD, Xie SX, Naj AC, et al. The unfolded protein response is activated in disease-affected brain regions in progressive supranuclear palsy and Alzheimer’s disease. Acta Neuropathol Commun. 2013;1, 31
22. Yoon SO, Park DJ, Ryu JC, et al. JNK3 perpetuates metabolic stress induced by Aβ peptides. Neuron. 2012;75, 824–837
23. Duran-Aniotz C, Martinez G, Hetz C. Memory loss in Alzheimer’s disease: are the alterations in the UPR network involved in the cognitive impairment? Front. Aging Neurosci. 2014;6, 8
24. Gerakis Y, Hetz C. Emerging roles of ER stress in the etiology and pathogenesis of Alzheimer’s disease. FEBS J. 2018;285, 995–1011
25. Özcan U, Yilmaz E, Özcan L, et al. Chemical chaperones reduce ER stress and restore glucose homeostasis in a mouse model of type 2 diabetes. Science. 2006;313, 1137–1140
26. Ozcan L, Ergin AS, Lu A, et al. Endoplasmic reticulum stress plays a central role in development of leptin resistance. Cell Metab. 2009;9, 35–51
27. Vettorazzi JF, Kurauti MA, Soares GM, et al. Bile acid TUDCA improves insulin clearance by increasing the expression of insulin-degrading enzyme in the liver of obese mice. Sci Rep. 2017;7, 14876
28. Zhang Z, Wang X, Zheng G, et al. Troxerutin attenuates enhancement of hepatic gluconeogenesis by inhibiting NOD activation-mediated inflammation in high-fat diet-treated mice. Int J Mol Sci. 2017;18, 31
29. DeVos SL, Miller TM. Direct intraventricular delivery of drugs to the rodent central nervous system. J Vis Exp. 2013;e50326
30. Contreras C, González-García I, Seoane-Collazo P, et al. Reduction of hypothalamic endoplasmic reticulum stress activates browning of white fat and ameliorates obesity. Diabetes. 2017;66, 87–99
31. Yamamoto K, Tanei Z, Hashimoto T, et al. Chronic optogenetic activation augments Aβ pathology in a mouse model of Alzheimer disease. Cell Rep. 2015;11, 859–865
32. Kaemmerer WF, Rodrigues CMP, Steer CJ, Low WC. Creatine-supplemented diet extends Purkinje cell survival in spinocerebellar ataxia type 1 transgenic mice but does not prevent the ataxic phenotype. Neuroscience. 2001;103, 713–724
33. Keene CD, Rodrigues CMP, Eich T, et al. Tauroursodeoxycholic acid, a bile acid, is neuroprotective in a transgenic animal model of Huntington’s disease. Proc Natl Acad Sci U S A. 2002;99, 10671–10676
34. Zhang X, Zhang G, Zhang H, et al. Hypothalamic IKKβ/NF-κB and ER stress link overnutrition to energy imbalance and obesity. Cell. 2008;135, 61–73
35. Naidoo N, Brown M. The endoplasmic reticulum stress response in aging and age-related diseases. Front Physiol. 2012;3, 263
36. Fontana L, Partridge L. Promoting health and longevity through diet: from model organisms to humans. Cell. 2015;161, 106–118
37. Van Cauwenberghe C, Vandendriessche C, Libert C, Vandenbroucke RE. Caloric restriction: beneficial effects on brain aging and Alzheimer’s disease. Mamm Genome. 2016;27, 300–319
38. Zangerolamo L, Vettorazzi JF, Rosa LRO, Carneiro EM, Barbosa HCL. The bile acid TUDCA and neurodegenerative disorders: An overview. Life Sci. 2021;272, 119252
39. Uddin MS, Tewari D, Sharma G, et al. Molecular mechanisms of ER stress and UPR in the pathogenesis of Alzheimer’s disease. Mol Neurobiol. 2020;57, 2902–2919
40. Halliday M, Mallucci GR. Review: Modulating the unfolded protein response to prevent neurodegeneration and enhance memory. Neuropathol Appl Neurobiol. 2015;41, 414–427
41. Liang L, Chen J, Zhan L, et al. Endoplasmic reticulum stress impairs insulin receptor signaling in the brains of obese rats. PLoS One. 2015;10, e0126384
42. Binayi F, Zardooz H, Ghasemi R, et al. The chemical chaperon 4-phenyl butyric acid restored high-fat diet- induced hippocampal insulin content and insulin receptor level reduction along with spatial learning and memory deficits in male rats. Physiol Behav. 2021;231, 113312
43. Krishna KB, Stefanovic-Racic M, Dedousis N, Sipula I, O’Doherty RM. Similar degrees of obesity induced by diet or aging cause strikingly different immunologic and metabolic outcomes. Physiol Rep. 2016;4, e12708
44. Nunes AF, Amaral JD, Lo AC, et al. TUDCA, a bile acid, attenuates amyloid precursor protein processing and amyloid-β deposition in APP/PS1 mice. Mol Neurobiol. 2012;45, 440–454
45. Lo AC, Callaerts-Vegh Z, Nunes AF, Rodrigues CMP, D’Hooge R. Tauroursodeoxycholic acid (TUDCA) supplementation prevents cognitive impairment and amyloid deposition in APP/PS1 mice. Neurobiol Dis. 2013;50, 21–29
46. Dionísio PA, Amaral JD, Ribeiro MF, et al. Amyloid-β pathology is attenuated by tauroursodeoxycholic acid treatment in APP/PS1 mice after disease onset. Neurobiol Aging. 2015;36, 228–240
47. Hashimoto S, Ishii A, Kamano N, et al. Endoplasmic reticulum stress responses in mouse models of Alzheimer’s disease: Overexpression paradigm versus knockin paradigm. J Biol Chem. 2018;293, 3118–3125
48. Hashimoto S, Saido TC. Critical review: involvement of endoplasmic reticulum stress in the aetiology of Alzheimer’s disease. Open Biol. 8, 180024
49. Lee JH, Won SM, Suh J, et al. Induction of the unfolded protein response and cell death pathway in Alzheimer’s disease, but not in aged Tg2576 mice. Exp Mol Med. 2010;42, 386–394
50. Ma T, Trinh MA, Wexler AJ, et al. Suppression of eIF2α kinases alleviates Alzheimer’s disease–related plasticity and memory deficits. Nat Neurosci. 2013;16, 1299–1305
51. Barbero-Camps E, Fernández A, Baulies A, et al. Endoplasmic reticulum stress mediates amyloid β neurotoxicity via mitochondrial cholesterol trafficking. Am J Pathol. 2014;184, 2066–2081
52. Guarente L. Calorie restriction and sirtuins revisited. Genes Dev. 2013;27, 2072–2085
53. Madeo F, Zimmermann A, Maiuri MC, Kroemer G. Essential role for autophagy in life span extension. J Clin Invest. 2015;125, 85–93
54. Katsouri L, Parr C, Bogdanovic N, Willem M, Sastre M. PPARγ co-activator-1α (PGC-1α) reduces amyloid-β generation through a PPARγ-dependent mechanism. J Alzheimers Dis. 2011;25, 151–162
55. Wang R, Li JJ, Diao S, et al. Metabolic stress modulates Alzheimer’s β-secretase gene transcription via SIRT1-PPARγ-PGC-1 in neurons. Cell Metab. 2013;17, 685–694
56. Qin W, Yang T, Ho L, et al. Neuronal SIRT1 activation as a novel mechanism underlying the prevention of Alzheimer disease amyloid neuropathology by calorie restriction. J Biol Chem. 2006;281, 21745–21754
57. Menzies FM, Fleming A, Caricasole A, et al. Autophagy and neurodegeneration: pathogenic mechanisms and therapeutic opportunities. Neuron. 2017; 93, 1015–1034
58. Pluvinage JV, Wyss-Coray T. Systemic factors as mediators of brain homeostasis, ageing and neurodegeneration. Nat Rev Neurosci. 2020;21, 93–102
59. Jack CR, Knopman DS, Jagust WJ, et al. Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol. 2010;9, 119–128



T. Mano1, K. Sato1, T. Ikeuchi2, T. Toda1, T. Iwatsubo3, A. Iwata4, Japanese Alzheimer’s Disease Neuroimaging Initiative5


1. Department of Neurology, Graduate School of Medicine, The University of Tokyo, Bunkyo-ku, Tokyo, Japan; 2. Department of Molecular Genetics, Brain Research Institute, Niigata University, Chuo-ku, Niigata, Japan; 3. Department of Neuropathology, Graduate School of Medicine, The University of Tokyo, Bunkyo-ku, Tokyo, Japan; 4. Tokyo Metropolitan Geriatric Hospital and Institute of Gerontology, Itabashi-ku, Tokyo, Japan; 5. The full list of members of the Japanese Alzheimer’s Disease Neuroimaging Initiative is provided in the supplementary file, «J-ADNI co-investigators».

Corresponding Author: Tatsuo Mano, Department of Neurology, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-8655, Japan, Email:, Phone +81-3-5800-8672, Fax +81-3-5800-6548

J Prev Alz Dis 2021;
Published online June 9, 2021,



BACKGROUND: Recent biomarker studies demonstrated that the central nervous system (CNS) environment can be observed from peripherally-derived samples. In a previous study, we demonstrated significant hypomethylation of the BRCA1 promoter region in neuronal cells from post-mortem brains of Alzheimer’s disease patients through neuron-specific methylome analysis. Thus, we investigate the methylation changes in the BRCA1 promoter region in the blood samples.
OBJECTIVES: To analyze the methylation level of the BRCA1 promoter in peripheral blood from AD patients and normal controls.
DESIGN, SETTING, PARTICIPANTS: Genomic DNA samples from peripheral blood were obtained from the J-ADNI repository, and their biomarker data were obtained J-ADNI from the National Bioscience Database Center. Genomic DNA samples from an independent cohort for validation was obtained from Niigata University Hospital (Niigata, Japan). Amyloid positivity was defied by visual inspection of amyloid PET or a CSF Aβ42 value ≤ 333 pg/mL at the baseline.
MEASUREMENTS: Methylation level of the BRCA1 promoter was analyzed by pyrosequencing.
RESULTS: Compared to normal controls, methylation of the BRCA1 promoter in AD patients was not significantly changed; however, in AD patients, it showed a positive correlation with AD risk factors.
CONCLUSIONS: Our data confirmed the importance of cell-type specific methylome analysis and also suggested that environmental changes in the CNS can be detected by observing the peripheral blood, implying that the peripheral BRCA1 methylation level can be a surrogate for AD.

Key words: Alzheimer’s disease, methylome, BRCA1, peripheral blood.

Abbreviations: AD: Alzheimer’s disease; Aβ: amyloid β; CNS: central nervous system; NC: normal control; NFT: neurofibrillary tangle; PET: positron emission tomography; CSF: cerebrospinal fluid.




BRCA1 is a nuclear DNA repair protein, and its loss-of-function causes breast and ovarian cancers. In familial cases, insufficient DNA repair caused by loss-of-function mutations leads to genomic instability and contributes to cancer development (20). In sporadic breast cancers, hypermethylation of BRCA1 leads to its downregulation and results in insufficient DNA repair, causing cancer (1, 19). Thus, proper BRCA1 function is thought to be crucial in maintaining genomic DNA homeostasis.
In our previous study, aberrant hypomethylation of the BRCA1 promoter was observed in brains from Alzheimer’s disease (AD) patients (13). In contrast to cancer, BRCA1 was upregulated in association with promoter demethylation. A series of experiments showed that BRCA1 upregulation was a cellular protective response to amyloid β (Aβ)-induced DNA double strand breaks. Despite its upregulation, BRCA1 was sequestered to neurofibrillary tangles (NFTs) and mis-localized to the cytoplasm. This suppressed its function and led to the accumulation of significant DNA damage in AD neurons. Aberrant hypomethylation was observed in neurons as well as glial cells, suggesting that Aβ toxicity affected all types of cells in the central nervous system (CNS).
In AD brains, neurons are far more vulnerable than glial cells to Aβ toxicity. This difference could be explained by the absence of NFTs in glial cells. Thus, functional BRCA1 provided resistance against DNA damage. Hypomethylation was also observed in the cerebellum, suggesting that Aβ was affecting the entire brain. However, its effect on other peripheral tissues and cells is not yet known.
The blood-brain-barrier was once thought to separate the CNS from peripheral blood, with the exception of several small molecules. However, there are a number of studies showing that the CNS environment can be observed from peripherally-derived samples (8, 12). Among patients with bipolar disorders and schizophrenia, there is evidence that genes encoding molecules reported to be involved in these diseases show an altered epigenome in peripheral blood (2, 9, 18). A series of recent studies on AD strongly indicates that fibrillar Aβ accumulation in the CNS can be detected by measuring Aβ levels in peripheral blood samples (5, 15, 16, 21). Thus, we believed that it was worthwhile to analyze the methylation level of the BRCA1 promoter in peripheral blood from AD patients.
Here, we report the levels of BRCA1 promoter methylation using peripheral blood DNA derived from clinically diagnosed AD patients. We also analyzed the methylation levels in J-ADNI participants who underwent either Pittsburgh compound B amyloid positron emission tomography (PET) or cerebrospinal fluid (CSF) Aβ42 analysis that identified them as amyloid pathology-positive, and examined the relationship between the methylation level and clinical features.


All participants provided written informed consent. This study was approved by the ethics committee of the University of Tokyo (approvals G2183-18 and 11628-(1)). All experiments were performed in accordance with the principles of the Declaration of Helsinki.

Genomic DNA samples

Peripheral blood samples were obtained from the J-ADNI repository upon approval from the sample sharing committee. We obtained J-ADNI biomarker data from the National Bioscience Database Center with approval from its data access committee ( We also validated the results of the J-ADNI samples in an independent cohort of patients admitted at Niigata University Hospital (Niigata, Japan). Patients were diagnosed clinically as AD or NC.

Amyloid positivity

Amyloid positivity was defined clinically by visual inspection of PET images using Pittsburgh Compound B as a ligand by trained radiologists who were blinded to any clinical information (22) and/or a CSF Aβ42 value ≤ 333 pg/mL at the baseline, using the cut-off value determined in a previous report (11). When both data were available, we defined a sample as «positive» when at least one test met the «positive» criterion.


Bisulfite-conversion of DNA was performed by applying 100 ng of genomic DNA to an EpiTect Bisulfite Kit (Qiagen, Hilden, Germany), following the manufacturer’s instructions; the final product was eluted with 50 μL buffer. The primers used in this study were published previously (13) (Supplementary Figure 1). Samples were applied to PyroMark Q24 using PyroMark Gold Q24 Reagents (Qiagen), following amplification of bisulfite-converted DNA by the polymerase chain reaction using a PyroMark PCR Kit (Qiagen). The results were analyzed using PyroMark Q24 software (Qiagen). Primer set designs were verified using universal methylated and de-methylated human DNA standards.

Statistical analyses

Prism 6 (GraphPad, San Diego, CA, USA) and R software were used for statistical analyses. Unless otherwise noted, the significance of differences between groups was determined using the t-test. P < 0.05 was considered significant.



We analyzed the methylation level of the BRCA1 promoter in peripheral blood from normal control (NC) and clinically diagnosed AD patients. The demographics of the subjects are shown in Table 1. We performed pyrosequencing analysis of genomic DNA samples derived from 76 subjects. In total, we analyzed seven CpGs of the BRCA1 promoter region that were differentially methylated in AD brains (13). The linearity of all primer sets was validated using universal methylated and unmethylated human genomic DNA (Figure S1). There were no statistically significant differences in the methylation levels between NC and AD patients in all the probes analyzed (Figure 1). Because these subjects were diagnosed based only on clinical features without knowing brain Aβ accumulation, there remains a possibility that insufficient diagnostic accuracy could have affected the results. Therefore, amyloid positivity should be confirmed for an accurate diagnosis of AD.

Table 1. Demographics of normal control (NC) and clinically diagnosed Alzheimer’s disease (AD) patients

* Fisher’s exact test; †The APOE ε4 genotype was not available in five NC samples.

Figure 1. Pyrosequenced methylation levels of each BRCA1 probe

The methylation levels of CpG probes in BRCA1 were analyzed using the validation cohort. Black dots and gray boxes represent normal control (NC) and Alzheimer’s disease (AD) patients, respectively. Error bars represent means + SD. Statistical significance was determined using a two-tailed t-test. n = 40 (NC) and 36 (AD).


Because the previous study on the neuron-specific methylome in AD brains demonstrated that the presence of Aβ was essential for aberrant methylation of the BRCA1 promoter region (13), we examined methylation in peripheral blood from NC and AD subjects with known amyloid pathology from the J-ADNI cohort. The demographics of this cohort are shown in Table 2. Among the 537 subjects enrolled in that study, there were 51 NC and 49 AD subjects for whom amyloid data from amyloid PET or CSF Aβ42 analysis were available. No significant change in the methylation level of the BRCA1 promoter region was observed (Figure 2).

Table 2. Demographics of participants from the J-ADNI study

CSF: cerebrospinal fluid

Figure 2. Differentially methylated BRCA1 CpG probes in neuronal cells

The CpG probes in BRCA1 that were differentially methylated in neuronal cells (13])were analyzed by pyrosequencing. Black dots and the gray boxes represent normal control (NC) and Alzheimer’s disease (AD) patients, respectively. Error bars represent means + SD. Statistical significance was determined using a two-tailed t-test. n = 51 (NC) and 49 (AD).


Because the methylation levels of some CpGs correlate with age, and the age of the participants examined in this study differed between NC and AD subjects, we considered the possibility that age-related effects might have diminished the difference in methylation levels of BRCA1 between NC and AD patients. However, the methylation level of the CpGs in the BRCA1 promoter region were not affected by age (Figure 3). Thus, we concluded that the methylation level of the BRCA1 promoter region derived from peripheral blood did not reflect the accumulation of fibrillar Aβ in the brain.

Figure 3. Relationship between the pyrosequenced methylation level of each BRCA1 probe and age at death

Correlation plots of the pyrosequenced methylation levels of each BRCA1 probe and age at death. For normal control (NC) patients, blue dots and lines represent individual data and regression lines, respectively. For Alzheimer’s disease (AD) patients, red dots and lines represent individual data and regression lines, respectively. Gray bands represent the 95% confidence interval of each linear regression. Pearson’s product-moment correlation coefficient (r) values between the methylation level and age of NC and AD patients, and subsequent significance, are shown in each graph.


In the brain, methylation levels of the BRCA1 promoter region in neuronal cells were associated with the number of APOE ε4 alleles only in the AD cohort (13). Thus, we analyzed the relationship between sex differences or APOE ε4 and the methylation level of this region in the NC and AD cohorts. As shown in Figure 4A, female sex had a significant positive effect on the methylation level of the BRCA1 promoter region only in the AD group. The number of APOE ε4 alleles had opposite effects on the methylation levels in NC and AD groups; in NCs, the methylation level was negatively correlated with the APOE ε4 allele number, while in AD the level had a positive correlation.

Figure 4. Relationship between the pyrosequenced methylation level of each BRCA1 probe and age at death

(A, B) Comparison of the methylation level of each differential methylation position based on sex difference (A) and the number of APOE ε4 alleles (B). (A) Orange, green, blue, and purple represent normal control (NC) males, NC females, Alzheimer’s disease (AD) males, and AD females, respectively. Significance was determined using a two-way analysis of variance (ANOVA) followed by the post hoc Sidak method. (B) Orange, yellow, light green, green, blue, and purple represent NC without the APOE ε4 allele, NC with one APOE ε4 allele, NC with two APOE ε4 alleles, AD without APOE ε4 allele, AD with one APOE ε4 allele, and AD with two APOE ε4 alleles, respectively. Whiskers were defined by Tukey’s boxplot method. Significance was determined for NC and AD subjects using a two-way ANOVA followed by the post hoc Tukey’s (for NC) or Sidak (for AD) method. Boxes extend from the 25th–75th percentile. Lines in the boxes are medians.



In this study, we clearly demonstrated that the methylation level of the BRCA1 promoter in peripheral blood was unchanged between NC and AD patients. This was in stark contrast to the global methylation change in the CNS. In the brain, toxic Aβ induces DNA damage regardless of the region or the cell type, resulting in BRCA1 up-regulation through promoter hypomethylation. This allows sufficient DNA repair only in the absence of cytoplasmic aggregated tau because BRCA1 co-aggregates with NFTs and loses its proper function exclusively in neurons. In peripheral blood cells that are apparently free from NFTs, even if peripheral toxic Aβ had any effect on DNA damage, proper repair should occur. The absence of different methylation levels in NC and AD subjects could be attributed to a low Aβ concentration in the peripheral blood leading to a less prominent response of peripheral cells towards Aβ toxicity. Another explanation is that, in peripheral blood, Aβ toxicity could differ from that of the CNS, even if blood biomarker studies show that CNS Aβ accumulation can be detected by an analysis of the peripheral blood. This suggests that the CNS and peripheral blood share common properties of Aβ species.
A sub-analysis revealed several interesting results. Upon stratifying by sex, AD females showed higher methylation levels than males. One of the CpGs showed significantly higher methylation levels even after multiple post-hoc comparisons (Figure 4A, probe cg18372208). As the methylation level of the promoter region generally shows an inverse relationship with the expression of downstream genes (1, 19), the peripheral response to Aβ could be insufficient in females, resulting in vulnerability to toxic Aβ. This response could be related to the fact that being female is a risk factor for AD (4, 17) and also for a rapid decline in cognitive function (3, 6, 7, 10, 14).
In our previous study, we did not observe any sex differences in the BRCA1 promoter methylation level in the CNS (13). This discrepancy could be explained by differences of the organs we analyzed or the disease stage. Specifically, J-ADNI AD patients exhibited mild to moderate dementia, while autopsy patients usually suffer from severe dementia.
When comparing the number of APOE ε4 alleles and BRCA1 methylation levels, the presence of APOE ε4 alleles was negatively associated in NC and positively associated in AD subjects. These opposing results could be explained by a potential protective effect of BRCA1 upregulation in response to peripheral Aβ exposure; APOE ε4 carriers are protected from Aβ toxicity when BRCA1 is upregulated (i.e., a lower methylation level), whereas they are not protected when BRCA1 is downregulated (i.e., higher methylation in AD patients).
The cause of these different responses is not fully clear. Recent studies have shown that Aβ accumulation in the CNS could be detected by analyzing the peripheral blood (5, 15, 16). One study even showed that patients with CNS Aβ accumulation had increased Aβ oligomerization activity in peripheral blood (21). These results imply that aberrant Aβ metabolism could be occurring both in central and peripheral tissues, and could drive BRCA1 upregulation in certain conditions.
In summary, no significant changes in methylation of the BRCA1 promoter were observed in NC and AD patients. Thus, we concluded that BRCA1 promoter methylation cannot be a biomarker for diagnosing AD. Nevertheless, we found distinctive hypomethylation of the BRCA1 promoter in the brain through a neuron-specific methylome analysis. These data emphasized the importance of analyzing specific cell types directly involved in the disease process. However, only in the AD group were risk factors for AD (i.e., female sex and APOE allele ε4 number) correlated with BRCA1 promoter methylation in the peripheral blood, implying that the response to toxic Aβ in terms of BRCA1 expression was shared in the peripheral blood. This could be attributed to recent findings that the brain environment can be partially detected by observing peripheral blood, and provides insights into how the CNS and peripheral blood cross-talk in terms of Aβ accumulation (5, 15, 16, 21). Increased hypomethylation of the BRCA1 promoter has a potentially protective effect against Aβ toxicity. Therefore, while its relevance remains unclear, higher levels of BRCA1 promoter methylation might indicate AD risk, suggesting that peripheral methylation could be a potential biomarker for AD progression.


Author Contributions: Conceptualization, TM and AI; Methodology, TM and KS; Resource, TI and TI; Supervision, TT; Writing – Original Draft Preparation, TM; Writing – Review & Editing, AI; Funding, TM and AI.

Funding: This study was supported by AMED under grant number 17dm0107069h0002, 18dk0207028h0003, 18dk0207020h0004, JP21dk0207057h0001, 21dk0207042h0003, 21dk0207046h0001. JSPS KAKENHI grant numbers 16H05316 and 19K17027, the Cell Science Research Foundation (Osaka, Japan), the Ichiro Kanehara Foundation for the Promotion of Medical Sciences and Medical Care (Tokyo, Japan), the Takeda Science Foundation (Osaka, Japan), The Mochida Memorial Foundation for Medical and Pharmaceutical Research (Tokyo, Japan), Janssen Pharmaceutical K.K. (Tokyo, Japan), and Eisai Co. (Tokyo, Japan).

Acknowledgements: We are grateful for the technical support provided by Yuki Inukai-Mizutani.

Conflicts of interest: None.






