jpad journal

AND option

OR option


H. Miao1,*, K. Chen2,*, X. Yan3, F. Chen4

1. Department of Endocrinology, Jiangjin Central Hospital of Chongqing, Chongqing, China; 2. Out-Patient Department, Jinzi Mountain Hospital of Chongqing Mental Health Center, Chongqing, 401147, China; 3. Department of Pediatrics, Jiangjin Central Hospital of Chongqing, Chongqing, China; 4. Department of Geriatrics, Chongqing Mental Health Center, Chongqing, China; * Hongyun Miao and Ken Chen are co-first author

Corresponding Author: Fei Chen, First Psychogeriatric Ward, Jinzi Mountain Hospital of Chongqing Mental Health Center, No102, Jinzi Mountain, Jiangbei, District, Chongqing, 401147, China. Phone:+86-023-67511695, Fax: +86-023-67511695, e-mail:; Xiaoyong Yan, Department of Pediatrics, Jiangjin Central Hospital of Chongqing , No.725 Jiangzhou Street, Jiangjin District, Chongqing,402260,China. Phone: +86-023-475213

J Prev Alz Dis 2020;
Published online November 9, 2020,




Background: This study aimed to investigate the association between sugar in beverage and dementia, Alzheimer