1. Baldwin RL, Nemeth E, Tran H, et al., BRCA1 promoter region hypermethylation in ovarian carcinoma: a population-based study, Cancer research 2000;60:5329-5333.
2. Booij L, Szyf M, Carballedo A, et al., DNA methylation of the serotonin transporter gene in peripheral cells and stress-related changes in hippocampal volume: a study in depressed patients and healthy controls, PLoS One 2015;10:e0119061.
3. Butchart J, Birch B, Bassily R, Wolfe L, Holmes C, Male sex hormones and systemic inflammation in Alzheimer disease, Alzheimer Dis Assoc Disord 2013;27:153-156.
4. Caracciolo B, Palmer K, Monastero R, et al., Occurrence of cognitive impairment and dementia in the community: a 9-year-long prospective study, Neurology 2008;70:1778-1785.
5. Fandos N, Perez-Grijalba V, Pesini P, et al., Plasma amyloid beta 42/40 ratios as biomarkers for amyloid beta cerebral deposition in cognitively normal individuals, Alzheimers Dement (Amst) 2017;8:179-187.
6. Ferretti MT, Iulita MF, Cavedo E, et al., Sex differences in Alzheimer disease – the gateway to precision medicine, Nat Rev Neurol 2018;14:457-469.
7. Fyfe I, Alzheimer disease: Sex-specific inflammatory link to early Alzheimer pathology, Nat Rev Neurol 2017;13:5.
8. Gaiottino J, Norgren N, Dobson R, et al., Increased neurofilament light chain blood levels in neurodegenerative neurological diseases, PLoS One 2013;8:e75091.
9. Ikegame T, Bundo M, Murata Y, et al., DNA methylation of the BDNF gene and its relevance to psychiatric disorders, J Hum Genet 2013;58:434-438.
10. Iwata A, Iwatsubo T, Ihara R, et al., Effects of sex, educational background, and chronic kidney disease grading on longitudinal cognitive and functional decline in patients in the Japanese Alzheimer’s Disease Neuroimaging Initiative study, Alzheimers Dement (N Y) 2018;4:765-774.
11. Iwatsubo T, Iwata A, Suzuki K, et al., Japanese and North American Alzheimer’s Disease Neuroimaging Initiative studies: Harmonization for international trials, Alzheimer’s & Dementia 2018.
12. Lu CH, Macdonald-Wallis C, Gray E, et al., Neurofilament light chain: A prognostic biomarker in amyotrophic lateral sclerosis, Neurology 2015;84:2247-2257.
13. Mano T, Nagata K, Nonaka T, et al., Neuron-specific methylome analysis reveals epigenetic regulation and tau-related dysfunction of BRCA1 in Alzheimer’s disease, Proceedings of the National Academy of Sciences 2017.
14. Mosconi L, Berti V, Quinn C, et al., Sex differences in Alzheimer risk: Brain imaging of endocrine vs chronologic aging, Neurology 2017;89:1382-1390.
15. Nakamura A, Kaneko N, Villemagne VL, et al., High performance plasma amyloid-β biomarkers for Alzheimer’s disease, Nature 2018.
16. Ovod V, Ramsey KN, Mawuenyega KG, et al., Amyloid beta concentrations and stable isotope labeling kinetics of human plasma specific to central nervous system amyloidosis, Alzheimers Dement 2017;13:841-849.
17. Roberts RO, Geda YE, Knopman DS, et al., The incidence of MCI differs by subtype and is higher in men The Mayo Clinic Study of Aging, Neurology 2012;78:342-351.
18. Turecki G, Ota VK, Belangero SI, Jackowski A, Kaufman J, Early life adversity, genomic plasticity, and psychopathology, The Lancet Psychiatry 2014;1:461-466.
19. Turner NC, Reis-Filho JS, Russell AM, et al., BRCA1 dysfunction in sporadic basal-like breast cancer, Oncogene 2007;26:2126-2132.
20. Venkitaraman AR, Cancer susceptibility and the functions of BRCA1 and BRCA2, Cell 2002;108:171-182.
21. Wang MJ, Yi S, Han JY, et al., Oligomeric forms of amyloid-beta protein in plasma as a potential blood-based biomarker for Alzheimer’s disease, Alzheimers Res Ther 2017;9:98.
22. Yamane T, Ishii K, Sakata M, et al., Inter-rater variability of visual interpretation and comparison with quantitative evaluation of (11)C-PiB PET amyloid images of the Japanese Alzheimer’s Disease Neuroimaging Initiative (J-ADNI) multicenter study, Eur J Nucl Med Mol Imaging 2017;44:850-857.



S. Saunders1, G. Muniz-Terrera1, S. Sheehan2, C.W. Ritchie1,3, S. Luz2


1. Centre for Clinical Brain Sciences, University of Edinburgh, UK; 2. Usher Institute of Population Health Sciences and Informatics; Molecular, Genetic and Population Health Sciences, University of Edinburgh, UK; 3. Brain Health Scotland, UK

Corresponding Author: Stina Saunders, University of Edinburgh, Centre for Clinical Brain Sciences, UK,

J Prev Alz Dis 2021;
Published online June 9, 2021,



BACKGROUND: It is important to use outcome measures for novel interventions in Alzheimer’s disease (AD) that capture the research participants’ views of effectiveness. The electronic Person-Specific Outcome Measure (ePSOM) development programme is underpinned by the need to identify and detect change in early disease manifestations and the possibilities of incorporating artificial intelligence in outcome measures.
Objectives: The aim of the ePSOM programme is to better understand what outcomes matter to patients in the AD population with a focus on those at the pre-dementia stages of disease. Ultimately, we aim to develop an app with robust psychometric properties to be used as a patient reported outcome measure in AD clinical trials.
Design: We designed and ran a nationwide study (Aug 2019 – Nov 2019, UK), collecting primarily free text responses in five pre-defined domains. We collected self-reported clinical details and sociodemographic data to analyse responses by key variables whilst keeping the survey short (around 15 minutes). We used clustering and Natural Language Processing techniques to identify themes which matter most to individuals when developing new treatments for AD.
Results: The study was completed by 5,808 respondents, yielding over 80,000 free text answers. The analysis resulted in 184 themes of importance. An analysis focusing on key demographics to explore how priorities differed by age, gender and education revealed that there are significant differences in what groups consider important about their brain health.
Discussion: The ePSOM data has generated evidence on what matters to people when developing new treatments for AD that target secondary prevention and therein maintenance of brain health. These results will form the basis for an electronic outcome measure to be used in AD clinical research and clinical practice.

Key words: Clinically meaningful change, electronic patient reported outcome measures, Alzheimer’s disease, outcome measures, brain health.




Attempts to develop disease modifying therapies for Alzheimer’s disease (AD) started over 20 years ago with little success to date. A recent estimate of the costs of AD was US$818B, which is equivalent to the combined GDP of Indonesia, The Netherlands, and Turkey (1).
The lack of progress in finding a pharmacological treatment for AD is however at odds with a rapid development in the understanding of the pathology of AD suggesting that clinical trial design and delivery may partially account for a lack of progress with insensitive outcome measures lacking clinical meaningfulness also playing a part in this lack of progress. It has been shown that the disease process starts long before an individual becomes symptomatic or eventually, the dementia syndrome manifests (2, 3). Increasingly, we are exploring AD processes at earlier disease stages through examining at-risk populations in mid-life which helps identify the earliest manifestation of declining brain health. In the absence of pharmacological interventions, it is estimated that approximately 40% of dementia cases could be prevented by targeting epidemiologically derived modifiable risk factors (4). Changes occurring years earlier than dementia develops have been observed in at-risk populations using exploratory and sensitive computerised tests assessing e.g. allocentric and egocentric spatial processing (5). These test results correlate with brain imaging findings in hippocampal subfields known to be sensitive to amyloid derived neurotoxicity (6); as well as in changes to brain β-amyloid in at risk populations aged between 63-81 years old who did not have dementia (7).
Whilst there are global initiatives focusing on dementia prevention through risk factor modification (8, 9), there remains a major and complementary need for effective AD pharmacological interventions. Irrespective of the type of intervention to reduce incident dementia rates, the fact is that these studies will engage at risk populations who will be, to the most part, in mid-life and healthy. Currently, there are 31 AD drugs being tested in Phase III clinical trials (19 of which are disease modifying) (10). We argue that using outcome measures assessing clinical symptoms and functioning in earlier disease stages is less valid than biological measures of disease and what the individual considers personally meaningful from a treatment. A treatment’s success should therefore be determined not only by the impact on the individual’s disease (as evidenced by biomarker change) but also by its effect on related well-being (as measured by patient reported outcomes).
To this end, whilst it is currently proposed by regulators that AD trials measure cognition as the primary outcome, as trials move to an earlier disease stage it could be argued that many commonly used (cognitive) measures lack ecological validity and are not sensitive enough to detect changes in the earlier stages of the AD continuum where the ideal intervention should take place (11). Moreover, it is recommended by both the US Food and Drug Administration (FDA) (12) and European Medicines Agency (EMA) (13) that AD trials incorporate measures which capture clinically meaningful results to the individual. Patient reported outcome measures (PROMs) are developed for the incorporation of the person’s own perspective regarding their treatment, though these measures are currently not used in AD clinical trials (14). PROMs reflect an individual’s view on what they define as an effective treatment and consider a meaningful change. Notably, PROMs are already more widely used in other disease areas. For example, a recent study of nearly 100,000 clinical trials published on found that a PROM had been used in 27% of all trials, primarily in oncology (15).
In light of the drive towards early detection, looking at younger at-risk populations and the main regulators’ recommendation for clinically meaningful outcome measures, we have established the electronic Person Specific Outcome Measure (ePSOM) development programme. As the target population in dementia prevention research is an at-risk population, our group took the view that what matters to people when developing new treatments for AD is approached by way of maintenance of brain health (16, 17). The ePSOM programme consists of four sequential steps, ultimately aiming to employ new technology to create an outcome measure to be used in AD clinical research and practice. This will be in the form of an outcome app used on any screen-based device which will assess aspects specific to the individual using it. At the start of the programme, we reviewed literature around PROMs in AD clinical trials which informed our focus group study with people with memory concerns, healthy volunteers and health care professionals (18). The focus group study yielded five domains of importance for what matters to people about brain health. These domains formed the basis for the next stage of the ePSOM development programme. In this paper, we report on a large UK-wide population-based study to understand what matters to people when developing new treatments for Alzheimer’s disease. We consider the respondents to the ePSOM study a representative population of individuals who may be enrolled in dementia primary and secondary prevention clinical trials and characterise what matters to people about brain health focusing on key demographic groups.



We designed and ran a UK-wide population-based online study collecting primarily free text answers (see Appendix 1). The study built on a previously run focus group study which yielded five domains of importance about brain health. The study obtained ethics approval from the ACCORD Medical Research Ethics Committee in Edinburgh, Scotland. The ePSOM study ran from Aug 2019 – Dec 2019 and was divided into sections, starting with an introductory video and informed consent.
Free text answers were collected across five pre-defined domains. These answers were clustered, leading to specific themes of what matters to people about brain health

The study was open to anyone over the age of 18 and was launched primarily via Alzheimer’s Research UK media channels through e-mails to individuals registered on their mailing lists and a social media campaign (with social media support from other dementia related organisations). We collected key sociodemographic and clinical data such as having been seen by a doctor because of any brain health issues though the primary method of the survey used a qualitative approach. Respondents were presented with the five domains derived from the earlier focus group work to orientate and channel free text responses: [1] Everyday functioning; [2] Sense of Identity; [3] Relationships and Social Connections; [4] Enjoyable Activities and [5] Thinking problems. They were then asked to provide free text answers on what they would like to retain or keep being able to do in those domains if their brain health got worse. At the end of the study, respondents were asked to identify five answers across all the answers they had given which they consider the most important. We used Natural Language Processing (NLP) techniques to analyse the free text data (see Figure 1).

Figure 1. Natural Language Processing techniques used to analyse the survey data


Free text answers were collected across five pre-defined domains. These answers were clustered, leading to specific themes of what matters to people about brain health

Step 1: Natural Language Processing to create clusters

We used NLP to create clusters of semantically similar free text answers. These clusters were then manually annotated with appropriate labels. We refer to the finally labelled clusters derived from this stepped NLP-manual annotation process as “themes”.
NLP employed word embeddings trained on vast amounts of text data to achieve fine-grained representation of semantic regularities in text. We were thus able to build robust representations of free text answers. To begin, “stop words” (i.e. words that occur very frequently and contribute little to semantic content) and punctuation were removed from the free text answers. The resulting texts were then converted to numerical vector representations by using GloVe vectors (19) to generate sentence embeddings. These vectors encode semantic relationships between words and can be used to cluster semantically similar text segments. This allowed us to use automated methods to identify words, and thus answers, of a similar “theme” or meaning. The K-means clustering algorithm was used to cluster the answer embeddings within each of the five domains. The K parameter, that is, the desired number of automatic clusters per domain, was determined analytically. The goal was to generate fine grained clusters which contain semantically similar answers while avoiding overfitting or creating so many clusters that important themes are not revealed. We found when the number of important items in the largest cluster changes by less than 10, between each of the previous five increments of K, that the majority of the clusters also exhibited minor changes in the number of important items. Using this criterion, we chose a value of 151 clusters across all five domains. This method resulted in a total of 755 clusters of free text answers, or 151 clusters for each of the five domains.

Step 2: Manual Annotation to create themes

The clusters that emerged within each of the five domains were reordered so that semantically similar clusters appeared close together. This was achieved using hierarchical clustering on the cluster centroids. We used the reordered clusters for manual annotation in each of the five domains. Each cluster was represented by the 200 most frequent unique answers, after punctuation and stopwords were removed. The annotation goals were to combine any clusters which fit together, exclude uninterpretable clusters and label the final clusters thus deriving the final themes. Six authors of the current paper annotated two domains each, ensuring two separate people analysed a single domain, which helped ensure inter-rater reliability between domains. Finally, two of the authors did quality control across the five domains and homogenised the labels across domains.

Statistical analyses

In this paper, we focus our analyses on key demographic groups: age (up to the age of 64 / age 65 and older); gender (men / women) and education (no degree / degree and higher). We present the largest themes for each of these demographic groups as well as themes which were identified as particularly important most often in the final question on the study forms. For both of these analyses, we report percentages for each theme by key demographic groups. As the demographic groups are unbalanced in terms of the number of respondents we use percentages rather than the absolute number of answers in the statistical analyses. The percentages are derived by dividing the count of answers within the demographic group by the total number of answers in that demographic group, thus providing proportions for comparison when dealing with imbalanced demographics. It should be noted that respondents were not bounded by a minimum or maximum number of free text answers they could give in each domain.
Finally, we conducted a Chi-squared test to analyse whether the differences in percentages between demographic groupings’ answers within each theme were statistically significant. A p-value of <0.01 was used in statistical significance testing.




The study was completed by 5,808 people from across the UK. They provided a total of 82,514 free text answers. These were clustered using automated NLP techniques resulting in 151 clusters in each of the five pre-defined domains, a total of 755 automated clusters across all domains, as described. Subsequent analysis reduced the number of clusters to 334 (due to a cluster being represented in two or more domains) which were all manually annotated by the research team. Many of the same themes emerged from different domains (e.g., the theme of Walking in the “Enjoyable activities” domain as well as the “Everyday activities” domain). After merging themes with the same label in different domains, the final number of unique themes was 184. Some respondents used more generic language (e.g., “Maintaining independence”) whereas others were more specific (e.g., “Driving”). Using NLP methods for free text analysis means that, in this example, the “Maintaining independence” theme contains 1100 answers, most containing either the word “independent” or “independence”. Analytically, this is therefore not a general theme for answers which relate to the concept of independence, but a cluster of answers in which the respondents are directly referencing the word independence as something which is important for them to maintain. This has therefore resulted in themes which are more or less specific but directly reflect the language used by the respondents.
Pre-defined answers: Characteristics of the ePSOM survey sample

The characteristics of the 5,808 respondents are presented above (see Table 1).

Table 1. ePSOM survey respondent characteristics

We used NLP techniques and manual annotation to group individual free text answers into clusters and then themes respectively. The most frequent themes across all demographics were reading, driving, friendships and following a storyline (Figure 2). We also calculated the proportion of answers within each key demographic, expressed as percentages of the total answers given by that demographic group.

Figure 2. What matters to people about brain health? The survey received 82,514 free text answers which were clustered into 184 themes

This figure shows themes which were mentioned the most, broken down by key demographics. Full figure of the survey themes in Appendix 2.

At the end of the survey, respondents were asked to identify the five most important answers to them across all their answers. We used this metric to rank the themes in terms of being selected as particularly important and observed a subtle difference between the largest themes (themes mentioned the most frequently) and themes which are identified as the most important. The 5 top important themes across all demographics were family connections, driving, socializing, reading and friendships (Figure 3).

Figure 3. What matters to people about brain health?

This figure shows themes with the highest number of answers selected as particularly important by key demographics. Full figure of themes with the most important answers in Appendix 3.


Cross-Tabulations of key demographics

The following tables show statistically significant proportional differences in theme sizes (Table 2) and identifying themes as particularly important (Table 3), focusing on demographic group dyads (younger vs older; men vs women; individuals with no degree vs individuals with a degree or higher).
Table 2 Top 10 themes selected as particularly important which had the highest Chi square values representing greater differentiation between demographic groups by age (younger/older); gender (male/female) and education (no degree / degree and higher). A full list of particularly important themes which were significantly different across key demographics can be found in Appendix 4.

Table 2. Top 10 themes selected as particularly important which had the highest Chi square values representing greater differentiation between demographic groups by age (younger/older); gender (male/female) and education (no degree / degree and higher)

Full list of particularly important themes which were significantly different across key demographics in Appendix 4.

Table 3. Top 10 largest themes which had the highest Chi square values representing greater differentiation between demographic groups by age (younger/older); gender (male/female) and education (no degree / degree and higher)

Full list of largest themes which were significantly different across key demographics in Appendix 5.



Building on the scientific foundation provided by previous stages of the ePSOM research programme, we designed and ran a nationwide study with open ended questions to derive free text answers exploring what matters to people about maintaining their brain health within five focus group-derived domains. To our knowledge, this is the first study collecting free and systematically analysing text responses from a very large number of respondents on what is important to them about brain health. The themes and granularity derived from our study are in line with the FDA’s guidance for capturing aspects relevant to AD research participants “e.g., [assessing] facility with financial transactions, adequacy of social conversation” (12).
As AD drug development moves to an earlier phase of the neurodegenerative disease spectrum and clinical research targets an earlier, younger population, it is crucial any outcomes are meaningful and relevant to that trial population. Additionally, as upcoming AD treatments are hoped to be disease modifying rather than reducing symptoms, the cognitive domains which respond to the medication may not be the same as with symptomatic treatments measured at a later disease stage (20). We also know from a recent review that lifestyle factors may influence brain health in midlife (21) so it is apposite to examine what matters to people about brain health including lifestyle dependant factors as this will be increasingly relevant in Brain Health Clinics which are developing throughout the UK (16) and Europe (17).
There has been other work collecting evidence on important outcomes focusing on the point of view of people living with dementia (22). The focus of the ePSOM programme though is the maintenance of brain health. As the majority of the individuals in our study had not received a diagnosis of neurodegenerative disease, the findings from our study provide evidence for what matters to people about brain health in normal lived experience which may include people at the earlier (asymptomatic) stages of disease rather than once the dementia syndrome develops. Our findings are supported by literature recognising that AD trials currently do not measure outcomes which are relevant to the patient themselves. Tochel et al. (23) carried out a literature review extracting data from studies where participants described outcomes which matter to them. Their review concluded by demonstrating an array of outcomes which are not commonly captured in clinical trials of new treatments (23).
Changes at the early stages of the AD continuum are currently detected by biomarker assessments, with functional measures used increasingly towards the more symptomatic and advanced stage of the continuum where ultimately impairment is evidenced in basic activities of daily living. However, dementia prevention cohorts have found differences in more than just biomarker assessed pathology, e.g. there is evidence that middle-aged adults at risk of dementia have poorer cognitive performance, principally in visuospatial functions (24) and memory (25). Lau et al. (26) concluded that observing early functional limitations at baseline in the at-risk population had prognostic value in identifying older adults at risk for developing functional disability a few years later (26).
A recent review also concluded that in the pre-dementia stages of AD, executive functions (such as inhibitory abilities), attentional and visuospatial functions can already be impacted (27). A PROM therefore could be viewed as an ecologically valid instrument for cognitive assessment measures which are proxies for what matters to people, especially if the PROM relates to a cognitive process affected early in the course of AD (e.g. activities requiring planning, judgement or navigation/orientation like confidence driving). The key questions here is: if an individual’s score changes on a particular domain using a cognitive assessment measure, does this correlate with a change of score in a PROM and is therefore a change meaningful (by definition) to the patient? While functional or Activities of Daily Living scales measure a more direct or practical effect a drug may have, these measures have limitations such as poor psychometric properties (28) and as evidenced by the analysis of key demographic groups in the ePSOM survey, what matters to people about brain health and their function is different depending on age, sex and education levels. By capturing data specific to the individual who in effect derives their own outcome measure, the ePSOM app in development would present an outcome measure for clinical trials that captures changes noticed by and meaningful to the person themselves and therefore more likely to be correlated to their own specific functional outcomes than generic outcomes which were derived by homogenising population level data. Ultimately, employing more meaningful, ecologically valid and sensitive measures will facilitate more drugs to be approved by regulatory bodies which will actually impact on well-being and not just impact on cognition and function ‘on average’ between groups (29). Moreover – ePSOMs are immune to cultural, educational and language variability as each outcome is unique to that individual and bears no reference to an external ‘population norm’.
We used an online study design as it was important to allow for free text answers and reach a large number of people. However, this is also a limitation in the study leading to inevitable sampling bias of individuals who are able to access an online survey. There was also a demographic imbalance among the survey respondents with reference to the UK population as a whole, but appropriate analysis focusing on proportions rather than absolute values of this relatively large sample mitigates the effects of the imbalances in the data. The main strength of the study was collecting free text answers and using NLP techniques in the data analysis. Employing NLP techniques to gather evidence for what outcomes matter in AD drug development is unique and we are not aware of any similar studies. Free text answers offer insights which go beyond rating themes on a scale which have been predefined as important by the researchers and are culturally biased and limited. Moreover, the open character of the questions may motivate respondents to reveal more (31). In some regards, our study results may be considered comparable to hundreds of focus group studies, though by using NLP techniques, we are able to extract patterns in answers by key demographic at a scale and level of detail not feasible using traditional qualitative methodologies.



There is a growing consensus that PROMs should be used in AD trials so that the patient can assess if they observe a change in their well-being which is meaningful and specific to them. Including the patient’s perspective is also recommended by regulatory bodies such as the EMA with whom we collaborated in the initial phases of this project, and the FDA. In our study, we included a large number of people collecting free text responses to understand what matters to people about their brain health – our analyses focussed on key demographic groups. This approach is novel in so much as it uses NLP approaches to create a range of outcomes from a theoretically limitless range of possible responses and then can apply these into quantifiable and ecologically valid outcomes. The main criticism and in many ways fatal flaw of current approaches to PROMs is that they are derived at a population level and therefore have to incorporate the characteristics of the population they were derived from. These populations will hold certain language, cultural and ethnic characteristics making their use in other limited in other populations. The ePSOM app will ultimately be used by people in earlier stages of neurodegenerative disease before dementia develops in populations across the globe, in clinical trials with seamless translation into clinical practice.


Acknowledgement: We thank Alison Evans from Alzheimer’s Research UK for her intellectual contribution to this study. We also thank the individuals who took part in the ePSOM study.

Conflicts: The authors declare no conflict of interest.

Funding Sources: The ePSOM survey was funded by Alzheimer’s Research UK.

Declarations of interest: none (all authors).

Ethical standards: The study obtained ethics approval from the ACCORD Medical Research Ethics Committee in Edinburgh, Scotland.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.









1. Wimo A, Guerchet M, Ali G-C, Wu Y-T, Prina AM, Winblad B, et al. The worldwide costs of dementia 2015 and comparisons with 2010. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2017;13(1):1-7. doi: 10.1016/j.jalz.2016.07.150
2. DeTure MA, Dickson DW. The neuropathological diagnosis of Alzheimer’s disease. Molecular Neurodegeneration. 2019;14(1):32. doi: 10.1186/s13024-019-0333-5
3. Association As. Alzheimer’s disease facts and figures. Alzheimer’s & Dementia. 2020;16(3):391-460. doi: 10.1002/alz.12068
4. Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-46. doi: 10.1016/s0140-6736(20)30367-6
5. Ritchie K, Carrière I, Howett D, Su L, Hornberger M, O’Brien JT, et al. Allocentric and Egocentric Spatial Processing in Middle-Aged Adults at High Risk of Late-Onset Alzheimer’s Disease: The PREVENT Dementia Study. Journal of Alzheimer’s Disease. 2018;65:885-96. doi: 10.3233/JAD-180432
6. Brien JT, Firbank MJ, Ritchie K, Wells K, Williams GB, Ritchie CW, et al. Association between midlife dementia risk factors and longitudinal brain atrophy: the PREVENT-Dementia study. Journal of Neurology, Neurosurgery &amp;amp; Psychiatry. 2020;91(2):158. doi: 10.1136/jnnp-2019-321652
7. Martikainen IK, Kemppainen N, Johansson J, Teuho J, Helin S, Liu Y, et al. Brain β-Amyloid and Atrophy in Individuals at Increased Risk of Cognitive Decline. AJNR Am J Neuroradiol. 2019;40(1):80-5. doi: 10.3174/ajnr.A5891
8. Ritchie CW, Muniz-Terrera G, Kivipelto M, Solomon A, Tom B, Molinuevo JL, et al. The European Prevention of Alzheimer’s Dementia (EPAD) Longitudinal Cohort Study: Baseline Data Release V500.0. J Prev Alzheimers Dis. 2020;7(1):8-13. doi: 10.14283/jpad.2019.46
9. Cummings J, Aisen P, Barton R, Bork J, Doody R, Dwyer J, et al. Re-Engineering Alzheimer Clinical Trials: Global Alzheimer’s Platform Network. J Prev Alzheimers Dis. 2016;3(2):114-20. doi: 10.14283/jpad.2016.93
10. Devenish SRA. The current landscape in Alzheimer’s disease research and drug discovery. Drug Discov Today. 2020;25(6):943-5. doi: 10.1016/j.drudis.2020.04.002
11. Snyder PJ, Kahle-Wrobleski K, Brannan S, Miller DS, Schindler RJ, DeSanti S, et al. Assessing cognition and function in Alzheimer’s disease clinical trials: do we have the right tools? Alzheimers Dement. 2014;10(6):853-60. doi:10.1016/j.jalz.2014.07.158
12. FDA UFaDA. Early Alzheimer’s Disease: Developing Drugs for Treatment Guidance for Industry. 2018.
13. EMA EMA. Guideline on the clinical investigation of medicines for the treatment of Alzheimer’s disease. Committee for Medicinal Products for Human Use (CHMP). 2018:1-36.
14. Saunders S, Muniz-Terrera G, Watson J, Clarke CL, Luz S, Evans AR, et al. Participant outcomes and preferences in Alzheimer’s disease clinical trials: The electronic Person-Specific Outcome Measure (ePSOM) development program. Alzheimers Dement (N Y). 2018;4:694-702. doi:10.1016/j.trci.2018.10.013
15. Vodicka E, Kim K, Devine EB, Gnanasakthy A, Scoggins JF, Patrick DL. Inclusion of patient-reported outcome measures in registered clinical trials: Evidence from (2007-2013). Contemp Clin Trials. 2015;43:1-9. doi:10.1016/j.cct.2015.04.004
16. Ritchie CW, Russ, T. C., Banerjee, S., Barber, B., Boaden, A., Fox, N. C., . . . Burns, A. The Edinburgh Consensus: preparing for the advent of disease-modifying therapies for Alzheimer’s disease. Alzheimers Res Ther. 2017;9(1):85. doi:doi:10.1186/s13195-017-0312-4
17. Frisoni GB, Molinuevo JL, Altomare D, Carrera E, Barkhof F, Berkhof J, et al. Precision prevention of Alzheimer’s and other dementias: Anticipating future needs in the control of risk factors and implementation of disease-modifying therapies. Alzheimers Dement. 2020;16(10):1457-68. doi:10.1002/alz.12132
18. Watson J, Saunders S, Muniz Terrera G, Ritchie C, Evans A, Luz S, et al. What matters to people with memory problems, healthy volunteers and health and social care professionals in the context of developing treatment to prevent Alzheimer’s dementia? A qualitative study. Health Expect. 2019;22(3):504-17. doi:10.1111/hex.12876
19. Pennington J, Socher, R., Manning, C. Glove: Global vectors for word representation Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP). 2014:1532–43. doi: 10.3115/v1/D14-1162
20. Hendrix SB. Measuring clinical progression in MCI and pre-MCI populations: enrichment and optimizing clinical outcomes over time. Alzheimers Res Ther. 2012;4(4):24-. doi:10.1186/alzrt127
21. Topiwala H, Terrera GM, Stirland L, Saunderson K, Russ TC, Dozier MF, et al. Lifestyle and neurodegeneration in midlife as expressed on functional magnetic resonance imaging: A systematic review. Alzheimer’s & Dementia: Translational Research & Clinical Interventions. 2018;4(1):182-94. doi:10.1016/j.trci.2018.04.001
22. Morbey H, Harding AJE, Swarbrick C, Ahmed F, Elvish R, Keady J, et al. Involving people living with dementia in research: an accessible modified Delphi survey for core outcome set development. Trials. 2019;20(1):12. doi:10.1186/s13063-018-3069-6
23. Tochel C, Smith M, Baldwin H, Gustavsson A, Ly A, Bexelius C, et al. What outcomes are important to patients with mild cognitive impairment or Alzheimer’s disease, their caregivers, and health-care professionals? A systematic review. Alzheimers Dement (Amst). 2019;11:231-47. doi:10.1016/j.dadm.2018.12.003
24. Ritchie K, Carrière I, Su L, O’Brien JT, Lovestone S, Wells K, et al. The midlife cognitive profiles of adults at high risk of late-onset Alzheimer’s disease: The PREVENT study. Alzheimer’s & Dementia. 2017;13(10):1089-97. doi: 10.1016/j.jalz.2017.02.008
25. Caselli RJ, Locke DEC, Dueck AC, Knopman DS, Woodruff BK, Hoffman-Snyder C, et al. The neuropsychology of normal aging and preclinical Alzheimer’s disease. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2014;10(1):84-92. doi:10.1016/j.jalz.2013.01.004
26. Lau KM, Parikh M, Harvey DJ, Huang CJ, Farias ST. Early Cognitively Based Functional Limitations Predict Loss of Independence in Instrumental Activities of Daily Living in Older Adults. J Int Neuropsychol Soc. 2015;21(9):688-98. doi:10.1017/s1355617715000818
27. Guarino A, Favieri F, Boncompagni I, Agostini F, Cantone M, Casagrande M. Executive Functions in Alzheimer Disease: A Systematic Review. Front Aging Neurosci. 2019;10:437-. doi:10.3389/fnagi.2018.00437
28. Weintraub S, Carrillo MC, Farias ST, Goldberg TE, Hendrix JA, Jaeger J, et al. Measuring cognition and function in the preclinical stage of Alzheimer’s disease. Alzheimer’s & dementia (New York, N Y). 2018;4:64-75. doi:10.1016/j.trci.2018.01.003
29. Sabbagh MN, Hendrix S, Harrison JE. FDA position statement “Early Alzheimer’s disease: Developing drugs for treatment, Guidance for Industry”. Alzheimer’s & dementia (New York, N Y). 2019;5:13-9. doi:10.1016/j.trci.2018.11.004
30. Ashour HM, Elkhatib WF, Rahman MM, Elshabrawy HA. Insights into the Recent 2019 Novel Coronavirus (SARS-CoV-2) in Light of Past Human Coronavirus Outbreaks. Pathogens. 2020;9(3). doi:10.3390/pathogens9030186
31. Rohrer JM, Brümmer M, Schmukle SC, Goebel J, Wagner GG. “What else are you worried about?” – Integrating textual responses into quantitative social science research. PLOS ONE. 2017;12(7):e0182156. doi:10.1371/journal.pone.0182156s



J.T. McDaniel, R.J. McDermott, T. Schneider


Southern Illinois University Carbondale, USA

Corresponding Author: Justin T McDaniel, Southern Illinois University Carbondale, USA,



BACKGROUND: Although studies have examined the geographic distribution of dementia among the general population in order to develop geographically targeted interventions, no studies have examined the geographic distribution of subjective cognitive decline (SCD) among military veterans specifically.
Objectives: To map the geographic distribution of subjective cognitive decline from 2011-2019 in the United States among military veterans.
Design: Cross-sectional.
Setting: United States.
Participants: Individuals reporting previous service in the United States Armed Forces.
Measurements: Using 2011 Behavioral Risk Factor Surveillance System (BRFSS) data, which is last year for which geocoded SCD data is publicly available, we estimated the survey-weighted county-level prevalence of veteran SCD for counties with >30 veterans (43 counties in 7 states). We then developed a Fay-Herriot small area estimation linear model using auxiliary data from the Census, with county-level veteran-specific covariates including % >65 years old, % female, % college educated, and median income. Following model validation, we created beta-weighted predictions of veteran SCD for all USA counties for 2011-2019 using relevant time-specific Census auxiliary data. We provide choropleth maps of our predictions.
Results: Results of our model on 43 counties showed that county-level rates of SCD were significantly associated with all auxiliary variables except annual income (F = 1.49, df = 4, 38). Direct survey-weighted rates were correlated with model-predicted rates in 43 counties (Pearson r = 0.32). Regarding predicted rates for the entire USA, the average county-level prevalence rate of veteran SCD in 2011 was 13.83% (SD = 7.35), but 29.13% in 2019 (SD = 14.71) – although variation in these rates were evident across counties.
Conclusions: SCD has increased since 2011 among veterans. Veterans Affairs hospitals should implement plans that include cognitive assessments, referral to resources, and monitoring patient progress, especially in rural areas.

Key words: Alzheimer’s disease, subjective cognitive decline, veterans, geographic distribution, country, maps.



Subjective cognitive decline (SCD), which is a self-reported measure of increasing severity of memory loss or confusion, may be one of the earliest symptoms of Alzheimer’s disease (1). Undertaking an intervention as soon as possible following noticeable SCD may delay its progress, and thus, its subsequent deleterious health effects, including mortality (2). In the United States (U.S.) an estimated 6.2 million people have Alzheimer’s disease (3, 4). An estimated one in nine people in the U.S. general population aged 45 and older (11.1%) report SCD, a figure that rises to 11.6% for persons 65 and older (5). The U.S. Centers for Disease Control and Prevention (CDC) labels SCD as a public health issue of growing concern (6).
Taylor et al. (7) show that military veterans are more likely to report SCD than nonveterans (13.6% vs. 10.8%), possibly due to the increased risk of exposure to risk factors during military service (8). Although some studies have examined the distribution of cognitive decline in the general population to develop geographically targeted interventions (9, 10), no studies have examined the geographic distribution of SCD among military veterans specifically. Given the lack of a published choropleth map of military veteran SCD rates in the U.S. that would enable a focus on geographically targeted interventions, we sought identification of temporal and geographic trends in SCD rates among military veterans.



Data Source and Sample

We acquired data from the CDC’s 2011 Behavioral Risk Factor Surveillance System (BRFSS), the most recent year for which subjective cognitive decline data are publicly available with county-level geocodes (11). BRFSS is a telephone survey of non-institutionalized adults (e.g., persons who are not in nursing homes or on active duty in the military) in the U.S., aged > 17 years. Topics covered in the BRFSS include, but are not limited to, health behaviors, healthcare access, and chronic diseases. In 2011, the BRFSS had a response rate of 53% for landline telephone respondents and 28% for cell phone respondents.
During the 2011 BRFSS survey waves, seven states included the subjective cognitive decline survey questions: Hawaii, Illinois, New Hampshire, South Carolina, Tennessee, West Virginia, and Wisconsin. Consequently, the analytic sample in this study was limited to residents in those seven states. We delimited the dataset to individuals who answered in the affirmative to the following question about military service status (n = 6,108): “Have you ever served on active duty in the United States Armed Forces, either in the regular military or in a National Guard or military reserve unit?”


SCD was measured using the following question: “During the past 12 months, have you experienced confusion or memory loss that is happening more often or is getting worse?” Closed-ended response options to the aforementioned question included yes, no, don’t know, or refuse to answer. Of 6,108 military veteran respondents, 53 veterans indicated that they did not know whether they had experienced SCD and 93 refused to answer the question. Subsequently, we delimited the dataset only to military veterans who responded with an answer of yes or no to the SCD question (n = 5,962).
The BRFSS protocol also asked participants: “What county do you live in?” In response to this question, survey participants stated the name of the county in which they lived at the time of the survey. The CDC subsequently coded each person’s geographic location response with a federal information processing standard (FIPS) number.

Data Analysis

Phase 1

For the present study, we followed the area-level Fay-Herriot (12) approach to estimate the prevalence of SCD among military veterans in small areas (i.e., counties). As Li and Lahiri (13) note, “the Fay-Herriot model has been widely used in small area estimation and related problems for a variety of reasons, including its simplicity, its ability to protect confidentiality of microdata and its ability to produce design-consistent estimators” (p. 882). Given our analysis involves a subset of the general population (e.g., military veterans), we selected the Fay-Herriot approach especially for it’s ability to retain the confidentiality of microdata. In the first phase of our analysis, we calculated direct survey-weighted prevalence rates of SCD for each county in the BRFSS dataset. Only counties with ≥ 30 military veterans in the denominator of this rate were included in the analysis given that the National Center on Health Statistics has suggested that rates with denominators < 30 produce unstable estimates (14). In total, our analyses revealed that 43 counties met the criteria for inclusion in the study, with county samples ranging from 32 to 599 (M = 87, SD = 92).

Phase 2

Following procedures outlined in Fay and Herriot’s (12) paper, we obtained auxiliary data (i.e., county-level independent variables) for each of the 43 counties in our dataset. Specifically, we obtained 2011 county-level summary measures for each of these counties in the following areas from the U.S. Census Bureau’s American Community Survey (15): (a) the percent of military veterans aged ≥ 65 years, (b) percent of military veterans reporting female as their sex, (c) percent of military veterans with at least a bachelor’s degree, and (d) median annual income for military veterans. The previously listed auxiliary variables have been shown to influence cognitive decline (16). Our selection of 4 auxiliary variables also was determined, in part, by considerations of power and model stability. That is, with only 43 counties in our dataset we did not want to overextend our model with > 4 predictor variables (17). Subsequently, we regressed our county-level prevalence rates of SCD on these 4 county-level auxiliary variables in a multivariable regression model. Although it would have been useful to explore interaction effects in this model, we were unable to do so owing to the limited sample size. Model estimation was performed in Stata IC version 16.

Phase 3

After we fit a model to the data, we conducted an internal model validation procedure following the guidelines of Zhang et al. (18). Specifically, we calculated a Pearson correlation coefficient for the relationship between our modeled county-level prevalence rates and the county-level prevalence rates based on the raw survey data. We also created a scatterplot for the relationship between these two variables using the “ggplot2” package in R Studio version 3.6.1. As in Zhang et al.’s (18) study, a statistically significant and positive correlation coefficient was indicative of consistency between the modeled rates and the direct survey estimates.

Phase 4

Following model validation, as described in phase 3, we obtained comprehensive and nationally representative county-level summary data (i.e., data for 3,220 counties in the U.S.) on our 4 previously described military veteran-specific auxiliary variables from the 2011-2019 U.S. Census Bureau’s American Community Surveys. Subsequently, we applied the beta coefficients calculated in our linear regression model (i.e., from phase 2) to the U.S. Census Bureau county-level summary data to estimate predicted prevalence rates of SCD for all U.S. counties from 2011 to 2019. Using ESRI ArcGIS version 10.5.1., we developed county-level choropleth maps of our predictions by year, as well as by percentage change from 2011-2019.



Phase 1

Table 1 shows the characteristics of the military veterans included in our 43 counties. All 7 states with available data on SCD also included at least one county with a military veteran sample size ≥ 30. The aggregated prevalence rate of SCD in the sample, Table 1 shows, was 11.96% (SD = 32.50). County-level prevalence rates of SCD ranged from 2.38% (i.e., Shelby County, Tennessee) to 22.02% (i.e., Horry County, South Carolina), with an average of 12.24% (SD = 4.91).

Table 1. Demographic characteristics of the military veteran sample with available data on subjective cognitive decline, BRFSS 2011 (n = 5,962)


Phase 2

Table 2 shows the results of our multivariable linear regression model, where the BRFSS survey-weighted rates of SCD (n = 43) were regressed on 4 auxiliary variables obtained from the U.S. Census Bureau. Although we fit models with 1-3 auxiliary variables, the model with all 4 auxiliary variables explained the greatest amount of variance in SCD among veterans. Model 4 indicates that rates of SCD at the county level were positively associated with the percent of veterans aged ≥ 65 years, positively associated with the percent of veterans reporting female as their sex, and negatively associated with the percent of veterans with a college degree.

Table 2. Fay Herriot models for the prediction of subjective cognitive decline among military veterans

a. Unstandardized beta and robust standard error; b. e-5, scientific notation


Phase 3

We obtained predictions from the model described in phase 2 for all 43 counties for which military veteran data was available. We calculated a Pearson correlation coefficient between the county-level predictions and the direct survey estimates. As Figure 1 shows, our internal model validation procedure revealed that SCD rates modeled with our linear regression model were consistent with the direct survey estimates (r = 0.32, p = 0.03).

Figure 1. Scatterplot illustrating the relationship between the predicted prevalence rates of SCD based on our linear regression model and the direct survey-weighted prevalence estimates (n = 43)


Phase 4

Given the results of our internal model validation procedure, we used the model coefficients [y = –0.19 + (0.02*X1) + (0.04*X2) + (0.000016*X3) – (0.02*X4)] to estimate the prevalence of SCD among military veterans in 3,220 counties in the U.S. from 2011 to 2019. Figure 2 shows the results of our predictions across the 9-year study period. In 2011, the average county-level prevalence rate of veteran SCD, across all counties in the U.S. (i.e., not just the original 43 counties), was 13.83% (SD = 7.35), but in 2019, the estimated average county-level prevalence rate of veteran SCD was 29.13% (SD = 14.71) – although variation in these rates were evident across U.S. counties.
Using the formula for percent change (i.e., Y2 – Y1/Y1), we examined changes in SCD rates among military veterans by county for 2011-2019. Our analysis revealed that 2,948 counties (91.55%) had increases in predicted military veteran SCD for 2011-2019, with counties in Texas, Georgia, Kentucky, Kansas, West Virginia, Montana, and Puerto Rico having the highest increases (i.e., > 1000% increase).

Figure 2. Predicted county-level prevalence rates of SCD among military veterans from 2011-2019 in the United States



Using publicly available, de-identified data from a nationally representative health indicator survey and auxiliary data from the U.S. Census Bureau, we were able to estimate a predictive model of SCD among military veterans. We subsequently used our predictive model to estimate the prevalence of SCD among military veterans in 3,220 counties across the U.S. Results showed that the prevalence of military veteran SCD increased in 91.55% of counties, with counties in Texas, Georgia, Kentucky, Kansas, West Virginia, Montana, and Puerto Rico having the greatest increases.
Research demonstrates that early intervention in Alzheimer’s disease can have quality of life and economic benefits (19). The prevention of Alzheimer’s disease may be carried out with pharmacological (20) or behavioral approaches, especially those that target cardiovascular problems, which are risk factors for cognitive decline (21). For military veterans, specifically, greater enrollment in the U.S. Department of Veterans Affairs (VA) programs focused on physical activity and nutrition, such as the “MOVE!” weight management program (22), may be particularly advantageous in preventing cognitive decline. Complementary and integrative medicine approaches, such as the VA THRIVE program (23), also have been shown to improve mental functioning in military veterans and may be applicable to the prevention of Alzheimer’s disease.


Several limitations accompany the methods and results of the present study. First, although the measure of SCD used in this study has been used widely, it only includes one self-report item and, therefore, may be subject to recall or social desirability bias. Second, several counties in the present study were excluded due to unstable or unavailable data, which prevented a more comprehensive picture of the geographic distribution of SCD in military veterans. Third, although other county level predictors may have been beneficial for inclusion in our predictive model, we were limited due to the small sample size of counties obtained from the BRFSS dataset.
Those limitations notwithstanding, this is the first study to map the prevalence of SCD among military veterans in U.S. counties. Our analysis revealed geographic variation in the distribution of SCD prevalence in the U.S., which implies a need for geographically targeted interventions that may prevent or moderate the progression of cognitive decline in veterans.


Funding: No funding was received for this study.

Conflicts of interest/Competing interests: The authors have no conflicts of interest to disclose.

Availability of data and material:

Code availability: Not applicable.

Ethical standards: Because this study used de-identiied data freely available on the web, this study was considered exempt from IRB review.



1. Alzheimer’s Association. 2018 Alzheimer’s disease facts and figures. Alzheimer’s and Dementia. 2018;14:367-429 1.
2. Crous-Bou M, Minguillón C, Gramunt N, et al. Alzheimer’s disease prevention: From risk factors to early intervention. Alzheimer’s Research and Therapy. 2017; 9:71 2.
3. Alzheimer’s Association. 2021 Alzheimer’s disease facts and figures. Special report. Race, ethnicity and Alzheimer’s in America. Chicago, IL: Alzheimer’s Association. Retrieved 03/292021 from
4. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population estimate of people with clinical AD and mild cognitive impairment in the United States (2020-2060). Alzheimer’s & Dementia. In press.
5. Alzheimer’s Association. Fact sheet. 2020. Retrieved 03/29/2021 from
6. Centers for Disease Control and Prevention. Subjective cognitive decline – a public health issue. 2019. Retrieved 03/29/2021 from
7. Taylor CA, Bouldin ED, McGuire LC. Subjective cognitive decline among adults aged >45 years – United States, 2015-2016. Morbidity and Mortality Weekly Report. 2018;67(27):753-757 7.
8. Weiner MW, Friedl KE, Pacifico A, et al. Military risk factors for Alzheimer’s disease. Alzheimer’s and Dementia. 2013;9(4):445-451 8. 10.1016/j.jalz.2013.03.005.
9. Kirson NY, Meadows ES, Desai U, Smith BP, Cheung HC, Zuckerman P, Matthews BR. Temporal and geographic variation in the incidence of Alzheimer’s disease diagnosis in the US between 2007-2014. Journal of the American Geriatrics Society. 2019;68(2): 346-353 9. 10.1111/jgs.16262.
10. Fowler M, DiBlasio C, Crowe M, Kennedy RE. Geographic disparities in county-level prevalence of Alzheimer’s disease across the United States. Innovations in Aging. 2019;3(1):S782.
11. Centers for Disease Control and Prevention. BRFSS 2011 survey data and documentation. 2011. Retrieved 1/2/2021 from
12. Fay RE, Herriot RA. Estimates of income for small places: An application of James-Stein procedures to census data. Journal of the American Statistical Association. 1979;74:269-277 12.
13. Li H, Lahiri P. An adjusted maximum likelihood method for solving small area estimation problems. J Multivariate Analysis. 2010;101(4):882-982 13.
14. Parker JD, Talih M, Malec DJ, et al. National Center for Health Statistics data -presentation standards for proportions. National Center for Health Statistics. Vital Health Statistics. 2017;2(175):1-22 14.
15. United States Census Bureau. Advanced search. 2021. Retrieved 1/5/2021 from
16. Evans DA, Hebert LE, Beckett LA, Scherr PA, Albert MS, Chown MJ, Pilgrim DM, Taylor JO. Education and other measures of socioeconomic status and risk of incident Alzheimer disease in a defined population of older persons. Archives of Neurology. 1997;54:1399-1405. 16. 10.1001/archneur.1997.00550230066019
17. VanHoorhis CRW, Morgan BL. Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology. 2017;3(2):43-50. 17. 10.20982/tqmp.03.2.p043
18. Zhang X, Holt JB, Yun S, Lu H, Greenlund KJ, Croft JB. Validation of multilevel regression and poststratification methodology for small area estimation of health indicators from the Behavioral Risk Factor Surveillance System. American Journal of Epidemiology. 2015;182(2):127-137. 18. 10.1093/aje/kwv002
19. Barnett JH, Lewis L, Blackwell AD, Taylor M. Early intervention in Alzheimer’s disease: A health economic study of the effects of diagnostic timing. BMC Neurology. 2014;14,101. 19. 10.1186/1471-2377-14-101
20. Lopez C, Caputo A, Liu F, Riviere ME, Rouzade-Dominguez ML, Thomas RG, Langbaum JB, Lenz R, Reiman EM, Graf A, Tariot PN. The Alzheimer’s prevention initiative generation program: Evaluating CNP520 efficacy in the prevention of Alzheimer’s disease. Journal of the Prevention of Alzheimer’s Disease. 2017;4(4):242-246. 20. 10.14283/jpad.2017.37
21. Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential for primary prevention of Alzheimer’s disease: An analysis of population based data. Lancet Neurology. 2014;13:788-794. 21. 10.1016/S1474-4422(14)70136-X
22. United States Department of Veterans Affairs. MOVE! Weight management program. 2021. Retrieved 3/9/2021 from
23. Haun JN, Paykel J, Alman AC, Patel N, Melillo C. A complementary and integrative health group-based program pilot demonstrates positive health outcomes with female veterans. EXPLORE. 2020;16(2):85-89. 23.





P.S. Aisen1, R.J. Bateman2, M. Carrillo3, R. Doody4, K. Johnson5, J.R. Sims6, R. Sperling7, B. Vellas8 and the EU/US CTAD Task Force


*EU/US/CTAD TASK FORCE: Rebecca Amariglio (Boston); Sandrine Andrieu (Toulouse); Rhoda Au (Boston); Clive Ballard (Exeter); Matthew Barton (Raleigh); Monika Baudler (Basel); Joanne Bell (Willmington); Kevin Biglan (Indianapolis); Adam Boxer (San Francisco); Sasha Bozeat (Basel); Claudine Brisard (Issy les Moulineaux); Miroslaw Brys (Indianapolis); Marc Cantillon (Gilbert); Bill Chan (Beijing); Ivan Cheung (Woodcliff Lake); Min Cho (Woodcliff Lake); Julia Coelho (San Francisco); Shobha Dhadda (Woodcliff Lake); Daniel Di Giusto (Basel); Michael Donohue (San Diego); John Dwyer (Washington); Michael Egan (North Wales); Rianne Esquivel (Malvern); Howard Fillit (New York); Wendy Galpern (New Jersey); Harald Hampel (Woodcliff Lake); Jason Hassenstab (St Louis); David Henley (New Jersey); Joseph Herring (North Wales); Carole Ho (South San Francisco); Michael Irizarry (Woodcliff Lake); Takeshi Iwatsubo (Tokyo); Gustavo Jimenez-Maggiora (San Diego); Jeffrey Kaye (Portland); Geoffrey Kerchner (South San Francisco); Gene Kinney (South San Francisco); Shailaja Korukonda (Woodcliff Lake); Lynn Kramer (Woodcliff Lake); Jaren Landen (Cambridge); Ishani Landri (Woodcliff Lake); Lars Lannfelt (Uppsala); Valérie Legrand (Nanterre); Simon Lovestone (Oxford); Manoj Malhotra (Woodcliff Lake); Eric McDade (St Louis); Fanni Natanegara (Indianapolis); Francisco Nogueira (South San Francisco); Gerald Novak (New Jersey); Gunilla Osswald (Stockholm); Susanne Ostrowitzki (South San Francisco); Amanda Paley (New York); Kathryn Papp (Boston); Martin Rabe (Woodcliff Lake); Rema Raman (San Diego); Elena Ratti (Cambridge); Laurie Ryan (Bethesda); Stephen Salloway (Providence); Peter Schüler (Langen); Hiroshi Sekiya (Malvern); Jiong Shi (Las Vegas); Melanie Shulman (Cambridge); Eric Siemers (Zionsville); Kaycee Sink (South San Francisco); Maria Soto-Martin (Toulouse); Joyce Suhy (Newark); Jina Swartz (London); Pierre Tariot (Phoenix); Edmond Teng (South San Francisco); Jacques Touchon (Montpellier); Martin Traber (Basel); Andrea Vergallo (Woodcliff Lake); Judy Walker (Singapore); Jon Walsh (San Francisco); Michael Weiner (San Francisco); Alette Wessels (Indianapolis); Haichen Yang (North Wales); Wagner Zago (San Francisco); Kenton Zavitz (Cambridge)

1. Alzheimer’s Therapeutic Research Institute (ATRI), Keck School of Medicine, University of Southern California, San Diego, CA, USA; 3. Alzheimer’s Association, Chicago, IL, USA; 4. Genentech/Roche, Basel, Switzerland; 5. Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 6 – Eli Lilly and Company, Indianapolis, IN, USA; 7. Brigham and Women’s Hospital, Boston MA, USA; 8. Gerontopole, INSERM U1027, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France

Corresponding Author: P.S. Aisen, University of Southern California Alzheimer’s Therapeutic Research Institute, San Diego, CA, USA,

J Prev Alz Dis 2021;3(8):306-312
Published online May 24, 2021,



A diverse range of platforms has been established to increase the efficiency and speed of clinical trials for Alzheimer’s disease (AD). These platforms enable parallel assessment of multiple therapeutics, treatment regimens, or participant groups; use uniform protocols and outcome measures; and may allow treatment arms to be added or dropped based on interim analyses of outcomes. The EU/US CTAD Task Force discussed the lessons learned from the Dominantly Inherited Alzheimer’s Network Trials Unit (DIAN-TU) platform trial and the challenges addressed by other platform trials that have launched or are in the planning stages. The landscape of clinical trial platforms in the AD space includes those testing experimental therapies such as DIAN-TU, platforms designed to test multidomain interventions, and those designed to streamline trial recruitment by building trial-ready cohorts. The heterogeneity of the AD patient population, AD drugs, treatment regimens, and analytical methods complicates the design and execution of platform trials, yet Task Force members concluded that platform trials are essential to advance the search for effective AD treatments, including combination therapies.

Key words: Alzheimer’s disease, anti-amyloid therapies, anti-tau therapies, platform trials, adaptive trial design, shared placebos, master protocols, secondary prevention.



Disappointing clinical trial results continue to mount in the search for effective treatments for Alzheimer’s disease (AD) despite increasing knowledge about underlying mechanisms and potential therapeutic targets (1,2). As trials move into earlier stages of the disease with a focus on secondary and even primary prevention, the need for innovative trial designs and improved biomarker technologies has become increasingly evident. In December 2020 as the COVID-19 pandemic caused the suspension of many non-COVID-19 clinical trials worldwide (3), the EU/US Clinical Trials in Alzheimer’s Disease Task Force met virtually to explore platform trials as a means of accelerating and improving the efficiency and success of AD drug development. The Task Force brought together leaders from existing and planned AD platform trials to discuss their experiences to date as well as plans for moving forward. Together, they examined recent experience from the Dominantly Inherited Alzheimer Network Trials Unit (DIAN-TU) platform trial and other platform trials that have launched or are in the planning stages, and that provide lessons applicable to the design of future trials.
Platform trials are those conducted within an infrastructure that enables simultaneous and perpetual assessment of multiple therapeutics, treatment regimens, and/or participant groups. They have the potential to increase efficiency by minimizing screen failures; using uniform outcome measures, protocols, and consents; and testing multiple targets and drugs in parallel rather than serially. Adaptive platform trials may allow new arms to be added or individual arms to be stopped if interim analysis indicates a failure to demonstrate efficacy.
However, platform trials can be highly complex both operationally and analytically, requiring a great deal of coordination to bring multiple stakeholders together at the same time and to produce convincing results.


Lessons learned from DIAN-TU

DIAN-TU and the DIAN Observational study (DIAN-Obs) are public-private partnerships created to facilitate the development of AD therapeutics and advance scientific understanding of the optimal ways to prevent and treat AD. To accomplish these objectives, DIAN enrolled participants from around the world with dominantly inherited AD (DIAD) mutations that confer nearly 100 percent certainty of developing AD. By longitudinally monitoring both symptomatic and presymptomatic participants with clinical assessments and physiologic and pathologic biomarkers, DIAN established the progression of biomarkers across the continuum of the disease and demonstrated a biomarker profile associated with presymptomatic disease (4). This provided a rationale for starting a prevention trial in the presymptomatic stage. Figure 1 illustrates the relationship of a range of biomarkers to disease stage and how that translates to the potential for primary and secondary prevention as well as symptomatic treatment.

Figure 1. Stages of Pathology and Disease in Dominantly Inherited Alzheimer’s Disease

Bateman, Randall J., Chengjie Xiong, Tammie L.S. Benzinger, Anne M. Fagan, Alison Goate, Nick C. Fox, Daniel S. Marcus, et al. “Clinical and Biomarker Changes in Dominantly Inherited Alzheimer’s Disease.” New England Journal of Medicine 367, no. 9 (August 30, 2012): 795–804. Reprinted with permission.


DIAN-TU built a flexible and robust platform based on observational data from DIAN-Obs along with disease progression models, comprehensive biomarker studies, input from participants and family members, and inclusive discussions with stakeholders. The platform enabled the design and execution of prevention studies that can accommodate a range of prevention and treatment trials. Since the rarity of these mutations limits the number of trial participants, DIAN-TU has established study sites around the globe: in North and South American, Europe, Japan, Korea, and Australia, with additional sites planned in China and other countries.
The DIAN-TU Trial Platform is designed to improve the efficiency of trials by:
• testing multiple drugs and targets in parallel
• optimizing the trial design in view of the limited number of participants
• using a pooled placebo and control group, building up data over time to accumulate power to detect a drug effect
• evaluating the magnitude of target engagement based on each drug’s mechanism of action
• adapting the trial in response to biomarker findings, magnitude of drug effects at different doses, and safety signals
• incorporating novel biomarkers
• applying learnings from each study into future studies in the platform.

The DIAN-TU platform launched in 2012 as a 4-year secondary prevention trial in participants with DIAD mutations and expected onset ranging from -15 to +10 years and Clinical Dementia Rating (CDR) score of 0 (cognitively normal), 0.5 (very mild cognitive impairment), or 1 (mild cognitive impairment consistent with early dementia) (6). Nearly 200 participants were enrolled, randomized to receive Eli Lilly’s solanezumab (a soluble anti-amyloid-beta monoclonal antibody), Roche’s gantenerumab (an aggregated anti-amyloid monoclonal antibody), or placebo. A third drug arm testing a beta-secretase inhibitor was launched but was terminated early because of safety concerns.
As the two-drug trial continued, adaptations to the platform were made. Planned as a two-year biomarker trial, the study transitioned to a four-year cognitive endpoint trial after the biomarker analysis confirmed that the drugs were engaging their targets. Part way through the trial, a decision was made to increase the dose of both drugs based on external data.
Other components were also added to the platform. These included a run-in period during which data from cognitive measures and tau PET studies were collected in preparation for launching three tau next generation (NexGen) drug arms. Meanwhile, a primary prevention trial in DIAD participants 10 years before symptom onset is in the launch preparation stage. This trial will use a beta amyloid (Aβ)-targeting drug with the aim of preventing the development of plaques before they start.
Topline results for the solanezumab/gantenerumab trial, reported in early 2020, showed that both drugs failed to meet the prespecified primary cognitive endpoint (7). Nonetheless, the platform itself succeeded in meeting many of its objectives and has provided extensive data that continues to be analyzed. While DIAN-TU has not yet confirmed the clinical or cognitive effectiveness of a therapy for DIAD, the ongoing studies have gathered important data relevant to the validity of the amyloid hypothesis, the timing of treatment, and the use of biomarkers to track therapeutic effectiveness.
These analyses suggest two possible reasons that may explain the failure to detect a significant cognitive difference between groups: 1) although the dose was increased part way through the trial, most participants had received lower doses; by the time they were switched to the high dose, many symptomatic participants had already declined substantially, thus limiting the ability to evaluate the effect of the high dose; 2) some asymptomatic trial participants failed to decline and in some cases improved; thus it was not possible to detect a drug effect in those participants.
The continuing data analysis also suggested that gantenerumab showed significant biologic effects on downstream biomarkers, suggesting that higher doses and longer treatment periods may be needed, especially for asymptomatic mutation carriers. While substantially advancing the field, these results were not able to definitively test the hypothesis. Nonetheless, gantenerumab is being studied in an ongoing exploratory open label extension (8).
Other lessons learned from the trial include the need to enrich the study for participants likely to show signs of cognitive decline during the treatment period. The optimal approach to this challenge is not clear, but further analysis of biomarker data from this trial may provide some insight. In addition, other research suggests that some novel biomarkers or combinations of markers may be predictive of incipient cognitive decline. How they would be implemented in screening for trials, however, would present additional complexity. Researchers also continue to search for more sensitive cognitive assessments and tools that minimize practice effects, which could also potentially improve the ability to detect a treatment effect. Yet even if a test is developed to detect subtle cognitive change, the ability of that test to be applied to an AD trial could be limited if the cognitive change results from non-AD aging effects.


AHEAD 3-45 Platform Study

The AHEAD Study is a platform-based study comprising two Phase 3 secondary prevention trials — the A45 trial and the A3 trial. Both four-year trials are testing Eisai’s anti-Aβ monoclonal antibody lecanemab (BAN 2401) but use a different dosing regimen depending on baseline amyloid burden. Together they are evaluating the efficacy of the treatment across the continuum of early-to-late preclinical AD as determined by positron emission tomography (PET) amyloid assessment. Using different radioligands, PET imaging enables the visualization of both amyloid plaques and tau neurofibrillary tangles. The AHEAD trials build on the findings of the Harvard Aging Brain Study (HABS) showing that early cortical amyloid accumulation is associated with subsequent tau accumulation in the inferior temporal neocortex, which is associated with neurodegeneration and cognitive decline (9). This study and others in separate cohorts have shown that a period of rapid acceleration in rate of Aβ accumulation immediately precedes reaching the threshold of amyloid positivity that heralds cognitive decline (10–12). These studies led to the hypothesis that decreasing amyloid accumulation at the earliest detectable stage will provide the best opportunity to slow progression and prevent cognitive decline.
The A3 trial aims to capture people with a measurable amyloid PET signal that is below a conventional threshold of amyloidosis. These participants with “intermediate” levels of amyloid (early preclinical or “pre-preclinical” AD) will be randomized to receive the drug or placebo monthly. Amyloid PET will be used as the primary outcome measure to determine whether lecanemab slows amyloid accumulation. Tau PET will be a key secondary outcome and cognitive testing will be an exploratory outcome.
In the A45 study, participants with elevated amyloid will be randomized to receive the drug or placebo biweekly for the first two years followed by monthly maintenance dosing, with cognitive testing as the primary outcome. Amyloid and tau PET will be key secondary outcomes, and biomarker studies will also be an essential component of the trial along with additional cognitive and participant-reported outcomes.
The first participants in AHEAD were screened in July and randomized in September. Worldwide, 100 sites have been identified. The AHEAD platform will use a common screening protocol for the two trials, 18F NAV4694 amyloid and 18F MK6240 tau PET at baseline, 2 and 4 years, and will enroll participants between the ages of 55 and 80. Participants under age 65 will need to have one additional risk factor (e.g., family history, APOE carriage, or known amyloid status). In response to the COVID-19 pandemic, the study team also plans to implement home infusion and potentially, supervised remote assessments.


Tau platforms

Other platforms are currently being planned to evaluate the efficacy of anti-tau drugs. While anti-tau therapies have gained support in the Alzheimer’s community because tau is thought responsible for neurodegeneration and cognitive impairment and because of the disappointing results of so many anti-amyloid therapies, many challenges remain due to the complexity of tau biology and the incomplete understanding of the relationship between tau and Aβ (13,14). Moreover, several recent trials of anti-tau monoclonal antibodies have failed to demonstrate efficacy, raising concerns that the disappointing search for an anti-amyloid therapy may be repeated for tau therapies (15). Nonetheless, these negative anti-tau trials may in fact highlight even more strongly the importance of a platform approach.
Eli Lilly and Company has proposed using a master protocol, which should facilitate greater efficiency and accelerated drug development through the use of a common protocol, trial design, and infrastructure. Master protocols can be used in umbrella, basket, or adaptive platform trials to evaluate multiple therapies in the context of a single disease, a single therapy in the context of multiple diseases or disease subtypes, or multiple therapies in a single disease with arms added and/or terminated over time (16). For example, Lilly developed an innovative master protocol to evaluate multiple interventions for chronic pain, in which three different therapeutics were tested across three different pain states.
Master protocols offer the opportunity for increased flexibility to explore multiple hypotheses and different drug combinations or patient populations. As with other platform trials they allow for shared placebo data, unbalanced randomization so that more participants are assigned to the active drug arm, and the potential for a lower screen failure rate. While master protocols may offer increased efficiency for testing multiple therapeutics in parallel, tradeoffs include increased complexity, reduced efficiency gains if fewer therapeutics enter as well as diminished flexibility compared to a single therapeutic trial. It is essential to consider the mechanistic differences among candidate tau therapeutics (and among immunotherapies targeting different epitopes) which may point to different selection criteria and outcome measures.
To study treatments for tauopathies, a master protocol could be used to test multiple therapeutics in different tauopathies (e.g. AD, Primary Supranuclear Palsy [PSP], and Corticobasal degeneration [CBD]); multiple therapeutics in one tauopathy at different stages of disease (late stage AD, early symptomatic AD, and preclinical AD); or multiple therapeutics in only one stage (e.g., preclinical AD). Lilly’s consideration of a master protocol approach to testing multiple anti-tau assets could use one or a combination of these structures, although all of them present substantial operational costs and risks.
A second tau platform protocol is being developed by the Alzheimer’s Clinical Trial Consortium (ACTC). Developers of the Alzheimer’s Tau Platform (ATP) envision a proof-of-concept platform to accelerate decision making in tau therapeutic development by simultaneously testing different anti-tau mechanisms in sporadic AD. They argue that a platform that enables testing multiple molecules and multiple regimens using factorial adaptive designs would improve the efficiency of recruitment, trial startup, and analyses. They also suggest that demonstrating efficacy may be possible more quickly with anti-tau versus anti-amyloid therapies since recent evidence suggest that tau PET agents can detect highly dynamic change in some individuals, e.g. those with higher levels of amyloid (17).
DIAN-TU is also planning to add three anti-tau drug arms to its study in 2021. Since platforms can be especially helpful in assessing treatment response at different stages of disease, each arm will likely enroll both presymptomatic and symptomatic participants. Moreover, by testing different mechanisms and targets in the platform and using short-term biomarker studies, the platform approach may be able to get information on which targets are druggable at which stages of disease.


The landscape of AD platform trials

The concept of platform trials is broad. DIAN-TU was designed to enable testing of multiple drugs using a shared placebo arm, with adaptive allocation based on randomization data. In contrast, the AHEAD 3-45 Platform provides benefits that are primarily operational rather than statistical comparisons between treatment regimens. Platform trials may also be used to explore multifaceted hypotheses, for example, to ask questions such as whether amyloid accumulation must be cleared before targeting tau. Different arms of a platform study can ask different questions simultaneously rather than sequentially to increase efficiency and accelerate drug development.
Longitudinal biomarker and clinical studies in clearly defined cohorts (e.g. APOE homozygotes and heterozygotes) are needed to inform the design of platform programs. A matrix of outcomes (including tau and amyloid biomarkers, imaging, and the neural tool kit) across the continuum of disease could eventually lead to a surrogate marker of efficacy, which could dramatically accelerate the search for effective therapies. For example, the European Prevention of Alzheimer’s Dementia (EPAD) platform began with a longitudinal cohort study to serve as a readiness cohort for proof-of-concept secondary prevention AD trials (18).
Meanwhile, an international network of multidomain intervention clinical trials has been established to replicate the groundbreaking Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER), which demonstrated that a combination of lifestyle interventions reduces the risk of cognitive decline in older adults with vascular risk factors that increase their risk of dementia. The Worldwide FINGERS initiative (WW-FINGERS) provides a platform for investigators in different countries and cultures to adapt the FINGERS protocol while using similar protocols and sharing data (19).
The Trial-Ready Cohort for Preclinical/prodromal Alzheimer’s Disease (TRC-PAD) project is a collaborative effort to establish an efficient mechanism for recruiting participants into very early stage Alzheimer’s disease trials (20, 21). Clinically normal and mildly symptomatic individuals are followed longitudinally in a web-based component called the Alzheimer’s Prevention Trial Webstudy (APT Webstudy), with quarterly assessment of cognition and subjective concerns. The Webstudy data is used to predict the likelihood of brain amyloid elevation; individuals at relatively high risk are invited for in-person assessment in the TRC screening phase, during which a cognitive battery is administered and APOE genotype is obtained followed by reassessment of risk of amyloid elevation (22–24). After an initial validation study, plasma amyloid peptide ratios will be included in this risk assessment. Based on this second risk calculation, individuals may have amyloid testing by PET scan or lumbar puncture, with those potentially eligible for trials followed in the TRC, while the rest are invited to remain in the APT Webstudy.


Addressing the challenges of AD platform trials

The heterogeneity of the AD patient population creates several complications in designing and executing platform trials. For example, the substantial clinical and biomarker heterogeneity in sporadic AD makes screening particularly challenging and increases the sample size needed to obtain adequate power. Regarding screening, accumulating data suggests that blood-based biomarkers have adequate sensitivity and specificity to be advantageous for selecting subjects with AD pathology although they are not yet supported for use as endpoints (25). Incorporating plasma Aβ testing into screening protocols could dramatically reduce the number of PET scans required. Many of the AD platform trials underway or in the planning stages rely on PET scans prior to randomization to determine how far the disease as progressed. For trials testing tau-based therapies, tau PET promises to be especially useful since deposition of neocortical tau is seen even in the presymptomatic stage. PET scans are also commonly used as endpoints and to assess target engagement and treatment response; however, the spectrum of baseline amyloidosis is broad. The growing utility of plasma measures indicative of tau abnormalities will extend biomarker coverage beyond the fibrillar deposits measurable by PET scans. The incorporation of brain donation protocols in DIAN-TU and other trials enables the comparison and confirmation of AD pathology to biomarkers and drug effects on AD pathology.
The heterogeneity of the drugs themselves also increase both the potential benefits and the complexity of platform trials. Multi-arm platforms enable the parallel testing of different approaches and are particularly suited to testing combination therapies. However, the complexity of combination trials is decreased if one drug in a combination approach has been approved for a certain population. While there may be theoretical reasons for pursuing a combination of drugs that target both amyloid and tau, designing a regimen that allows for maximal biological engagement remains unclear, i.e., should amyloid be cleared or neutralized prior to targeting tau, or can targeting tau even in the presence of amyloid stop tau spread and downstream tau effects? Another issue with combination therapies is that two drugs with separate development plans may not be compatible from a logistical standpoint.
Different modes of administration further complicate platform trials. Participants’ acceptance of a treatment may vary by mode of administration (e.g., oral more likely to be accepted than an intrathecal treatment). Combining placebos when there are different modes of administration may not be feasible.
Platform trials may employ different analytical methods to test hypotheses, address subgroup effects, and compare outcomes against control groups (26). Interim analysis offers both benefits and risks. While interim analysis may be informative and can protect subjects from risks of harm or better target limited resources by eliminating a drug that has little chance of success, early decisions may be inaccurate and prematurely ending an arm may also limit what can be learned from the study or reduce confidence in an approach that may ultimately work. Task Force members agreed that planning for interim analysis ahead of time is essential. This includes determining the types of data to be analyzed (e.g. biomarkers vs. cognitive outcomes) and the timepoints at which interim analyses should be conducted. For exploratory, hypothesis-generating platform trials where the goal is to elucidate pathophysiological interactions between targets and drugs, shorter studies with no interim analysis may be appropriate. However, in platform trials with registration as an ultimate goal, carefully designed interim analysis may be appropriate.
Platform studies may also be challenging from a company perspective since they are expensive, complex, and may compete for resources with individual molecule programs. Collaboration of different company sponsors has the potential to share costs and risks, but presents other challenges including protection of intellectual property and data. Industry representatives acknowledged, however, that as data accumulate in a platform and the ability to use that data grows, care must be taken to ensure the data are appropriately used. The Task Force also recognized the importance of ensuring that the community of funders and sponsors understand the benefits and challenges associated with platform trials and that both public and private support are essential for this complex endeavor.
Despite the challenges, many Task Force members believe that platform trials are essential to ensure that the field builds on and moves past the previous disappointments that have plagued amyloid trials. Moreover, they are necessary since conducting trials sequentially, drug by drug and severity stage by severity stage, takes much too long. Platform trials provide other opportunities that are essential for AD drug development by testing the predictive value of biomarker status and the surrogate value of biomarkers; evaluating individual drugs across multiple stages of disease; and accelerating the timing of testing combinations of drugs, which have historically waited until each drug in the combination has shown efficacy. Additionally, the Task Force advocated the importance of continued collaborations and partnerships to ensure that multiple investigators analyze multiple datasets, share data and knowledge, and learn from one another to gain insight into how to defeat this deadly disease.


Acknowledgements: The authors thank Lisa J. Bain for assistance in the preparation of this manuscript. Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen, outside the su

Conflicts of interest: The Task Force was partially funded by registration fees from industrial participants. These corporations placed no restrictions on this work. Dr. Aisen reports grants from Janssen, NIA, FNIH, Alzheimer’s Association, and Eisai, personal fees from Merk, Biogen, Roche, ImmunoBrain Checkpoint, Abbvie, Rainbow Medical, and personal fees from Shionogi, outside the submitted work. Dr. Bateman reports grants from the Alzheimers Association, NIH, FNIH, GHR Foundation, Eli Lilly and Company, Hoffman-LaRoche, Avid Radiopharmaceuticals, Janssen, Eisai, Genetech Abbvie, Biogen, Centene, United Neuroscience, and an anonymous organization. In-kind support from CogState and Signant. Personal fees from Hoffman-LaRoche, Janssen, Eisai, C2N Diagnostics, AC Immune, Amgen, and Pfizer. Dr. Carrillo does not have any COI and is a full time empolyee of the Alzheimer’s Assn. Dr. Doody is a full-time employee of F. Hoffman LaRoche/Genentech; Dr. Johnson reports personal fees from Novartis, AC Immune, Janssen and Cerveau, outside the submitted work. Dr. Sims is an empolyee of Lilly. Dr. Sperling reports grants from Eli Lilly, NIA, Alzheimer’s Association, Janssen, Eisai, personal fees from Shionogi, Genentech, Oligomerix, Inc., Cytox, Prothena, Acumen, JOMDD, Renew, Alnylam Pharmaceuticals, Neuraly, Janssen, Neurocentria, AC Immune, Biogen, Eisai, Roche and Takeda Pharmaceuticals, outside the submitted work. Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen, outside the submitted work.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Rafii MS, Aisen PS. The search for Alzheimer disease therapeutics – same targets, better trials? Nat Rev Neurol. 2020 Nov;16(11):597–8.
2. Cummings J, Lee G, Ritter A, Sabbagh M, Zhong K. Alzheimer’s disease drug development pipeline: 2020. Alzheimer’s & Dementia: Translational Research & Clinical Interventions. 2020;6(1):e12050.
3. van Dorn A. COVID-19 and readjusting clinical trials. The Lancet. 2020 Aug 22;396(10250):523–4.
4. Bateman RJ, Benzinger TL, Berry S, Clifford DB, Duggan C, Fagan AM, et al. The DIAN-TU Next Generation Alzheimer’s prevention trial: adaptive design and disease progression model. Alzheimers Dement. 2017 Jan;13(1):8–19.
5. Bateman RJ, Xiong C, Benzinger TLS, Fagan AM, Goate A, Fox NC, et al. Clinical and biomarker changes in dominantly inherited Alzheimer’s disease. N Engl J Med. 2012 Aug 30;367(9):795–804.
6. Morris JC. Clinical Dementia Rating: A Reliable and Valid Diagnostic and Staging Measure for Dementia of the Alzheimer Type. Int Psychogeriatr. 1997 Dec;9(S1):173–6.
7. Strobel G. Topline Result for First DIAN-TU Clinical Trial: Negative on Primary | ALZFORUM [Internet]. ALZFORUM. 2020 [cited 2021 Feb 12]. Available from:
8. Strobel G. In DIAN-TU, Gantenerumab Brings Down Tau. By a Lot. Open Extension Planned [Internet]. ALZFORUM. 2020 [cited 2021 Feb 12]. Available from:
9. Hanseeuw BJ, Betensky RA, Jacobs HIL, Schultz AP, Sepulcre J, Becker JA, et al. Association of Amyloid and Tau With Cognition in Preclinical Alzheimer Disease: A Longitudinal Study. JAMA Neurol. 2019 Jun 3;
10. Farrell ME, Jiang S, Schultz AP, Properzi MJ, Price JC, Becker JA, et al. Defining the lowest threshold for amyloid-PET to predict future cognitive decline and amyloid accumulation. Neurology. 2020 Nov 16;
11. Villain N, Chételat G, Grassiot B, Bourgeat P, Jones G, Ellis KA, et al. Regional dynamics of amyloid-β deposition in healthy elderly, mild cognitive impairment and Alzheimer’s disease: a voxelwise PiB-PET longitudinal study. Brain. 2012 Jul;135(Pt 7):2126–39.
12. Jack CR, Therneau TM, Wiste HJ, Weigand SD, Knopman DS, Lowe VJ, et al. Transition rates between amyloid and neurodegeneration biomarker states and to dementia: a population-based, longitudinal cohort study. Lancet Neurol. 2016 Jan;15(1):56–64.
13. Cummings J, Blennow K, Johnson K, Keeley M, Bateman RJ, Molinuevo JL, et al. Anti-Tau Trials for Alzheimer’s Disease: A Report from the EU/US/CTAD Task Force. J Prev Alzheimers Dis. 2019;6(3):157–63.
14. Barthélemy NR, Li Y, Joseph-Mathurin N, Gordon BA, Hassenstab J, Benzinger TammieLS, et al. A soluble phosphorylated tau signature links tau, amyloid and the evolution of stages of dominantly inherited Alzheimer’s disease. Nat Med. 2020 Mar;26(3):398–407.
15. Mullard A. Failure of first anti-tau antibody in Alzheimer disease highlights risks of history repeating. Nature Reviews Drug Discovery [Internet]. 2020 Dec 10 [cited 2020 Dec 10]; Available from:
16. Woodcock J, LaVange LM. Master Protocols to Study Multiple Therapies, Multiple Diseases, or Both. N Engl J Med. 2017 06;377(1):62–70.
17. Sanchez JS, Becker JA, Jacobs HIL, Hanseeuw BJ, Jiang S, Schultz AP, et al. The cortical origin and initial spread of medial temporal tauopathy in Alzheimer’s disease assessed with positron emission tomography. Sci Transl Med. 2021 Jan 20;13(577).
18. Ritchie CW, Muniz-Terrera G, Kivipelto M, Solomon A, Tom B, Molinuevo JL. The European Prevention of Alzheimer’s Dementia (EPAD) Longitudinal Cohort Study: Baseline Data Release V500.0. J Prev Alzheimers Dis. 2020;7(1):8–13.
19. Rosenberg A, Mangialasche F, Ngandu T, Solomon A, Kivipelto M. Multidomain Interventions to Prevent Cognitive Impairment, Alzheimer’s Disease, and Dementia: From FINGER to World-Wide FINGERS. J Prev Alzheimers Dis. 2020 Jan 1;7(1):29–36.
20. Aisen PS, Sperling RA, Cummings J, Donohue MC, Langford O, Jimenez-Maggiora GA, et al. The Trial-Ready Cohort for Preclinical/Prodromal Alzheimer’s Disease (TRC-PAD) Project: An Overview. J Prev Alzheimers Dis. 2020;7(4):208–12.
21. Walter S, Clanton TB, Langford OG, Rafii MS, Shaffer EJ, Grill JD, et al. Recruitment into the Alzheimer Prevention Trials (APT) Webstudy for a Trial-Ready Cohort for Preclinical and Prodromal Alzheimer’s Disease (TRC-PAD). J Prev Alzheimers Dis. 2020;7(4):219–25.
22. Walter S, Langford OG, Clanton TB, Jimenez-Maggiora GA, Raman R, Rafii MS, et al. The Trial-Ready Cohort for Preclinical and Prodromal Alzheimer’s Disease (TRC-PAD): Experience from the First 3 Years. J Prev Alzheimers Dis. 2020;7(4):234–41.
23. Jimenez-Maggiora GA, Bruschi S, Raman R, Langford O, Donohue M, Rafii MS, et al. TRC-PAD: Accelerating Recruitment of AD Clinical Trials through Innovative Information Technology. J Prev Alzheimers Dis. 2020;7(4):226–33.
24. Langford O, Raman R, Sperling RA, Cummings J, Sun C-K, Jimenez-Maggiora G, et al. Predicting Amyloid Burden to Accelerate Recruitment of Secondary Prevention Clinical Trials. J Prev Alzheimers Dis. 2020;7(4):213–8.
25. Palmqvist S, Janelidze S, Quiroz YT, Zetterberg H, Lopera F, Stomrud E, et al. Discriminative Accuracy of Plasma Phospho-tau217 for Alzheimer Disease vs Other Neurodegenerative Disorders. JAMA. 2020 Aug 25;324(8):772.
26. Park JJH, Harari O, Dron L, Lester RT, Thorlund K, Mills EJ. An overview of platform trials with a checklist for clinical readers. Journal of Clinical Epidemiology. 2020 Sep 1;125:1–8.



A.P. Porsteinsson1, R.S. Isaacson2, S. Knox3, M.N. Sabbagh4, I. Rubino5


1. University of Rochester School of Medicine and Dentistry, Rochester, NY, USA; 2. Weill Cornell Medical Center and New York-Presbyterian, New York, NY, USA; 3. Biogen International GmbH, Baar, Switzerland; 4. Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, NV, USA; 5. Biogen Inc, Cambridge, MA, USA

Corresponding Author: Sean Knox, MBChB. Biogen International GmBH, Neuhofstrasse 30, 6340 Baar, Switzerland. Phone: +41413921976; Email:

J Prev Alz Dis 2021;
Published online May 12, 2021,



Alzheimer’s disease is a progressive, irreversible neurodegenerative disease impacting cognition, function, and behavior. Alzheimer’s disease progresses along a continuum from preclinical disease, to mild cognitive and/or behavioral impairment and then Alzheimer’s disease dementia. Recently, clinicians have been encouraged to diagnose Alzheimer’s earlier, before patients have progressed to Alzheimer’s disease dementia. The early and accurate detection of Alzheimer’s disease-associated symptoms and underlying disease pathology by clinicians is fundamental for the screening, diagnosis, and subsequent management of Alzheimer’s disease patients. It also enables patients and their caregivers to plan for the future and make appropriate lifestyle changes that could help maintain their quality of life for longer. Unfortunately, detecting early-stage Alzheimer’s disease in clinical practice can be challenging and is hindered by several barriers including constraints on clinicians’ time, difficulty accurately diagnosing Alzheimer’s pathology, and that patients and healthcare providers often dismiss symptoms as part of the normal aging process. As the prevalence of this disease continues to grow, the current model for Alzheimer’s disease diagnosis and patient management will need to evolve to integrate care across clinical disciplines and the disease continuum, beginning with primary care. This review summarizes the importance of establishing an early diagnosis of Alzheimer’s disease, related practical ‘how-to’ guidance and considerations, and tools that can be used by healthcare providers throughout the diagnostic journey.

Key words: Alzheimer’s disease, early diagnosis, diagnostic work-up.



Dementia is among the greatest global health crises of the 21st century. Currently, more than 50 million people are living with dementia worldwide (1), with this number estimated to triple to 152 million by 2050 as the world’s population grows older (2). Alzheimer’s disease (AD) is the most common cause of dementia and is thought to account for 60–80% of dementia cases (3). Currently, the total annual cost for AD and other dementias in the USA is $305 billion and is predicted to increase to more than $1.1 trillion by 2050 (3). This substantial economic burden includes not only healthcare and hospice support for patients with AD (3) but also lost productivity from patients and caregivers (4).
AD is a progressive, neurodegenerative disease associated with cognitive, functional, and behavioral impairments, and characterized by two underlying pathological hallmarks: the progressive accumulation of extracellular amyloid beta (Aβ) plaques and intracellular neurofibrillary tangles (NFTs) (3). In AD, aggregated Aβ plaques are deposited within the brain as a result of either reduced Aβ clearance or excessive production (5); plaque deposition typically occurs ~20 years before the onset of cognitive impairment (6,7). NFTs are formed by the abnormal accumulation of hyperphosphorylated-tau protein (5); these can be detected 10–15 years before the onset of symptoms (6, 7).
AD follows a progressive disease continuum that extends from an asymptomatic phase with biomarker evidence of AD (preclinical AD), through minor cognitive (mild cognitive impairment [MCI]) and/or neurobehavioral (mild behavioral impairment [MBI]) changes to, ultimately, AD dementia. A number of staging systems have been developed to categorize AD across this continuum (7–9). While these systems vary in terms of how each stage is defined, all encompass the presence/absence of pathologic Aβ and NFTs, as well as deficits in cognition, function, and behavior (7–9). As a result, subtle but important differences exist in the nomenclature for each stage of AD depending on the selected clinical and research classifications (Figure 1).

Figure 1. Stages within the Alzheimer’s disease continuum

The AD continuum can be classified into different stages from preclinical AD to severe AD dementia; the nomenclature associated with each stage varies between the different clinical and research classifications. This figure provides a summary of the different naming conventions that are used within the AD community and the symptoms associated with each stage of the continuum; *Mild behavioral impairment is a construct that describes the emergence of sustained and impactful neuropsychiatric symptoms that may occur in patients ≥50 years old prior to cognitive decline and dementia (112); Abbreviations: Aβ, amyloid beta. AD, Alzheimer’s disease. FDA, Food and Drug Administration. IWG, International Working Group. MCI, mild cognitive impairment. NIA-AA, National Institute on Aging—Alzheimer’s Association


Preclinical AD, as the earliest stage in the AD continuum, comprises a long asymptomatic phase, in which individuals have evidence of AD pathology but no evidence of cognitive or functional decline, and their daily life is unaffected (8) (Figure 1). The duration of preclinical AD can vary between individuals, but typically lasts 6–10 years depending on the age of onset (10,11). The risk of progression from preclinical AD to MCI due to AD (with/without MBI) depends on a number of factors, including age, sex, and apolipoprotein E (ApoE) status (11,12); however, not all individuals who have underlying AD pathology will go on to develop MCI or AD dementia (13,14). A recent meta-analysis of six longitudinal cohorts followed up for an average of 3.8 years found that 20% of patients with preclinical AD progressed to MCI due to AD (11). A further study by Cho et al., with an average follow-up rate of 4 years, found that 29.1% of patients with preclinical AD progressed to MCI due to AD (12).
For patients who do progress to MCI due to AD (with/without MBI), initial clinical symptoms typically include short-term memory impairment, followed by subsequent decline in additional cognitive domains (15) (Figure 1). On a day-to-day basis, an individual with MCI due to AD may struggle to find the right word (language), forget recent conversations (episodic memory), struggle with completing familiar tasks (executive function), or get lost in familiar surroundings (visuospatial function) (15,16). As individuals have varying coping mechanisms and levels of cognitive reserve, patients’ experiences and symptomology vary widely; however, patients tend to remain relatively independent at this stage, despite potential marginal deficits in function. The prognosis for patients with MCI due to AD can be uncertain; one study that followed up patients with MCI due to AD for an average of 4 years found that 43.4% progressed to AD dementia (12). Other studies reported 32.7% and 70.0% of individuals with MCI due to AD progress to AD dementia within 3.2 and 3.6 years of follow-up, respectively (17,18). Patients who do progress to AD dementia will develop severe cognitive deficits that interfere with social functioning and will require assistance with activities of daily living (7) (Figure 1). As the disease progresses further, increasingly severe behavioral symptoms will develop that significantly burden patients and their caregivers, and the disease ultimately results in severe loss of independence and the need for round-the-clock care (3).
An early diagnosis of AD can provide patients the opportunity to collaborate in the development of advanced care plans with their family, caregivers, clinicians, and other members of the wider support team. Importantly, it also enables patients to seek early intervention with symptomatic treatment, lifestyle changes to maintain quality of life, and risk-reduction strategies that can provide clinically meaningful reductions in cognitive, functional, and behavioral decline (19–22). It can also help reduce healthcare system costs and constraints: a study by the Alzheimer’s Association found that diagnosing AD in the early stages could save approximately $7 trillion. These savings were due to lower medical and long-term care costs for patients with managed MCI than for those with unmanaged MCI and dementia (3). Furthermore, an early diagnosis will be vital for patients when a therapy addressing the underlying pathology of AD becomes available; currently 19 biologic compounds are under Phase 2 or 3 investigation (23). Physicians will need to be prepared for the approval of these treatments, to optimize the potential benefit and prolong preservation of patients’ cognitive function and independence beyond that associated with current standard of care (19).
As the prevalence of AD continues to grow, the advancement of AD patient diagnosis will require an orchestrated effort, starting in the primary care setting and subsequently involving multiple healthcare provider (HCP) specialties (e.g., nurse practitioner [NP] or physician assistant [PA]) throughout the disease continuum. Galvin et al. recently highlighted the need for HCPs to work as an integrated, patient-centered care team to accommodate the growing and diverse population of patients with AD, beginning with diagnosis (24). For patients to receive a timely diagnosis, it is vital to implement an approach that minimizes the burden placed on the patient, clinician, and healthcare system (25). Here, we summarize the importance of establishing an early diagnosis of AD, related practical ‘how-to’ guidance and considerations, and tools that can be used by healthcare providers throughout the diagnostic journey.


The importance of an early diagnosis

Historically, a diagnosis of AD has been one of exclusion, and one only made in the latter stages of disease (26); however, the disease process can take years to play out, exacting a significant toll on the patient, caregiver, and healthcare system along the way (27).
To mitigate this burden, the early and accurate detection of AD-associated symptoms in clinical practice represents a critically needed but challenging advancement in AD care (19, 28–30). Usually, a patient with early signs/symptoms of AD will initially present in a primary care setting (30). For some patients, minor changes in cognition and/or behavior may be detected during a routine wellness visit or an appointment to discuss other comorbidities (24). As the PCP is often the first to observe a patient’s initial symptomatology, it is vital they recognize the early signs and symptoms, and understand how to use the most appropriate assessment tools designed to detect these early clinical effects of the disease.
Because the neuropathologic hallmarks of AD (Aβ plaques and NFTs) can be detected decades prior to the onset of symptoms (6, 7), biomarkers reflecting this underlying pathology represent an important opportunity for early identification of patients at greatest risk of developing MCI due to AD. Biomarkers support the diagnosis of AD (especially important early on when symptoms can be subtle), and the U.S. Food and Drug Administration (FDA) has recently published guidelines that endorse their use in this population (9). The National Institute on Aging—Alzheimer’s Association (NIA-AA) has recently created a research framework that acknowledges the use of biomarkers for diagnosing AD in vivo and monitoring disease progression (7).
Important biomarker information can be gathered from imaging modalities such as magnetic resonance imaging (MRI) and positive emission tomography (PET) that visualize early structural and molecular changes in the brain, respectively (25, 30). Fluid biomarker testing, such as cerebrospinal fluid (CSF) can also be used; CSF biomarkers can directly reflect the presence of Aβ and aggregated tau within the brain (7, 31). As will be discussed in more depth later in this article, a large number of clinical studies have shown that Aβ and tau biomarkers can contribute diagnostically important information in the early stages of disease (32). There is ongoing research to expand the current range of tests that can be used by clinicians as part of the multistage diagnostic process (25). For instance, once approved, blood-based biomarkers could be used to identify patients at risk of developing AD and for monitoring disease progression (33, 34), which would also reduce the current capacity constraints associated with PET imaging (25).


Practical guide for an early diagnosis of Alzheimer’s disease in clinical practice

As already raised, recent recommendations for evolving AD care to a more patient-centric, transdisciplinary model include guidance on realizing an efficient diagnostic process—one in which HCPs, payers, and specialists are encouraged to combine their efforts to ensure the early warning signs of AD are not overlooked (24). The recommendations include dividing the diagnosis of AD into the following steps: detect, assess/differentiate, diagnose, and treat (Figure 2). We present here a practical guide for the early diagnosis of AD, based on this outlined approach, including a case study to highlight each of these key steps.

Figure 2. A stepwise infographic to highlight key stages within the diagnostic process, along with the recommended tests to support each step

The diagnostic process for AD can be divided into the following steps: detect, assess/differentiate, diagnose, and treat. It is important for clinicians to utilize appropriate tests when investigating a patient suspected of having AD in the early stages. Here, we highlight the most valuable tests for each step and which ones should be used in a primary care or specialist setting; *FDG-PET is usually considered after a diagnostic work-up; Abbreviations: A-IADL-Q, Amsterdam Instrumental Activities of Daily Living Questionnaire. Aβ, amyloid beta. Ach, acetylcholine. BG, blood glucose. CSF, cerebrospinal fluid. FAQ, Functional Activities Questionnaire. FAST, Functional Analysis Screening Tool. FDG-PET, fluorodeoxyglucose-PET. GDS, Geriatric Depression Scale. IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly. Mini-Cog, Mini Cognitive Assessment Instrument. MMSE, Mini-Mental State Examination. MoCA, Montreal Cognitive Assessment. MRI, magnetic resonance imaging. NMDA, N-Methyl-D-aspartic acid. NPI-Q, Neuropsychiatric Inventory Questionnaire. PCP, primary care physician. PET, positive emission tomography. p-tau, phosphorylated tau. QDRS, Quick Dementia Rating System. TSH, thyroid-stimulating hormone. t-tau, total tau


Step 1: Detect

The role of primary care in the early detection of AD

The insidious and variable emergence of symptoms associated with AD and other dementias can make recognition extremely challenging, particularly in a primary care setting (30, 35). Clinicians often have limited time with patients, so it is vital that they are able to quickly and accurately recognize the early signs and symptoms associated with AD (Table 2) (3, 30, 36), and training for nurses, NPs, and PAs who may have more time to observe patients should provide substantial benefits. Although extremely variable, initial symptoms may include short-term memory loss or psychological concerns, including depressive symptoms and a loss of purpose (36).
Patients, family members, and even HCPs themselves may present barriers to the diagnosis of early-stage AD. Patients may hide their symptoms or even avoid making an appointment until their symptoms significantly affect their day-to-day life due to fear of the stigma associated with a diagnosis of AD (19). Additionally, patients, family members, and PCPs/HCPs may dismiss or misinterpret symptoms as simply part of the normal aging process (30). Retrieving information from a trusted family member or informant/caregiver is essential when trying to assess a patient for suspected AD, as this perspective can provide a more objective understanding of the daily routine, mood, and behavior of the patient, and how this may have changed over time (30). For patients presenting with even subtle symptoms associated with AD, it is important that the PCP/HCP conducts an initial assessment to confirm the presence of symptoms using a validated assessment for early-stage AD detection (Figure 2; Step 2: Assess/Differentiate).

Case study: Presentation

A 63-year-old Caucasian male (J.K.) presented to his PCP with short-term memory loss over the last 2 years (Table 1A). Accompanied by his wife, he acknowledged his job had been affected by issues with his short-term memory; however, he considered his memory similar to that of his peers. His wife reported that people at work had started to notice him struggling to keep up, and also that family had to remind him of his upcoming appointments. He admitted to having intermittent depressive symptoms and anxiety, as well as irritability. Based on the patient’s symptoms, the PCP felt his presentation warranted further clinical assessment.

Table 1. Patient case study

Abbreviations: Aβ, amyloid beta. ApoE, apolipoprotein E. HgbA1c, hemoglobin A1c. MoCA, Montreal Cognitive Assessment. MRI, magnetic resonance imaging. PCP, primary care physician. p-tau, phosphorylated tau. t-tau, total tau

Table 2. Symptoms associated with suspected early stage Alzheimer’s disease


Step 2: Assess and differentiate

Primary care: Initial assessment when a patient presents

When a patient initially presents with symptoms consistent with early stages of AD, a clinician must first conduct a comprehensive clinical assessment to rule out other potential non-AD causes of cognitive impairment (Figure 2). PCPs are well placed to conduct these initial assessments, as they may not require specialist input or hospital tests. During the initial assessment, the primary objective of the clinician should be to exclude possible reversible causes of cognitive impairment, such as depression, or vitamin, hormone, and electrolyte deficiencies (37). The initial assessment should include a thorough history to identify potential risk factors associated with AD, including a family history of AD or related dementias in first-degree relatives (31, 38). Other known risk factors for AD that should be identified include age, female sex, ApoE ε4 status, physical inactivity, low education, diabetes, and obesity (3). It is also important to review for pre-existing medical conditions or prescribed medications that could be a cause of the patient’s cognitive impairment (36). Additionally, when conducting a thorough history, open-ended, probing questions should be directed to both the patient and the informant to ascertain how the patient’s cognition has changed over time and how the cognitive deficits affect their everyday activities; example questions for the initial assessment are detailed in Table 3 (30). Engaging with informants/caregivers is key to capturing additional information to help support all assessments. A routine differential diagnosis of AD begins with a detailed history, physical and neurologic examinations, and bloodwork analyses, followed by cognitive assessments and functional evaluation (Figure 2).

Table 3. Example questions for a clinician conducting an initial assessment with a patient and caregiver (30)


Primary care: Physical examination and blood analyses

A physical examination and blood tests can identify comorbid contributory medical conditions and reversible causes of cognitive impairment. A physical examination, including a mental status and neurological assessment, should be conducted to detect conditions such as depression and, for example, to look for signs such as issues with speaking or hearing as well as signs that could indicate a stroke (37). As part of the physical exam, a physician may ask the patient about diet and nutrition, review all medications (to see if these are the cause of any cognitive impairment, e.g. anti-cholinergics, analgesics, or sleep aids and anxiolytics), check blood pressure, temperature and pulse, and listen to the heart and lungs (36, 39).
Blood tests can rule out potentially treatable illnesses as a cause of cognitive impairment, such as vitamin B12 deficiency or thyroid disease (37). Suggested blood analyses include: 1) complete blood cell count; 2) blood glucose; 3) thyroid-stimulating hormone; 4) serum B12 and folate; 5) serum electrolytes; 6) liver function; and 7) renal function tests (30). Although not routinely used in clinical practice, clinicians may request ApoE genotyping, as this can help assess the genetic risk of developing AD. ApoE is the dominant cholesterol carrier within the brain that supports lipid transport and injury repair (40,41), and the APOE gene exists as three polymorphic alleles: APOE ε2, ε3, and ε4. The ε4 allele of ApoE is associated with increased AD risk, whereas the ε2 allele is protective (40,42). The number of ApoE ε4 alleles a person carries increases their risk of developing AD and the age of disease onset (43). Homozygous ε4 carriers (those with two copies of the ε4 allele) have the greatest risk of developing AD and the lowest average age of onset (43). In some practice settings, ApoE genotyping can only be conducted by a genetic counselor; a referral for more comprehensive genetic testing may be considered by the HCP if there is a family history of early-onset AD or dementia. Consumer tests are also becoming more readily available for patients wanting to determine their risk of developing diseases such as AD based on genetic risk factors (44).

Primary care: Cognitive, functional, and behavioral assessments

Cognitive assessments

If a patient is suspected of having AD following an initial assessment in primary care, and they are <65 years old, or if the case is complex, a referral to a dementia specialist such as a neurologist, geriatrician, or geriatric psychiatrist may be required for further evaluation. The specialist would then use an appropriate battery of cognitive, functional, and behavioral tests to assess the different aspects of disease, and ultimately to confirm diagnosis. However, not all patients with suspected cognitive deficits are immediately referred to a dementia specialist at this stage, which is only partly due to limited numbers of specialists (25) (Figure 2). In clinical practice, a two-stage process is often employed. This involves an initial ‘triage’ step conducted by non-specialists to clinically assess and select those patients who require further evaluation by a dementia specialist (45). During this ‘triage’ step, there are several clinical assessments available to non-specialists for assessing the presence of cognitive and functional impairments and behavioral symptoms (Table 4) (28, 35, 46–55)
Previous research has shown that clinicians have a tendency to choose one assessment over another due to their familiarity with the assessment, time constraints, or specific resources available to them within their clinic (30), but clinicians need to be aware of, and prepared to use, the most patient-appropriate assessments: the cultural, educational, and linguistic needs of the patient are important considerations (30,36,56–58). Some assessments have been translated into different languages or shortened, or have education-adjusted scoring classifications, where required (56–58).
Cognitive assessments that can be conducted quickly (<10 minutes), such as the Mini-Mental State Examination (MMSE) or Montreal Cognitive Assessment (MoCA), can be used by non-specialists to identify the presence and severity of cognitive impairment in patients before referring to a dementia specialist (Table 4) (36). Both the MMSE and MoCA are used globally in clinical practice, particularly in primary care, but vary in terms of their sensitivity to identify AD in the early stages (28,59). The MMSE is sensitive and reliable for identifying memory and language deficits in general but has limitations in identifying impairments in executive functioning (59). MoCA was originally developed to improve the detection of MCI (28) and is more sensitive than the MMSE in its assessment of memory, visuospatial, executive, and language function, and orientation to time and place (59). Both tests are relatively easy to administer and take around 10 minutes to complete. Neither assessment requires extensive training by the clinician, although MoCA users do need to undergo a 1-hour certification as mandated by the MoCA Clinic and Institute (28,60).
For time-constrained clinicians, the Mini Cognitive Assessment Instrument (Mini-Cog) may be an appropriate tool to assess cognitive deficits that focus on memory, and components of visuospatial and executive function (Table 4). The assessment includes the individual learning three items from a list, drawing a clock, and then recalling the three-item list. The Mini-Cog can be useful for clinicians in primary care, as it requires no training and the results are easy to interpret. As an alternative to these tests, PCPs might also consider using an informant-based structured questionnaire such as the AD8 or Informant Questionnaire on Cognitive Decline in the Elderly to help guide discussions with the patient and caregiver (Table 4) (28).

Table 4. Cognitive, functional, and behavioral assessments to support the diagnosis of Alzheimer’s disease in a primary care and specialist setting

*Personal communication; Abbreviations: AD, Alzheimer’s disease. A-IADL-Q, Amsterdam Instrumental Activities of Daily Living Questionnaire. FAQ, Functional Activities Questionnaire. FAST, Functional Assessment Screening Tool. GDS, Geriatric Depression Scale. IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly. MCI, mild cognitive impairment. Mini-Cog, Mini Cognitive Assessment Instrument. MMSE, Mini-Mental State Examination. MoCA, Montreal Cognitive Assessment. NPI-Q, Neuropsychiatric Inventory Questionnaire. QDRS, Quick Dementia Rating System


Functional assessments

Functional assessments are valuable in identifying changes in a patient’s day-to-day functioning through the evaluation of their instrumental activities of daily living (IADLs). IADLs are complex activities that are necessary for the individual to function independently (e.g., cooking, shopping, and managing finances) and can be impaired during the early stages of cognitive impairment. While it is possible that functional decline may occur as a part of normal aging, a decline in a person’s IADL performance is strongly associated with neurodegenerative diseases such as AD (61). In the early stages of AD, patients may be functionally independent, and any impairment in IADLs may be subtle, such as difficulties paying bills or driving to new places. A patient’s functional independence is essential for their well-being and mental health (62), particularly in the early stages of the disease when the individual may still be working and socializing relatively independently (3). Consequently, functional independence is one of the most important clinical features for patients with AD. As the disease progresses, and patients have increasing functional impairment, this significantly impacts on their independence, and subsequently their and their family/caregiver’s quality of life.
Functional assessment is, therefore, an integral part of the diagnostic process for AD. The Functional Activities Questionnaire (FAQ) is an informant questionnaire that assesses the patient’s performance over a 4-week period and may take only a few minutes to complete (Table 4). The questionnaire is scored from ‘normal’ to ‘dependent’, using numerical values assigned to categories, with higher scores indicative of increasing impairment (47). Previous research has shown that the FAQ has high sensitivity and reliability for detecting mild functional impairment in patients with MCI (47).
Determining an individual’s functional independence can be challenging and the clinician may require additional input from an informant to determine a patient’s functional decline and their ongoing ability to conduct activities of daily living (37). The clinician can gain greater insight through the informant into the patient’s day-to-day life and any issues the patient is having at home. This type of information is vital to the clinician, and when combined with other assessment tools, can help to narrow the differential diagnosis.

Behavioral assessments

Patients with suspected AD may experience several behavioral symptoms such as anxiety, disinhibition, apathy, and depression (Table 2). In the early stages of disease, such symptoms are generally associated with poor long-term outcomes and caregiver burden, and are particularly distressing to both patients and their families (63). It is important for clinicians to use appropriate assessments to identify behavioral and psychiatric symptoms that are caused by neurodegenerative diseases, such as AD, rather than by alternative causes, such as a mood disorder.
The Geriatric Depression Scale (GDS) and Neuropsychiatric Inventory Questionnaire (NPI-Q) can be used by clinicians to assess neuropsychiatric symptoms in patients for whom early-stage AD is suspected (Table 4). The GDS is a 15-item (or longer 30-item) questionnaire that assesses mood, has good reliability in older populations for detecting depression, and can be completed by the patient within 5–10 minutes (63). The NPI-Q can be used in conjunction with or as an alternative to the GDS. The NPI-Q is completed by a knowledgeable informant or caregiver who can report on the patient’s neuropsychiatric symptoms. The NPI-Q can be conducted in around 5 minutes to determine both the presence and severity of symptoms across several neuropsychiatric domains including depression, apathy, irritability, and disinhibition (49). Consequently, as it assesses depression, it can be used as an alternative to GDS if time constraints do not allow for both to be completed.
Behavioral symptoms can be non-specific, so it is important for clinicians to consider and rule out other potentially treatable causes of impairment when assessing this domain. For example, depression is associated with concentration and memory issues (64); apathy can occur in non-depressed elderly individuals and can impact cognitive function (65). Signs/symptoms such as social withdrawal, feelings of helplessness, or loss of purpose should be investigated closely, as these could be indicative of depression alone. It is important for clinicians to recognize that if changes over time in cognitive symptoms and mood symptoms match, then depression is most likely to be the root cause of subtle cognitive decline, rather than AD (28).

Primary care clinician checklist

If AD is still suspected following clinical assessment, referral to a specialist for further diagnostic testing, including imaging and fluid biomarkers, may be required. It is important the clinician confirms the following checks/assessments before the patient undergoes further evaluation:

Primary care clinician checklist

• Confirm medical and family history
• Review the patient’s medications for any that could cause cognitive impairment
• Perform blood tests to eliminate potential reversible causes of cognitive impairment
• Conduct a quick clinical assessment to confirm the presence of cognitive impairment

Specialist role in assessment

Following the initial assessment in primary care, further cognitive, behavioral, functional, and imaging assessments can be carried out in a specialist setting. With their additional AD experience, access to other specialties, and possibly fewer time constraints than the PCP, the specialist is able to conduct a more comprehensive testing battery, using additional clinical assessments and biomarkers to determine causes of impairment and confirm diagnosis (Figure 2).

Cognitive assessments

Because the cognitive impacts of early-stage AD may vary from patient to patient, it is important to consider which cognitive domains are affected in these early stages when considering which assessments to use. Specialists are able to conduct a full neuropsychological test battery that covers the major cognitive domains (executive function, social cognition/emotions, language, attention/concentration, visuospatial and motor function, learning and memory); preferably, a battery should contain more than one test per domain to ensure adequate sensitivity in capturing cognitive impairment (66). This step can help with obtaining an in-depth understanding of the subtle changes in cognition seen in the early stages of AD and enables the clinician to monitor subsequent changes over time.
Typically, episodic memory, executive function, visuospatial function, and language are the most affected cognitive domains in the early stages of AD (29,67,68). Currently, most cognitive assessment tools focus on a subset of the overall dimensions of cognition; it is therefore vital the clinician chooses the correct test to assess impairment in these specific cognitive domains that could be indicative of AD in the early stages. As cognitive impairment in the early stages of AD can be subtle and vary significantly between individuals (29), clinicians must choose appropriate, sensitive tests that can detect these changes and account for a patient’s level of activity and cognitive reserve (29). If there is large disparity in results across cognitive assessments, it is important for the clinician to shape their assessments based on the patient’s history. If the patient’s history is positive for neurodegenerative disease, but one assessment does not reflect this, it is important to conduct further tests to ascertain the cause of the cognitive impairment.
The Quick Dementia Rating System (QDRS) can be used by specialists to assess cognitive impairment (Table 4). This short questionnaire (<5 minutes) is completed by a caregiver/informant and requires no training. The QDRS assesses several cognitive domains known to be affected by AD, including memory, language and communication abilities, and attention. The questionnaire can reliably discriminate between individuals with and without cognitive impairment and provides accurate staging for disease severity (28).

Functional assessments

The Amsterdam IADL Questionnaire (A-IADL-Q) and Functional Assessment Screening Tool (FAST) can both be used to assess a patient’s functional ability (Table 4) (53). The A-IADL-Q is a reliable computerized questionnaire that monitors a patient’s cognition, memory, and executive functioning over time. This questionnaire is completed by an informant of the patient and takes 10 minutes to complete (53). For patients with suspected early stage AD, the A-IADL-Q is a useful tool to monitor subtle changes in IADL independence over time and is less influenced by education, gender, and age than other functional assessments (53). The FAST is a useful assessment for clinicians to identify the occurrence of functional and behavioral problems in patients with suspected AD. The questionnaire is completed by informants who interact with the patient regularly; informants are required to answer Yes/No to a number of questions focusing on social and non-social scenarios (55).

Structural imaging

Structural imaging, such as MRI, provides clinically useful information when investigating causes of cognitive impairment (69) (Figure 2). MRI is routinely conducted to exclude alternative causes of cognitive impairment, rather than support a diagnosis of AD (37,70). It is well known that medial temporal lobe atrophy is the best MRI marker for identifying patients in the earliest stages of AD (70,71); however, specific patterns of atrophy may also be indicative of other neurodegenerative diseases. Atrophy alone is rarely sufficient to make a diagnosis. MRI findings can help to narrow the differential diagnosis, and the results should be considered in the context of the patient’s age and clinical examination (69–71).
Clinicians are advised to take a stepwise approach when reviewing structural imaging reports of a patient with suspected AD. These steps include: 1) excluding brain pathology that may be amenable to surgical intervention (e.g., the scan will show regions of hyper- or hypointensity rather than a uniform signal); 2) assessing for brain microbleeds (e.g., looking at signal changes within different areas of the brain can identify vascular comorbidities); and 3) assessing atrophy (e.g., medial temporal lobe atrophy is characteristic of AD) (69). Radiologists can conduct a quick and easy visual rating of any medial temporal lobe atrophy; these results can then be utilized by the specialist, in conjunction with a clinical assessment, to determine the likely cause of cognitive impairment. If the clinician is unable to determine a differential diagnosis, additional confirmatory tests can be requested.
Fluorodeoxyglucose-PET (FDG-PET) is a useful structural imaging biomarker that can support an early and differential diagnosis (72); however, specialists usually prefer to use this after their initial diagnostic work-up. As the brain relies almost exclusively on glucose as its source of energy, FDG (a glucose analog) can be combined with PET to identify regional patterns of reduced brain metabolism and neurodegeneration (70,72). FDG-PET is not recommended for diagnosing patients with preclinical AD, as there is no way to ascertain whether the hypometabolism is directly related to AD pathology (73); however, clinicians may refer patients with more established symptomatology for an FDG-PET scan to identify regions of glucose hypometabolism and neurodegeneration that could be indicative of AD (70).

Case study: Assess/differentiate

The initial assessment by the primary care clinician revealed that J.K.’s medical history was significant for hypertension, dyslipidemia, mild obesity, and glucose intolerance (Table 1B). There was no history of cerebrovascular events, significant head injuries, or focal findings on the neurologic exam. Besides the vascular risk factors, no medical conditions or current medications were found to be likely contributors to the cognitive deficit. The patient had a positive family history of dementia, where the onset typically occurred in the late 60s. Genotyping showed the patient to be a homozygous carrier of two ApoE ε4 alleles. Blood tests revealed elevated serum glucose and C-reactive protein but were otherwise normal. The patient had an unremarkable mental status examination, and his MoCA score was 21/30, with points lost on orientation, recall, and naming (Table 1C).
The patient was referred to a memory clinic for further assessment. The dementia specialist referred the patient for an MRI that predominantly showed mild small vessel disease and mild generalized atrophy with a significant reduction in hippocampal volume and ratio. Based on his medical and family history, cognitive assessments, and structural imaging results, the specialist deemed the severity of cognitive impairment to be in the mild range; consequently, the specialist referred the patient for biomarker assessment to determine the underlying cause.

Step 3: Diagnose

Historically, AD was only diagnosed postmortem until we developed the ability to ascertain the underlying pathology associated with the disease in new ways, namely imaging and fluid biomarkers. However, despite supportive results from single- and multicenter trials, the use and reimbursement of imaging and fluid biomarkers to support the diagnosis of AD still vary considerably between countries (70).

Imaging biomarkers

Recent advances have allowed physicians to visualize the proteins associated with AD, namely Aβ and tau, via PET scanning. Amyloid PET is currently the only imaging approach recommended by the Alzheimer’s Association and the Amyloid Imaging Task Force to support the diagnosis of AD (70). Amyloid PET utilizes tracers (florbetapir, flutemetamol, and florbetaben) that specifically bind to Aβ within amyloid plaques; a positive amyloid PET scan will show increased cortical retention of the tracer in regions of Aβ deposition within the brain (74), thus confirming the presence of Aβ plaques in the brain (74,75) and directly quantify brain amyloid pathology (76), thus making it a useful tool to supplement a clinical battery to diagnose AD (3,74). However, a positive amyloid PET scan alone does not definitively diagnose clinical AD, and these results must be combined with other clinical assessments, such as cognitive assessment, for an accurate diagnosis (74). It is also important to note that amyloid PET is expensive and not readily reimbursed by health insurance providers (70); if it is not possible to access amyloid PET, biomarker confirmation can be assessed using CSF.

Fluid biomarkers

An additional or alternative tool to amyloid PET is the collection and analysis of CSF for the presence of biomarkers associated with AD pathology. Patients who have symptoms suggestive of AD can be referred for a lumbar puncture to analyze their CSF for specific AD-associated biomarkers (3). CSF biomarkers are measures of the concentrations of proteins in CSF from the lumbar sac that reflect the rates of both protein production and clearance at a given timepoint (7). Lumbar punctures can be conducted safely and routinely in an outpatient setting or memory clinic (77). However, many patients still worry about the pain and possible side effects associated with the procedure and may require additional information and support from the clinician to undertake the procedure (77). Appropriate use criteria are available for HCPs to help identify suitable patients for lumbar puncture and CSF testing (78). For example, individuals presenting with persistent, progressing, and unexplained MCI, or those with symptoms suggestive of possible AD, should be referred for lumbar puncture and CSF testing (78). However, lumbar puncture and CSF testing are not recommended for determining disease severity in patients who have already received a diagnosis of AD or in lieu of genotyping for suspected autosomal dominant mutation carriers (78).
Because there is strong concordance between CSF biomarkers and amyloid PET, either can be used to confirm Aβ burden (79). As such, CSF biomarkers are widely accepted within the AD community to support a diagnosis (80). AD biomarkers from the brain can be detected in CSF well before the onset of overt clinical symptoms in early-stage AD (6,7). Core AD CSF biomarkers, such as Aβ42 (one of two main isoforms of Aβ and a major constituent of Aβ plaques) and phosphorylated tau (p-tau) and total tau (t-tau), can be measured to determine the presence of disease (80).
When interpreting CSF analyses for a patient with suspected AD, it is important to remember that AD is associated with decreased CSF Aβ42 and increased tau isoforms (32). Decreased CSF Aβ42 levels are a reflection of increased Aβ aggregation and deposition within the brain (32), and the concentration of CSF Aβ42 directly relates to the patient’s amyloid status (e.g., the presence or absence of significant amyloid pathology) and the total amount of Aβ peptides (e.g., Aβ42 and Aβ40) (32). Specialists’ use of ratios of these CSF biomarkers (e.g., Aβ42/40) rather than single CSF biomarkers alone has been shown to adjust for potential differences in Aβ production and provide a better index of the patient’s underlying amyloid-related pathology (81). The increase in CSF p-tau and t-tau associated with AD may directly reflect the aggregation of tau within the brain and neurodegeneration, respectively (32). P-tau in CSF provides a direct measure of the amount of hyperphosphorylated tau in the brain, which is strongly suggestive of the presence of NFTs, whereas CSF t-tau can predict the level of neurodegeneration in a patient with suspected AD; however, t-tau is also increased in other neurologic conditions (32).
Ultimately, the clinical decision to use amyloid PET or CSF to confirm amyloid and tau pathology can be affected by several practical factors (Table 5) (70,77,80,82–85).

Table 5. Comparison of key CSF and amyloid PET considerations for amyloid confirmation

Abbreviations: CSF, cerebrospinal fluid. PET, positron emission tomography


Emerging diagnostic tools

Access constraints for amyloid PET have driven the need for alternative sensitive and specific CSF and blood-based biomarkers that can detect AD-associated pathology in the early stages (86). Significant efforts have been undertaken over the last decade to identify blood-based biomarkers to: 1) detect AD pathology; 2) identify those at risk of developing AD in the future; and 3) monitor disease progression (33,34,87). At present, only a limited number of approved blood-based assays are available to clinicians to detect AD pathology (88); however, several novel assays are currently under investigation, including those measuring various phosphorylated forms of tau, including p-tau181 and p-tau217 (89). Investigational use of plasma p-tau181 (an isoform of tau) has been shown to differentiate AD from other neurodegenerative diseases and predict cognitive decline in patients with AD (33). CSF p-tau217 (a different isoform of tau) is a promising biomarker under investigation for detecting preclinical and advanced AD (86,90). Given that blood testing is already a well-established part of clinical routines globally and can easily be performed in a variety of clinical settings, blood-based biomarkers could in future serve as the potential first step of a multistage diagnostic process. This would be a benefit to clinicians, particularly those in primary care, by helping to identify individuals requiring a referral to a specialist for diagnostic testing (87).

Case study: Diagnose

J.K. underwent a lumbar puncture for CSF analysis, which showed decreased Aβ42 and increased p-tau and t-tau protein (Table 1D). Based on the results from the genotyping, cognitive assessments, MRI, and CSF biomarkers, the clinician confirmed that the likely cause of the patient’s cognitive deficits was early-stage AD, especially in view of a positive family history of dementia with similar age of onset.

Step 4: Treat

The role of the clinician following a diagnosis of early-stage AD is to discuss the available management and treatment options while providing emotional and practical support to the patient, caregiver, and family where appropriate (37). Clinicians can also refer the patient and their caregiver(s) to social services for further support, as well as help connect them with reliable sources of information and even local research opportunities and clinical trials.
One important role for a clinician treating a patient diagnosed with early-stage AD is to closely monitor the patient’s disease progression through regular follow-up appointments (e.g., every 6–12 months); clinicians should encourage patients (and the caregiver) to make additional follow-up appointments, especially should symptoms worsen. Routine cognitive and functional assessments (Table 4) should be used to monitor disease progression; these tools can be used to identify unexpected trends, such as rapid decline, which could prompt the need for additional medical evaluation such as blood tests, imaging, or biomarker analyses. Results from such tests could help guide management and/or treatment decisions over the course of the patient’s disease.
Non-pharmacologic therapies (e.g., diet and exercise) may be employed for patients with early AD, with the goal to maintain or even improve cognitive function and retain their ability to perform activities of daily living. For patients in the early stages of disease, dietary changes (e.g., following a healthy diet high in green, leafy vegetables, fish, nuts, and berries), physical exercise, and cognitive training have demonstrated small but significant improvements in cognition (36,91). Non-pharmacologic therapies can have a positive impact on quality of life and are generally safe and inexpensive (36); however, compliance with these non-pharmacologic therapies should be monitored by the clinician. Research suggests that multimodal therapies, such as cognitive stimulation therapy, may also be more effective when used in combination with pharmacologic treatments (91).
Several pharmacologic treatments have received regulatory approval to treat the symptoms of mild to severe AD dementia. Acetylcholinesterase inhibitors (donepezil, rivastigmine, and galantamine) and N-methyl-D-aspartate receptor antagonists (memantine) can be prescribed to patients to temporarily ameliorate the symptoms of AD dementia such as cognitive and functional decline (92–96). Meta-analyses of donepezil, rivastigmine, and galantamine have shown that patients with mild-to-moderate AD dementia experience some benefits in cognitive function, activities of daily living, and clinician-rated global clinical state (93,94,97). Furthermore, treatment with acetylcholinesterase inhibitors and/or memantine has also been shown to modestly improve measures of global function and temporarily stabilize measures of activities of daily living (96). However, it is important to note that these drugs provide only temporary, symptomatic benefit and that not all patients respond to treatment (36,98). Critically, none of the current drugs available address the underlying pathophysiology or alter the ultimate disease course.
Following AD diagnosis, a comprehensive approach toward clinical care can be individualized based on the patient’s specific AD risk factors (20,21). Clinicians should consider managing uncontrolled vascular risk factors (e.g., hypertension, hyperlipidemia, diabetes) with antithrombotics, antihypertensives, lipid-lowering, and/or antidiabetic agents, respectively, to reduce the risk of cerebrovascular ischemia and stroke, and subsequent cognitive decline (36,99). They should also consider the management of the patient’s behavioral symptoms. For most patients in the early stages of disease, behavioral symptoms will be relatively mild, and no pharmacologic management is required; however, pharmacologic treatment, such as a low-dose selective serotonin reuptake inhibitor, can be prescribed for patients with AD-associated depression and anxiety (100,101).

Specialist clinician checklist

The specialist’s role is critical to further evaluating the initial checks/assessments, providing the diagnosis, and developing the individualized patient management plan:
• Identify deficits to specific cognitive domains using appropriate tests
• Confirm functional performance, using patient and caregiver assessments
• Perform structural imaging to complete assessment of the patient
• Confirm diagnosis with imaging or fluid biomarkers
• Develop a personalized management and follow-up plan
• Direct the patient to additional support resources such as the Alzheimer’s Association

Case study: Treat

Following diagnosis, J.K. was advised on the available management options and research opportunities (Table 1E). The specialist emphasized the need to control his vascular risk factors and suggested lifestyle modifications to optimize the management of his other medical problems. The patient’s neuropsychiatric symptoms were considered mild and did not require pharmacologic intervention. The patient was also provided with details for a local social worker and directed toward further disease-specific information from the Alzheimer’s Association related to his disease. The patient was encouraged to return for additional follow-up visits so that his disease and associated symptoms could be appropriately monitored and managed.


Future perspectives

An early diagnosis of AD will become increasingly important as treatments that alter the underlying disease pathology become available—particularly given the expectation that such treatments will be more effective in preserving cognitive function, and thus prolonging independence, when given early in the course of the disease (19). The approval of such treatments will likely lead to an increased awareness of cognitive impairment and other AD-associated symptoms among both the public and non-specialists, such as those in primary care settings. This may encourage more patients/family members to seek help at an earlier stage of disease than is currently seen in community practice. Increased use of sensitive screening measures to proactively assess for the presence of AD symptoms will help identify patients suspected of having early AD. Assessment of cognitive impairment during a Medicare Annual Wellness Visit is inconsistent; the U.S. Preventative Services Task Force, whilst recognizing the importance of MCI, has maintained its decision that there is insufficient evidence to support the mandate of cognitive screening. However, sensitive screening procedures, along with the availability of disease-modifying treatments, are likely to change their recommendations. There is also a need for a mandated, standardized screening approach internationally. Together, this will result in an increase in patients requiring diagnosis, increasing the demand for specialists to evaluate and diagnose, the need for amyloid confirmation, and wait times for patients, which will collectively put further pressure on an already-stretched healthcare infrastructure (25).
Nevertheless, efforts continue within the AD field to streamline the diagnostic process. Planning for and implementing change will not only improve patient management now but also help prepare healthcare systems for an approved disease-modifying treatment for AD. A flexible, multidisciplinary team approach is recommended to integrate the care needed to detect, assess, differentiate, diagnose, treat, and monitor a diverse AD population (24). The development of tests that could be carried out routinely in a primary care setting, such as blood-based AD biomarkers, would help PCPs and non-specialists identify which patients may need further evaluation or referral to a specialist (25). Interest also remains high in advancing imaging techniques, such as amyloid and tau PET, to support a diagnosis of AD. Although amyloid and tau PET are not currently readily available, they may be useful for specialists in the future to determine disease staging or track progression, or as a surrogate marker of cognitive status (74). The introduction of new screening and diagnostic tools could ultimately help lower the burden on specialists and ensure patients are diagnosed in a timely manner.



Consensus within the AD community has recently shifted to encourage the diagnosis of AD as early as possible. This shift will enable patients to plan their future and consider symptomatic therapies and lifestyle changes that could reduce cognitive deficits and ultimately help preserve their quality of life. Promisingly, new, potentially disease-modifying therapeutic candidates are on the horizon that could be effective in early AD by targeting and ameliorating the underlying biological mechanisms (92,102). This paper has outlined a menu of practical tools for clinicians to use in the real world to support an early diagnosis of AD and how they may best be incorporated into current clinical practice. Ultimately, a coordinated, multidisciplinary approach that encompasses primary care and specialist expertise is required to ensure timely detection, assessment and differentiation, diagnosis, and management of patients with AD.


Authors’ contributions: All authors participated in the review of the literature and in the drafting and reviewing of the manuscript. All authors read and approved the final version of the manuscript for submission.

Funding: The authors developed this manuscript concept during an assessment of Alzheimer’s disease educational needs. The development of this manuscript was funded by Biogen. Editorial support was provided by Jodie Penney, MSc, PhD, Helios Medical Communications, Cheshire, UK, which was funded by Biogen.

Acknowledgements: The authors would like to acknowledge and thank Dr. Giovanni Frisoni, Geneva University Neurocenter, for his contribution towards the development of this manuscript.

Conflict of Interest: AP reports personal fees from Acadia Pharmaceuticals, Alzheon, Avanir, Biogen, Cadent Therapeutics, Eisai, Functional Neuromodulation, MapLight Therapeutics, Premier Healthcare Solutions, Sunovion, and Syneos; grants from Alector, Athira, Avanir, Biogen, Biohaven, Eisai, Eli Lilly, Genentech/Roche, and Vaccinex. RI has nothing to disclose. MS reports personal fees from Alzheon, Athira, Biogen, Cortexyme, Danone, Neurotrope, Regeneron, Roche-Genentech, and Stage 2 Innovations; stock options from Brain Health Inc, NeuroReserve, NeuroTau, Neurotrope, Optimal Cognitive Health Company, uMethod Health, and Versanum Inc. Additionally, he has intellectual property rights with Harper Collins. SK and IR report employment with Biogen.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Hughes J. This is one of the biggest global health crises of the 21st century. World Econ Forum 2017.’s%20Day%2C%20the,crises%20of%20the%2021st%20century (accessed May 18, 2020).
2. Alzheimer’s Disease International. World Alzheimer Report 2019. Attitudes to dementia 2019. (accessed February 13, 2020).
3. Alzheimer’s Association. 2020 Alzheimer’s disease facts and figures. Alzheimers Dement 2020;16:391–460.
4. Deb A, Thornton JD, Sambamoorthi U, Innes K. Direct and indirect cost of managing alzheimer’s disease and related dementias in the United States. Expert Rev Pharmacoecon Outcomes Res 2017;17:189–202.
5. Serrano-Pozo A, Frosch MP, Masliah E, Hyman BT. Neuropathological alterations in Alzheimer disease. Cold Spring Harb Perspect Med 2011;1:a006189.
6. Bateman RJ, Xiong C, Benzinger TL, et al. Clinical and biomarker changes in dominantly inherited Alzheimer’s disease. N Engl J Med 2012;367:795–804.
7. Jack CR, Bennett DA, Blennow K, et al. NIA-AA Research Framework: toward a biological definition of Alzheimer’s disease. Alzheimers Dement 2018;14:535–62.
8. Dubois B, Feldman HH, Jacova C, et al. Revising the definition of Alzheimer’s disease: a new lexicon. Lancet Neurol 2010;9:1118–27.
9. U.S. Food and Drug Administration (FDA). Early Alzheimer’s disease: developing drugs for treatment guidance for industry 2018. (accessed August 11, 2020).
10. Insel PS, Weiner M, Mackin RS, et al. Determining clinically meaningful decline in preclinical Alzheimer disease. Neurology 2019;93:e322–33.
11. Vermunt L, Sikkes SA, van den Hout A, et al. Duration of preclinical, prodromal, and dementia stages of Alzheimer’s disease in relation to age, sex, and APOE genotype. Alzheimers Dement 2019;15:888–98.
12. Cho SH, Woo S, Kim C, et al. Disease progression modelling from preclinical Alzheimer’s disease (AD) to AD dementia. Sci Rep 2021;11:4168.
13. Knopman DS, Parisi JE, Salviati A, et al. Neuropathology of cognitively normal elderly. J Neuropathol Exp Neurol 2003;62:1087–95.
14. Bennett DA, Schneider JA, Arvanitakis Z, et al. Neuropathology of older persons without cognitive impairment from two community-based studies. Neurology 2006;66:1837–44.
15. Kazim SF, Iqbal K. Neurotrophic factor small-molecule mimetics mediated neuroregeneration and synaptic repair: emerging therapeutic modality for Alzheimer’s disease. Mol Neurodegener 2016;11:50.
16. Tolbert S, Liu Y, Hellegers C, et al. Financial management skills in aging, MCI and dementia: cross sectional relationship to 18F-florbetapir PET cortical β-amyloid deposition. J Prev Alzheimers Dis 2019;6:274–82.
17. Ye BS, Kim HJ, Kim YJ, et al. Longitudinal outcomes of amyloid positive versus negative amnestic mild cognitive impairments: a three-year longitudinal study. Sci Rep 2018;8:5557.
18. Roberts RO, Aakre JA, Kremers WK, et al. Prevalence and outcomes of amyloid positivity among persons without dementia in a longitudinal, population-based setting. JAMA Neurol 2018;75:970–9.
19. Dubois B, Padovani A, Scheltens P, Rossi A, Dell’Agnello G. Timely diagnosis for Alzheimer’s disease: a literature review on benefits and challenges. J Alzheimers Dis 2016;49:617–31.
20. Isaacson RS, Ganzer CA, Hristov H, et al. The clinical practice of risk reduction for Alzheimer’s disease: a precision medicine approach. Alzheimers Dement 2018;14:1663–73.
21. Isaacson RS, Hristov H, Saif N, et al. Individualized clinical management of patients at risk for Alzheimer’s dementia. Alzheimers Dement 2019;15:1588–602.
22. Gauthier SG. Alzheimer’s disease: the benefits of early treatment. Eur J Neurol 2005;12:11–6.
23. Cummings J, Lee G, Ritter A, Sabbagh M, Zhong K. Alzheimer’s disease drug development pipeline: 2020. Alzheimers Dement NY 2020;6:e12050.
24. Galvin JE, Aisen P, Langbaum JB, et al. Early stages of Alzheimer’s disease: evolving the care team for optimal patient management. Front Neurol 2021;11:592302.
25. Liu JL, Hlavka JP, Hillestad R, Mattke S. Assessing the preparedness of the U.S. Health Care System infrastructure for an Alzheimer’s treatment 2017. (accessed May 5, 2018).
26. Sabbagh MN, Lue L-F, Fayard D, Shi J. Increasing precision of clinical diagnosis of Alzheimer’s disease using a combined algorithm incorporating clinical and novel biomarker data. Neurol Ther 2017;6:83–95.
27. Balasa M, Gelpi E, Antonell A, et al. Clinical features and APOE genotype of pathologically proven early-onset Alzheimer disease. Neurology 2011;76:1720–5.
28. Galvin JE. Using informant and performance screening methods to detect mild cognitive impairment and dementia. Curr Geriatr Rep 2018;7:19–25.
29. Sabbagh MN, Boada M, Borson S, et al. Early detection of mild cognitive impairment (MCI) in primary care. J Prev Alzheimers Dis 2020;7:165–70.
30. Galvin JE, Sadowsky CH, NINCDS-ADRDA. Practical guidelines for the recognition and diagnosis of dementia. J Am Board Fam Med 2012;25:367–82.
31. Aisen PS, Cummings J, Jack CR, et al. On the path to 2025: understanding the Alzheimer’s disease continuum. Alzheimers Res Ther 2017;9:60.
32. Blennow K, Zetterberg H. Biomarkers for Alzheimer disease – current status and prospects for the future. J Intern Med 2018;284:643–63.
33. Karikari TK, Pascoal TA, Ashton NJ, et al. Blood phosphorylated tau 181 as a biomarker for Alzheimer’s disease: a diagnostic performance and prediction modelling study using data from four prospective cohorts. Lancet Neurol 2020;19:422–33.
34. Janelidze S, Mattsson N, Palmqvist S, et al. Plasma P-tau181 in Alzheimer’s disease: relationship to other biomarkers, differential diagnosis, neuropathology and longitudinal progression to Alzheimer’s dementia. Nat Med 2020;26:379–86.
35. Iliffe S, Robinson L, Brayne C, et al. Primary care and dementia: 1. diagnosis, screening and disclosure. Int J Geriatr Psychiatry 2009;24:895–901.
36. Arvanitakis Z, Shah RC, Bennett DA. Diagnosis and management of dementia: review. JAMA 2019;322:1589–99.
37. Robinson L, Tang E, Taylor J-P. Dementia: timely diagnosis and early intervention. BMJ 2015;350:h3029.
38. Zucchella C, Bartolo M, Pasotti C, Chiapella L, Sinforiani E. Caregiver burden and coping in early-stage Alzheimer disease. Alzheimer Dis Assoc Disord 2012;26:55–60.
39. Pfistermeister B, Tümena T, Gaßmann K-G, Maas R, Fromm MF. Anticholinergic burden and cognitive function in a large German cohort of hospitalized geriatric patients. PLoS One 2017;12:e0171353.
40. Liu C-C, Liu C-C, Kanekiyo T, Xu H, Bu G. Apolipoprotein E and Alzheimer disease: risk, mechanisms, and therapy. Nat Rev Neurol 2013;9:106–18.
41. Sadigh-Eteghad S, Sabermarouf B, Majdi A, Talebi M, Farhoudi M, Mahmoudi J. Amyloid-beta: a crucial factor in Alzheimer’s disease. Med Princ Pract 2015;24:1–10.
42. Karch CM, Goate AM. Alzheimer’s disease risk genes and mechanisms of disease pathogenesis. Biol Psychiatry 2015;77:43–51.
43. Ungar L, Altmann A, Greicius MD. Apolipoprotein E, gender, and Alzheimer’s disease: an overlooked, but potent and promising interaction. Brain Imaging Behav 2014;8:262–73.
44. 23andMe. What to know about health report test result – 23andMe. 23&Me® 2020. (accessed October 15, 2020).
45. Hendry K, Green C, McShane R, et al. AD-8 for detection of dementia across a variety of healthcare settings. Cochrane Database Syst Rev 2019;3:CD011121.
46. Patnode CD, Perdue LA, Rossom RC, et al. Screening for cognitive impairment in older adults: an evidence update for the U.S. Preventive Services Task Force. Rockville (MD): Agency for Healthcare Research and Quality (US); Report No.: 19-05257-EF-1, 2020.
47. Teng E, Becker BW, Woo E, Cummings JL, Lu PH. Subtle deficits in instrumental activities of daily living in subtypes of mild cognitive impairment. Dement Geriatr Cogn Disord 2010;30:189–97.
48. Marshall GA, Amariglio RE, Sperling RA, Rentz DM. Activities of daily living: where do they fit in the diagnosis of Alzheimer’s disease? Neurodegener Dis Manag 2012;2:483–91.
49. Rosenberg PB, Mielke MM, Appleby BS, Oh ES, Geda YE, Lyketsos CG. The association of neuropsychiatric symptoms in MCI with incident dementia and Alzheimer disease. Am J Geriatr Psychiatry 2013;21:685–95.
50. Ismail Z, Emeremni CA, Houck PR, et al. A comparison of the E-BEHAVE-AD, NBRS and NPI in quantifying clinical improvement in the treatment of agitation and psychosis associated with dementia. Am J Geriatr Psychiatry 2013;21:78–87.
51. Bowden VM, Bowden CL. The Journal of Neuropsychiatry and Clinical Neurosciences. JAMA 1992;268:1473–4.
52. Galvin JE. The quick dementia rating system (QDRS): a rapid dementia staging tool. Alzheimers Dement Amst 2015;1:249–59.
53. Koster N, Knol DL, Uitdehaag BM, Scheltens P, Sikkes SA. The sensitivity to change over time of the Amsterdam IADL Questionnaire(©). Alzheimers Dement 2015;11:1231–40.
54. Sikkes SA, Pijnenburg YA, Knol DL, de Lange-de Klerk ES, Scheltens P, Uitdehaag BM. Assessment of instrumental activities of daily living in dementia: diagnostic value of the Amsterdam Instrumental Activities of Daily Living Questionnaire. J Geriatr Psychiatry Neurol 2013;26:244–50.
55. LaRue RH. Functional Assessment Screening Tool (FAST). In: Volkmar FR, editor. Encycl. Autism Spectr. Disord., New York, NY: Springer New York, 2018:1–2.
56. Lindgren N, Rinne JO, Palviainen T, Kaprio J, Vuoksimaa E. Prevalence and correlates of dementia and mild cognitive impairment classified with different versions of the modified Telephone Interview for Cognitive Status (TICS-m). Int J Geriatr Psychiatry 2019;34:1883–91.
57. Wang H, Fan Z, Shi C, et al. Consensus statement on the neurocognitive outcomes for early detection of mild cognitive impairment and Alzheimer dementia from the Chinese Neuropsychological Normative (CN-NORM) Project. J Glob Health 2019;9:020320.
58. Franzen S, van den Berg E, Goudsmit M, et al. A systematic review of neuropsychological tests for the assessment of dementia in non-western, low-educated or illiterate populations. J Int Neuropsychol Soc 2020;26:331–51.
59. Costa A, Bak T, Caffarra P, et al. The need for harmonisation and innovation of neuropsychological assessment in neurodegenerative dementias in Europe: consensus document of the Joint Program for Neurodegenerative Diseases Working Group. Alzheimers Res Ther 2017;9:27.
60. MoCA Test Inc. Upcoming mandatory training for MoCA testing. MoCA Montr – Cogn Assess 2021. (accessed February 3, 2021).
61. Tabira T, Hotta M, Murata M, et al. Age-related changes in instrumental and basic activities of daily living impairment in older adults with very mild Alzheimer’s disease. Dement Geriatr Cogn Disord Extra 2020;10:27–37.
62. Martyr A, Nelis SM, Quinn C, et al. The relationship between perceived functional difficulties and the ability to live well with mild-to-moderate dementia: findings from the IDEAL programme. Int J Geriatr Psychiatry 2019;34:1251–61.
63. Ismail Z, Smith EE, Geda Y, et al. Neuropsychiatric symptoms as early manifestations of emergent dementia: provisional diagnostic criteria for mild behavioral impairment. Alzheimers Dement 2016;12:195–202.
64. McAllister-Williams RH, Bones K, Goodwin GM, et al. Analysing UK clinicians’ understanding of cognitive symptoms in major depression: a survey of primary care physicians and psychiatrists. J Affect Disord 2017;207:346–52.
65. Richard E, Schmand B, Eikelenboom P, Yang SC, Ligthart SA. Symptoms of apathy are associated with progression from mild cognitive impairment to Alzheimers disease in non-depressed subjects. Dement Geriatr Cogn Disord 2012;33:204–9.
66. Weintraub S, Besser L, Dodge HH, et al. Version 3 of the Alzheimer Disease Centers’ Neuropsychological Test Battery in the Uniform Data Set (UDS). Alzheimer Dis Assoc Disord 2018;32:10–7.
67. Harrison JE, Hendrix S. Chapter 21 – The assessment of cognition in translational medicine: a contrast between the approaches used in Alzheimer’s disease and major depressive disorder. In: Nomikos GG, Feltner DE, editors. Handb. Behav. Neurosci., vol. 29, Elsevier, 2019:297–308.
68. Roberts R, Knopman DS. Classification and epidemiology of MCI. Clin Geriatr Med 2013;29:753–72.
69. Harper L, Barkhof F, Scheltens P, Schott JM, Fox NC. An algorithmic approach to structural imaging in dementia. J Neurol Neurosurg Psychiatry 2014;85:692–8.
70. Frisoni GB, Boccardi M, Barkhof F, et al. Strategic roadmap for an early diagnosis of Alzheimer’s disease based on biomarkers. Lancet Neurol 2017;16:661–76.
71. Dubois B, Feldman HH, Jacova C, et al. Advancing research diagnostic criteria for Alzheimer’s disease: the IWG-2 criteria. Lancet Neurol 2014;13:614–29.
72. Johnson KA, Fox NC, Sperling RA, Klunk WE. Brain imaging in Alzheimer disease. Cold Spring Harb Perspect Med 2012;2:a006213.
73. Dubois B, Hampel H, Feldman HH, et al. Preclinical Alzheimer’s disease: definition, natural history, and diagnostic criteria. Alzheimers Dement 2016;12:292–323.
74. Villemagne VL, Doré V, Burnham SC, Masters CL, Rowe CC. Imaging tau and amyloid-β proteinopathies in Alzheimer disease and other conditions. Nat Rev Neurol 2018;14:225–36.
75. Clark CM, Pontecorvo MJ, Beach TG, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-β plaques: a prospective cohort study. Lancet Neurol 2012;11:669–78.
76. Wong DF, Rosenberg PB, Zhou Y, et al. In vivo imaging of amyloid deposition in Alzheimer disease using the radioligand 18F-AV-45 (florbetapir F 18). J Nucl Med 2010;51:913–20.
77. Duits FH, Martinez-Lage P, Paquet C, et al. Performance and complications of lumbar puncture in memory clinics: results of the multicenter lumbar puncture feasibility study. Alzheimers Dement 2016;12:154–63.
78. Shaw LM, Arias J, Blennow K, et al. Appropriate use criteria for lumbar puncture and cerebrospinal fluid testing in the diagnosis of Alzheimer’s disease. Alzheimers Dement 2018;14:1505–21.
79. Hansson O, Seibyl J, Stomrud E, et al. CSF biomarkers of Alzheimer’s disease concord with amyloid-β PET and predict clinical progression: a study of fully automated immunoassays in BioFINDER and ADNI cohorts. Alzheimers Dement 2018;14:1470–81.
80. Blennow K, Dubois B, Fagan AM, Lewczuk P, de Leon MJ, Hampel H. Clinical utility of cerebrospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease. Alzheimers Dement 2015;11:58–69.
81. Hansson O, Lehmann S, Otto M, Zetterberg H, Lewczuk P. Advantages and disadvantages of the use of the CSF Amyloid β (Aβ) 42/40 ratio in the diagnosis of Alzheimer’s Disease. Alzheimers Res Ther 2019;11:34.
82. Lee SA, Sposato LA, Hachinski V, Cipriano LE. Cost-effectiveness of cerebrospinal biomarkers for the diagnosis of Alzheimer’s disease. Alzheimers Res Ther 2017;9:18.
83. Dolgin E. Alzheimer’s disease is getting easier to spot. Nature 2018;559:S10–2.
84. Morris E, Chalkidou A, Hammers A, Peacock J, Summers J, Keevil S. Diagnostic accuracy of 18F amyloid PET tracers for the diagnosis of Alzheimer’s disease: a systematic review and meta-analysis. Eur J Nucl Med Mol Imaging 2016;43:374–85.
85. Engelborghs S, Niemantsverdriet E, Struyfs H, et al. Consensus guidelines for lumbar puncture in patients with neurological diseases. Alzheimers Dement Amst 2017;8:111–26.
86. Barthélemy NR, Bateman RJ, Hirtz C, et al. Cerebrospinal fluid phospho-tau T217 outperforms T181 as a biomarker for the differential diagnosis of Alzheimer’s disease and PET amyloid-positive patient identification. Alzheimers Res Ther 2020;12:26.
87. O’Bryant SE, Mielke MM, Rissman RA, et al. Blood-based biomarkers in Alzheimer disease: current state of the science and a novel collaborative paradigm for advancing from discovery to clinic. Alzheimers Dement 2017;13:45–58.
88. C2N Diagnostics. Press release. Alzheimer’s breakthrough: C₂N first to offer a widely accessible blood test. C2N Diagn 2021. (accessed January 25, 2021).
89. Barthélemy NR, Horie K, Sato C, Bateman RJ. Blood plasma phosphorylated-tau isoforms track CNS change in Alzheimer’s disease. J Exp Med 2020;217:e20200861.
90. Janelidze S, Stomrud E, Smith R, et al. Cerebrospinal fluid p-tau217 performs better than p-tau181 as a biomarker of Alzheimer’s disease. Nat Commun 2020;11:1683.
91. Chen J, Duan Y, Li H, Lu L, Liu J, Tang C. Different durations of cognitive stimulation therapy for Alzheimer’s disease: a systematic review and meta-analysis. Clin Interv Aging 2019;14:1243–54.
92. Cummings J, Fox N. Defining disease modifying therapy for Alzheimer’s disease. J Prev Alzheimers Dis 2017;4:109–15.
93. Birks JS, Harvey RJ. Donepezil for dementia due to Alzheimer’s disease. Cochrane Database Syst Rev 2018;6:CD001190.
94. Birks JS, Chong LY, Grimley Evans J. Rivastigmine for Alzheimer’s disease. Cochrane Database Syst Rev 2015;9:CD001191.
95. Tariot PN, Farlow MR, Grossberg GT, et al. Memantine treatment in patients with moderate to severe Alzheimer disease already receiving donepezil: a randomized controlled trial. JAMA 2004;291:317–24.
96. Cummings J. New approaches to symptomatic treatments for Alzheimer’s disease. Mol Neurodegener 2021;16:2.
97. Loy C, Schneider L. Galantamine for Alzheimer’s disease and mild cognitive impairment. Cochrane Database Syst Rev 2006;1:CD001747.
98. Atri A. The Alzheimer’s Disease Clinical Spectrum. Med Clin N Am 2019;103:263–93.
99. Berlowitz DR, Foy CG, Kazis LE, et al. Effect of intensive blood-pressure treatment on patient-reported outcomes. N Engl J Med 2017;377:733–44.
100. Porsteinsson AP, Drye LT, Pollock BG, et al. Effect of citalopram on agitation in Alzheimer’s disease – the citAD randomized controlled trial. JAMA 2014;311:682–91.
101. Sheline YI, Snider BJ, Beer JC, et al. Effect of escitalopram dose and treatment duration on CSF Aβ levels in healthy older adults: a controlled clinical trial. Neurology 2020;95:e2658–65.
102. Sheehan B. Assessment scales in dementia. Ther Adv Neurol Disord 2012;5:349–58.
103. Haubois G, Annweiler C, Launay C, et al. Development of a short form of Mini-Mental State Examination for the screening of dementia in older adults with a memory complaint: a case control study. BMC Geriatr 2011;11:59.
104. Horton DK, Hynan LS, Lacritz LH, Rossetti HC, Weiner MF, Cullum CM. An Abbreviated Montreal Cognitive Assessment (MoCA) for dementia screening. Clin Neuropsychol 2015;29:413–25.
105. Mini-Cog©. Mini-Cog© In other languages. Mini-Cog© Lang 2021. (accessed April 21, 2021).
106. Carnero Pardo C, de la Vega Cotarelo R, López Alcalde S, et al. Assessing the diagnostic accuracy (DA) of the Spanish version of the informant-based AD8 questionnaire. Neurologia 2013;28:88–94.
107. Harrison JK, Fearon P, Noel-Storr AH, McShane R, Stott DJ, Quinn TJ. Informant questionnaire on cognitive decline in the elderly (IQCODE) for the diagnosis of dementia within a secondary care setting. Cochrane Database Syst Rev 2015;3:CD010772.
108. Sanchez MA, Correa PC, Lourenço RA. Cross-cultural adaptation of the “Functional Activities Questionnaire – FAQ” for use in Brazil. Dement Neuropsychol 2011;5:322–7.
109. Kim G, DeCoster J, Huang C-H, Bryant AN. A meta-analysis of the factor structure of the Geriatric Depression Scale (GDS): the effects of language. Int Psychogeriatr 2013;25:71–81.
110. Mapi Research Trust. NPI – Officially distributed by Mapi Research Trust. Neuropsychiatr Inventory Quest NPI-Q 2021. (accessed April 21, 2021).
111. Mapi Research Trust. A-IADL-Q-SV – Amsterdam Instrumental Activity of Daily Living Questionnaire – Short version. Amst Instrum Act Dly Living Quest – Short Version -IADL-Q-SV n.d. (accessed April 21, 2021).
112. Ismail Z, Agüera-Ortiz L, Brodaty H, et al. The Mild Behavioral Impairment Checklist (MBI-C): a rating scale for neuropsychiatric symptoms in pre-dementia populations. J Alzheimers Dis 2017;56:929–38.



H. Chen1,*, S. Liu2,3,*, B. Ge2, D. Zhou2, M. Li2,5, W. Li2,5, F. Ma4,5, Z. Liu6, Y. Ji3,*, G. Huang2,5


1. School of Nursing, Tianjin Medical University, Tianjin, China; 2. Department of Nutrition and Food Science, School of Public Health, Tianjin Medical University, Tianjin, China; 3. Department of Neurology, and Tianjin Key Laboratory of Cerebrovascular and Neurodegenerative Diseases, Tianjin Huanhu Hospital, Tianjin, China; 4. Department of Epidemiology & Biostatistics, School of Public Health, Tianjin Medical University, Tianjin, China; 5. Tianjin Key Laboratory of Environment, Nutrition, and Public Health, Tianjin, China; 6. Department of Immunology, Biochemistry and Molecular Biology, 2011 Collaborative Innovation Center of Tianjin for Medical Epigenetics, Tianjin Key Laboratory of Medical Epigenetics, Tianjin Medical University, Tianjin, China; * These authors contributed equally to work.

Corresponding Author: Guowei Huang, Department of Nutrition and Food Science, School of Public Health, Tianjin Medical University, Tianjin, China,, Tel: +86-22-83336603; Yong Ji, Department of Neurology, and Tianjin Key Laboratory of Cerebrovascular and Neurodegenerative Diseases, Tianjin Huanhu Hospital, Tianjin, China,; Tel: + 86-22-59065149

J Prev Alz Dis 2021;3(8):249-256
Published online May 11, 2021,



Objectives: To evaluate the combined action of folic acid and vitamin B12 supplementation on cognitive performance and inflammation in patients with Alzheimer’s disease (AD).
Design: This was a randomized, single-blind, placebo-controlled trial.
Participants: Patients (n=120) diagnosed clinically as probable AD and in stable condition from Tianjin Key Laboratory of Cerebrovascular and Neurodegenerative Diseases.
Measurements: Individuals were randomly divided into the intervention group (n=60, folic acid 1.2 mg/d + vitamin B12 50 μg/d) and the placebo group (n=60). Cognitive performance, blood folate, vitamin B12, one carbon cycle metabolite, and inflammatory cytokine levels were measured at baseline and after 6 months. The data were analyzed using linear mixed models for repeated measures.
Results: A total of 101 participants (51 in the intervention group and 50 in the placebo group) completed the trial. Folic acid plus vitamin B12 supplementation had a beneficial effect on the MoCA total scores (P=0.029), naming scores (P=0.013), orientation scores (P=0.004), and ADAS-Cog domain score of attention (P=0.008), as compared to those of the control subjects. Moreover, supplementation significantly increased plasma SAM (P<0.001) and SAM/SAH (P<0.001), and significantly decreased the levels of serum Hcy (P<0.001), plasma SAH (P<0.001), and serum TNFα (P<0.001) compared to in the control subjects.
Conclusions: Folic acid and vitamin B12 supplementation showed a positive therapeutic effect in AD patients who were not on a folic acid-fortified diet. The findings of this study help to delineate nutrient intervention as far as public health management for the prevention of dementia is concerned.

Key words: Alzheimer’s disease, folic acid, vitamin B12, inflammation, cognitive performance.



Alzheimer’s disease (AD) is one of the most common causes of dementia, and while recent advances in symptomatic treatments targeting cognitive functions in dementia have been made, effective disease treatment remains to be achieved (1, 2).
Modifying variable risk factors has attracted special attention in the prevention of dementia, especially through nutritional strategies (3). An international consensus statement suggested that vitamin B supplementation should not be ignored for its public health significance (4). The B vitamins folic acid and vitamin B12 have many functions in the nervous system required for brain health. Folic acid and vitamin B12 can downregulate the Hcy level in AD; Hcy is an independent risk factor for AD, and folate and vitamin B12 insufficiency impels S-adenosylmethionine (SAM) to convert to S-adenosylhomocysteine (SAH), which is then converted to Hcy, and decreases methylation potential, which is associated with AD patients (5). Elevated levels of total plasma Hcy causes central nervous system damage, cognitive impairment, dementia, and AD (6). Hcy can also drive an immuno-inflammatory response by enhancing the production of pro-inflammatory cytokines related to psoriasis (7). Inflammation is reportedly involved in AD pathogenesis, and a number of important pro-inflammatory cytokines, including tumor necrosis factor alpha (TNFα), interleukin 6 (IL6), and monocyte chemotactic protein 1 (MCP1), and anti-inflammatory cytokines, including interleukin 2 (IL2) and interleukin 10 (IL10) (8), may play a role in AD progression.
However, findings on the efficacy of folic acid and vitamin B12 in the treatment of AD are inconsistent. A clinical trial of high-dose B vitamin supplementation in individuals with AD showed that it had no beneficial effect on primary cognitive function or rate of change in ADAS-cog score during 18 months in environments with folate enrichment of grains, such as in the United States (9). However, folic acid fortification might have undermined and concealed the statistical outcomes of the treatment. In the present study, we aimed to investigate the treatment outcomes of folic acid and vitamin B12 together in AD subjects who were not on a folate-fortified diet, as well as the modification of specific inflammatory markers.


Materials and methods

Research design and subject characteristics

The study was conducted by the Department of Nutrition and Food Science, Tianjin Medical University, and Department of Neurology, Tianjin Key Laboratory of Cerebrovascular and Neurodegenerative Diseases, Tianjin Huanhu Hospital, China (ChiCTR-IOR-16009731). The trial was a single-center, single-blind, placebo-controlled, parallel-group, randomized controlled trial. The effectiveness of the combination of 1.2 mg/d of folic acid and 50 μg/d of vitamin B12 in mitigating the progression of cognitive decline in AD subjects, which were on acetylcholinesterase inhibitor (AChEI) or memantine treatment but without vitamin B supplementation for three months before baseline, was monitored. Participants did not know if they were in the treatment or placebo groups.

Study design

The participants were enrolled between September 2016 and August 2018 by neurologists at Tianjin Huanhu Hospital. Patients diagnosed clinically as probable AD and in a stable condition were included in the study (10). Inclusion criteria included a Montreal Cognitive Assessment (MoCA) score of less than 22 for people over 45 years of age (11).Exclusion criteria included the presence of encephalopathy with overlapping clinical symptoms as that of AD, or the consumption of any kind of nutritional supplement within three months before baseline that interfered with nutritional status, folate, and vitamin B12 metabolism. The details of the recruitment process are shown in Figure 1.

Randomization, treatment, and compliance

All participants were on dementia medication as a basic routine therapy. After baseline diagnosis, eligible participants were randomly and equally distributed into two groups, in accordance with the random number table. The subjects did not know the random number.
The folic acid and vitamin B12 groups received three tablets of 400μg of folic acid (1.2 mg/d) and two tablets of 25 μg of vitamin B12 (50μg/d), while the placebo group received three starch tablets similar to folic acid tablets and two starch tablets similar to the vitamin B12 tablets. There were no differences in flavor, shape, color, or size of medication between the two groups. Vitamins and placebo were administered in the form of identical oral tablets. Folic acid tablets were produced by Beijing Scrien Pharmaceutical Co., Ltd., China, 400μg/tablet, national medical license no. h10970079, and vitamin B12 tablets were produced by Shanxi Lifeng Huarui Pharmaceutical Co., Ltd., 25μg/tablet, national medical license no. h14023061.
The participants were instructed to take five tablets daily immediately after breakfast for six months. The participants received vitamins per prescription by hospital doctors. The researchers encouraged and monitored the compliance of the two groups through regular telephonic follow-up calls and blood analysis. Data were captured by two different individuals and monitored electronically to ensure accuracy.

Standard protocol approvals, registrations, and patient consent

Each participant had a main caregiver, and both the participants and their caregivers signed informed consent before participating in the trial. The study was carried out in accordance with the Declaration of Helsinki and the International Conference on Harmonization Good Clinical Practice guidelines. This study was approved by the ethics committee of the Tianjin Health Service.

Outcome measures

Assessment of cognitive function

To evaluate cognitive function, the subjects underwent a professional neuropsychological assessment at the baseline and six months thereafter by a senior neuropsychologist. Cognitive function was assessed using the MoCA and the Alzheimer’s Disease Assessment Scale-Cognitive subscale (ADAS-cog) (11,12). The MoCA is a 10-minute, 30-point cognitive screening test designed to assess global cognitive function, with lower scores indicating greater cognitive impairment. The MoCA contains seven subtests, including visuospatial/ executive ability, naming, attention(attention digits, attention letters, and attention subtraction), orientation, language(language repeat, and language fluency), abstraction, and delayed recall. The ADAS-Cog was used to assess the severity of AD. Total scores on the ADAS-cog range from 0 to 70, and a higher total score indicates poorer cognitive performance. The ADAS-cog subscales were grouped into three domains: language(spoken language, comprehension, word finding, naming, and remembering test instructions), memory (word recall, word recognition, and orientation), and praxis (commands, constructional praxis, and ideational praxis). Nurses and biomedical scientists experienced in cognitive testing administered the MoCA and ADAS-Cog evaluations.

Blood sampling and analysis

Venous blood was collected at baseline and after six months. Blood was collected after overnight fasting for 10-12 h into two test tubes for evaluation. The first tube containing the clotting agent was allowed to settle for 30 min and then centrifuged for 10 min at 3000 rpm to obtain the serum, which was subsequently frozen at -80°C until further use. The folate, vitamin B12, Hcy, IL2, IL6, IL10, MCP1, TNFα, Aβ40, and Aβ42 levels were evaluated. Another tube containing EDTA was centrifuged immediately for 10 min at 2500 rpm at 4°C to obtain the plasma, which was frozen at -80 °C until further use. Plasma SAH and SAM levels were assessed.
The concentrations of serum folate and vitamin B12 were evaluated using an Abbott Architect-i2000SR automated chemiluminescence immunoassay system and its supporting kit (Abbott, USA). Serum Hcy levels were analyzed using a Hitachi 7180 automatic biochemistry analyzer (Japan). The concentrations of plasma SAH and SAM were quantified relative to standards (Sigma Chemical Co., St. Louis, MO, USA) using high-performance liquid chromatography. Serum IL2, IL6, IL10, MCP1, and TNFα, were determined by the Merck Millipore liquid-chip multi-factor product detection technology. Serum Aβ40, and Aβ42 levels were quantitatively analyzed using a standard ELISA kit (Biospec, Camarillo, CA).

Statistical analysis

Statistical analysis was performed using SPSS version 18.0 (SPSS Inc., Released 2009: Chicago, IL, USA). Data are represented as medians (quartiles), mean±standard deviations (SD), or proportions. The chi-square test for categorical variables and two-tailed Student’s t-tests for quantitative variables at baseline were used to compare the groups. Linear mixed models for repeated measures were used to evaluate the difference between groups in each outcome variable at different time points as compared to the baseline. Linear mixed models for repeated measures were used to estimate β(95% CI) to evaluate the differences among treatments. In the linear mixed models, the β coefficient indicates the change in neuropsychological index after six months of treatment. A positive β coefficient for longitudinal analysis indicated that an increase in the neuropsychological index was associated with treatment over time. The group effect indicates the difference between the two groups, the time effect indicates the difference between time points, and the interaction effect indicates the difference between the two groups over time. Statistical significance was set at P<0.05. The enrolled patients were randomly assigned to the intention-to-treat analysis at baseline.



Participant characteristics

The process of participant recruitment and enrollment during the trial are shown in Figure 1. Out of a total of 162 AD patients, 42 patients with AD were excluded because 8 had an uncertain diagnosis, 18 had caregiver issues, 11 refused to receive a cognitive evaluation, and 5 were excluded for unknown reasons. A total of 120 patients were included in the trial and randomly divided into two groups. Sixty patients were placed in the vitamin B12 and folic acid supplementation groups, and 60 patients were placed in the placebo group.
The baseline characteristics of the participants are presented in Table 1. Family history refers to patients with dementia in their parents or siblings. No significant differences were observed between individuals in the two groups. Furthermore, no significant differences were observed in the folate and vitamin B12 levels at baseline between individuals in the treatment and placebo groups (P>0.05).

Figure 1. Flow diagram for recruitment, randomization, and follow-up in the trial

Table 1. Baseline characteristics of the study population

The range of complications included hypertension, diabetes, cerebral infarction, cerebral trauma, coronary heart disease, CO poisoning, cerebral hemorrhage, exposure to heavy metals, Parkinson’s disease, tumor, anxiety, and depression. BMI, body mass index; AChEI, acetylcholinesterase inhibitor; MoCA, Montreal Cognitive Assessment


Cognitive status

The effects of treatment on cognitive performance are shown in Table 2. Linear mixed models for repeated measures indicated that after 6 months, the significance of the interaction effect in MoCA total scores (β [95% CI]:1.033 [0.105, 1.961], P=0.029) indicates that an increase in MoCA total scores was associated with treatment over time. The significance of the interaction effect in MoCA naming scores (β [95% CI]:0.250 [0.053, 0.447], P=0.013) indicates that an increase in MoCA naming scores was associated with treatment over time. The significance of the group effect in MoCA orientation scores (P=0.015) indicates that there was a difference between individuals in the intervention and placebo groups.The significance of the time effect in MoCA orientation scores (P=0.024) indicates that there was a difference between time points.The significance of the interaction effect in MoCA orientation scores (β [95% CI]: 0.550 [0.183, 0.917], P=0.004) indicates that an increase in MoCA orientation scores was associated with treatment over time. The significance of the time effect in MoCA abstraction scores (P=0.036) indicates that there was a difference between the time points. The significance of the group effect in ADAS-Cog language scores (P=0.027) indicates that there was a difference between individuals in the intervention and placebo groups, and the significance of the time effect in ADAS-Cog language scores (P=0.047) indicates that there was a difference between time points. The significance of the group effect in ADAS-Cog attention scores (P=0.017) indicates that there was a difference between individuals in the intervention and placebo groups, and the significance of the interaction effect in ADAS-Cog attention scores (β [95% CI]: -0.675[-1.162, -0.188], P=0.008) indicates that a decrease in ADAS-Cog attention scores was associated with treatment over time.

Table 2. MoCA and ADAS-Cog test outcomes at baseline and at 6 months

Variables are presented as median ( P25, P75) or mean±SD. The cognitive function was analyzed by using linear mixed models for repeated measures.* P<0.05 compared with placebo group.


Level of vitamin cofactors in Alzheimer’s dementia

The levels of vitamin cofactors were evaluated at baseline and six-months in this trial (Table 3). According to the linear mixed models for repeated measures, individuals in the folic acid and vitamin B12 groups showed a significant group effect in serum folate level(P <0.001), indicating that there was a difference between individuals in the intervention and placebo groups. Significance of interaction effect in serum folate level (β [95% CI]:29.690 [21.315, 38.066], P<0.001) indicates that an increase in folate was associated with treatment over time. The significance of the interaction effect in serum vitamin B12 level (β [95% CI]:110.632 [56.070, 165.193], P<0.001) indicates that an increase in vitamin B12 was associated with treatment over time. The significance of the interaction effect in serum Hcy level (β [95% CI]:-3.101[-4.351, -1.852], P<0.001) indicates that a decrease in serum Hcy level was associated with treatment over time. The significance of group effect in plasma SAM level (P <0.001) indicates that there was a difference between individuals in the intervention and placebo groups. The significance of the interaction effect in plasma SAM level (β [95% CI]:52.707 [37.555, 67.858], P<0.001) indicates that an increase in plasma SAM level was associated with treatment over time. The significance of group effect in plasma SAH level (P <0.001) indicates that there was a difference between individuals in the intervention and placebo groups, and the significance of the time effect in plasma SAH level (P <0.001) indicates that there was a difference between time points. Significance of interaction effect in plasma SAH level (β [95% CI]:-29.380[-36.534, -22.225], P<0.001) indicates that a decrease in plasma SAH level was associated with treatment over time. Significance of group effect in plasma SAM/SAH (P <0.001) indicates that there was a difference between individuals in the intervention and placebo groups. The significance of the interaction effect in plasma SAM/SAH (β [95% CI]:1.000 [0.587, 1.414], P<0.001) indicates that an increase in plasma SAM/SAH was associated with treatment over time.

Table 3. The concentrations of vitamin cofactors in blood at the baseline and 6 months

Variables are presented as median ( P25, P75) or mean±SD. Variables were analyzed by using linear mixed models for repeated measures.* P<0.05 compared with placebo group.


Levels of peripheral inflammatory cytokines and biomarkers in Alzheimer’s dementia

The concentrations of inflammatory cytokines and biomarkers in Alzheimer’s dementia were evaluated for six months in the current trial (Table 4). According to the linear mixed models for repeated measures, the significance of group effect in serum TNFα level (P=0.003) indicates that there was a difference between individuals in the intervention and placebo groups. The significance of the time effect in serum TNFα level (P =0.043) indicates that there was a difference between time points; significance of interaction effect in serum TNFα level (β [95% CI]:-37.105[-56.811, -17.398], P<0.001) indicates that a decrease in serum TNFα level was associated with treatment over time. The significance of the effect of time on serum IL2 levels (P=0.036) indicates that there was a difference between the time points.

Table 4. The concentration of inflammatory cytokines and biomarkers in Alzheimer’s dementia at baseline and 6 months in the trial

Variables are presented as median ( P25, P75) or mean±SD. Variables were analyzed using linear mixed models for repeated measures.* P<0.05 compared with placebo group.



This was a single-blind, randomized, placebo-controlled trial of daily oral folic acid (1.2 mg/d) and vitamin B12 (50 μg/d) supplementation in individuals with AD conducted over six months. In this trial, we found that the MoCA total scores, naming, and orientation increased in individuals in the treatment group compared to individuals supplemented with placebos. The variation in total ADAS-Cog score did not differ between individuals in these two groups, but the ADAS-Cog domain score of attention decreased significantly in subjects in the treatment group compared to subjects in the placebo group. This study also demonstrated that vitamin B supplements decreased serum Hcy, serum TNFα and plasma SAH levels and increased plasma SAM levels and SAM/SAH.

Outcome of folic acid and vitamin B12 supplementation on cognition decline

This trial revealed the significance of the interaction effect in MoCA total, naming, and orientation scores, indicating that an increase in the respective scores was associated with treatment over time. We also found that the ADAS-Cog domain score of attention decreased significantly in subjects in the treatment group compared to subjects in the placebo group. A previous study showed that a tHcy concentration between 10.0 to 18.0 μmol/L in the blood of AD patients had a significant positive correlation with the rate of cognitive decline (13). Another study demonstrated that the incidence rate of dementia in the elderly with blood tHcy concentrations>15.0 μmol/L was almost 5 times higher than that in the elderly with tHcy<10.1μmol/L (14). Here, the treatment and placebo groups had 91.67% and 85% of the subjects, respectively, with Hcy levels greater than 10μmol/L. Of note, China does not have a folate-fortified food policy, which implies that the outcomes of this study were undoubtedly due to the effect of vitamin B supplementation, highlighting the significantly beneficial effect of this therapy.
The outcomes of this study are in line with those of previous studies on AD patients with mild dementia (15, 16). The combination of folic acid and vitamin B12 decreased Hcy levels and mitigated cognitive decline in subjects with mild cognitive impairment (MCI) (17). In a randomized trial,140 participants with mild to moderate AD or vascular dementia received 1 mg and 5 mg of folic acid and methylcobalamin, respectively, or placebo, once a day for 24 months. No apparent differences were observed in global cognition between the two groups, but the decline in the Chinese Mattis Dementia Rating Scale (construction domain) was lower in subjects in the treatment group than in subjects inthe placebo group, which contained patients with elevated total Hcy plasmalevels (>13 μmol/L) (18). Increased levels of Hcy are associated with a high risk of coronary stroke, heart diseases, and cognitive impairment, which commonly accompany AD and dementia (19,20). In a study in MCI, vitamin B supplementation demonstrated efficacy as it reduced Hcy levels and delayed cognitive decline (21). Moreover, we found that folic acid could delay cognitive decline in patients with AD in previous studies (22). As folic acid and vitamin B12 interact with each other, this study explored the therapeutic potential of the combined effect of folic acid and vitamin B12 supplementation in AD patients.

Outcome of folic acid and vitamin B12 supplements on nucleotide derivatives of methionine

In this study, the significance of the interaction effect in serum folate level indicated that an increase in folate level was associated with treatment over time, and the significance of the interaction effect in serum vitamin B12 level indicates that an increase in vitamin B12 was associated with treatment over time. Moreover, the interaction effects showed that the significant decrease in Hcy and SAH level, and significant increase in SAM and SAM/SAH level, was associated with treatment over time. We found that after 6 months, the level of folate, vitamin B12, and SAM increased by 39.2 nmol/L, 209.17 pmol/L, and 46.28 nmol/L, respectively, as compared to individuals in the placebo group, in which folate and SAM levels decreased by 0.13 nmol/L and 6.43 nmol/L, respectively, and vitamin B12 increased by16.97 pmol/L.
Vitamin B12 acts together with folate to provide methyl groups for Hcy-mediated SAM formation. SAM provides a methyl group for the methylation of proteins, DNA, and RNA. Demethylation of SAM leads to its conversion to SAH, and SAH, in turn, is hydrolyzed to form homocysteine to continue the methionine cycle (23,24).Deficiency of folate and vitamin B12 can not only affect the formation and transformation of nucleotides, but also deplete the methylation levels of DNA and protein and increase the Hcy levels(25). Hyperhomocysteinemia can damage vascular endothelial cells and increase the risk for cardiovascular, cerebrovascular, and neurodegenerative diseases (26,27). A meta-analysis indicated that the effects of homocysteine-lowering on cognitive decline was inconclusive, as the trials typically did not include individuals who were experiencing such decline (28). Trials in high-risk subjects, which have taken into account the baseline B vitamin status, show a slowing of cognitive decline and atrophy in critical brain regions which is consistent with the modification of the AD process (29).

Outcome of folic acid and vitamin B12 supplementation on inflammatory cytokines

In this study, the daily oral intake of folic acid and vitamin B12 supplements decreased serum TNFα levels, which was associated with treatment over time. Previous studies have shown that the TNFα level in thecerebrospinal fluid (CSF) of AD and MCI patients is higher than in individuals with normal cognitive functions (30, 31). TNFα synthesized by peripheral immunocompetent cells enters the brain parenchyma and CSF through a well-structured blood-brain barrier (32). The increased levels of peripheral cytokines are associated with stem cell malfunction and decreased hippocampal volume and memory (33-35). A systematic review showed that the concentration of inflammatory markers significantly changed in AD and MCI subjects compared to healthy subjects (36). This suggests that the inflammatory markers observed in the periphery and CSF are associated with AD and MCI.
In this study, we found a significant expression of inflammatory factors, such as TNFα protein, in subjects in the vitamin B group compared to subjects in the placebo group. This indicates that folic acid, coupled with vitamin B12, significantly affects the inflammatory process in AD. In addition, this study has shown that inflammatory factors play a crucial role in AD. We propose that the underlying mechanism of the effects of folic acid plus vitamin B12 on AD is that folic acid, a methyl donor, and vitamin B12 as a cofactor, inhibits the expression of inflammation-related proteins in the TNFα pathway by DNA methylation; however, further in-depth investigation is required to verify this.
This study had certain limitations. The short duration of the study would have undermined the therapeutic effects of folate and vitamin B12. In addition, the methylation levels of the inflammatory factors were not evaluated, which should be examined in future studies.



In this study, we found that folic acid together with vitamin B12 improved cognitive performance, decreased Hcy and SAH levels, increased SAM levels, and inhibited the expression of inflammatory factors. Folic acid and vitamin B12 supplementation showed a certain therapeutic effect in AD subjects who were not on a folate-fortified diet. These findings have public health management implications for the prevention of dementia. Moreover, the combined action of folic acid and vitamin B12 significantly affects the inflammatory process in patients with AD, which deserves future consideration in subsequent investigations.


Conflicts of interest: None declared.

Acknowledgments: The authors thank all the subjects for their participation. The authors declare no potential conflicts of interest relevant to this article. This study was supported by a grant from the National Natural Science Foundation of China (No. 81730091), Tianjin Science and Technology Support Program (No. 15ZCZDSY01040), China Postdoctoral Science Foundation (No. 2017M611179), and the Scientific Program of the Tianjin Education Committee (No. 2016YD20).

Ethical standards: The study was carried out in accordance with the Declaration of Helsinki and the International Conference on Harmonization Good Clinical Practice Guidelines.



1. Kivimäki M, Luukkonen R, Batty GD, et al. Body mass index and risk of dementia: analysis of individual-level data from 1.3 million individuals. Alzheimers Dement 2018; 14(5):601-609.
2. Dubois B, Epelbaum S, Nyasse F, et al. Cognitive and neuroimaging features and brain β-amyloidosis in individuals at risk of Alzheimer’s disease (Insight-preAD): a longitudinal observational study. Lancet Neurol 2018; 17(4):335-346.
3. Barnard ND, Bush AI, Ceccarelli A, et al. Dietary and lifestyle guidelines for the prevention of Alzheimer’s disease. Neurobiol Aging 2014; 35 Supplement 2:S74-S78.
4. Smith AD, Refsum H, Bottiglieri T, et al. Homocysteine and dementia: an international consensus statement. J Alzheimers Dis 2018; 62(2):561-570.
5. Zheng M, Zhang M, Yang J, et al. Relationship between blood levels of methyl donor and folate and mild cognitive impairment in Chinese patients with type 2 diabetes: a case-control study. J Clin Biochem Nutr 2014; 54(2):122-128.
6. Li JG, Barrero C, Gupta S, et al. Homocysteine modulates 5-lipoxygenase expression level via DNA methylation. Aging Cell 2017; 16(2):273-280.
7. Lin X, Meng X, Song Z. Homocysteine and psoriasis. Biosci Rep 2019; 39(11):BSR20190867.
8. Yin P, Wang X, Wang S, et al. Maresin 1 improves cognitive decline and ameliorates inflammation in a mouse model of Alzheimer’s disease. Front Cell Neurosci 2019; 13:466.
9. Aisen PS, Schneider LS, Sano M, et al. High-dose B vitamin supplementation and cognitive decline in Alzheimer disease: a randomized controlled trial. JAMA 2008; 300(15):1774-1783.
10. Jack CR Jr, Albert MS, Knopman DS, et al. Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011; 7(3):257-262.
11. Roalf DR, Moore TM, Mechanic-Hamilton D, et al. Bridging cognitive screening tests in neurologic disorders: a crosswalk between the short Montreal Cognitive Assessment and mini-mental state examination. Alzheimers Dement 2017; 13(8):947-952.
12. Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer’s disease. Am J Psychiatry 1984; 141(11):1356-1364.
13. Oulhaj A, Refsum H, Beaumont H, et al. Homocysteine as a predictor of cognitive decline in Alzheimer’s disease. Int J Geriatr Psychiatry 2010; 25(1):82-90.
14. Ravaglia G, Forti P, Maioli F, et al. Homocysteine and folate as risk factors for dementia and Alzheimer disease. Am J Clin Nutr 2005; 82(3):636-643.
15. Ravi SK, Narasingappa RB, Vincent B. Neuro-nutrients as anti-Alzheimer’s disease agents: A critical review. Crit Rev Food Sci Nutr 2019; 59(18):2999-3018.
16. Remington R, Bechtel C, Larsen D, et al. A Phase II randomized clinical trial of a nutritional formulation for cognition and mood in Alzheimer’s disease. J Alzheimers Dis 2015; 45(2):395-405.
17. Ma F, Zhou X, Li Q, et al. Effects of folic acid and vitamin B12, alone and in combination on cognitive function and inflammatory factors in the elderly with mild cognitive impairment: A single-blind experimental design. Curr Alzheimer Res 2019; 16(7):622-632.
18. Kwok T, Lee J, Law CB, et al. A randomized placebo controlled trial of homocysteine lowering to reduce cognitive decline in older demented people. Clin Nutr 2011; 30(3):297-302.
19. Seshadri S, Beiser A, Selhub J, et al. Plasma homocysteine as a risk factor for dementia and Alzheimer’s disease. N Engl J Med 2002; 346(7):476-483.
20. Ford ES, Smith SJ, Stroup DF, et al. Homocyst(e)ine and cardiovascular disease: a systematic review of the evidence with special emphasis on case-control studies and nested case-control studies. Int J Epidemiol 2002; 31(1):59-70.
21. DeJager CA, Oulhaj A, Jacoby R, et al. Cognitive and clinical outcomes of homocysteine-lowering B-vitamin treatment in mild cognitive impairment: a randomized controlled trial. Int J Geriatr Psychiatry 2012; 27(6):592-600.
22. Chen H, Liu S, Ji L, et al. Folic acid Supplementation Mitigates Alzheimer’s disease by Reducing Inflammation: A randomized Controlled Trial. Mediators Inflamm 2016; 2016:5912146.
23. Laurino P, Tawfik DS. Spontaneous emergence of S-adenosylmethionine and the evolution of methylation. Angew Chem Int Ed Engl 2017; 56(1):343-345.
24. Crider KS, Yang TP, Berry RJ, et al. Folate and DNA methylation: a review of molecular mechanisms and the evidence for folate’s role. Adv Nutr 2012; 3(1):21-38.
25. Wang L, Chang S, Wang Z, et al. Altered GNAS imprinting due to folic acid deficiency contributes to poor embryo development and may lead to neural tube defects. Oncotarget 2017; 8(67):110797-110810:29340017.
26. Ostrakhovitch EA, Tabibzadeh S. Homocysteine and age-associated disorders. Ageing Res Rev 2019; 49:144-164.
27. Chrysant SG, Chrysant GS. The current status of homocysteine as a risk factor for cardiovascular disease: a mini review. Expert Rev Cardiovasc Ther 2018; 16(8):559-565.
28. McCaddon A, Miller JW. Assessing the association between homocysteine and cognition: reflections on Bradford Hill, meta-analyses, and causality. Nutr Rev 2015 Oct; 73(10):723-735.
29. Smith AD, Refsum H. Homocysteine, B vitamins, and cognitive impairment. Annu Rev Nutr 2016 Jul 17; 36:211-239.
30. Brosseron F, Krauthausen M, Kummer M, et al. Body fluid cytokine levels in mild cognitive impairment and Alzheimer’s disease: a comparative overview. Mol Neurobiol 2014; 50(2):534-544.
31. Tarkowski E, Liljeroth AM, Minthon L, et al. Cerebral pattern of pro- and anti-inflammatory cytokines in dementias. Brain Res Bull 2003; 61(3):255-260.
32. Wong D, Prameya R, Dorovini-Zis K. In vitro adhesion and migration of T lymphocytes across monolayers of human brain microvessel endothelial cells: regulation by ICAM-1, VCAM-1, E-selectin and PECAM-1. J Neuropathol Exp Neurol 1999; 58(2):138-152.
33. Das S, Basu A. Inflammation: a new candidate in modulating adult neurogenesis. J Neurosci Res 2008; 86(6):1199-1208.
34. Marsland AL, Gianaros PJ, Abramowitch SM, et al. Interleukin-6 covaries inversely with hippocampal grey matter volume in middle-aged adults. Biol Psychiatry 2008; 64(6):484-490.
35. McAfoose J, Baune BT. Evidence for a cytokine model of cognitive function. Neurosci Biobehav Rev 2009; 33(3):355-366.
36. Shen XN, Niu LD, Wang YJ, et al. Inflammatory markers in Alzheimer’s disease and mild cognitive impairment: a meta-analysis and systematic review of 170 studies. J Neurol Neurosurg Psychiatry 2019; 90(5):590-598.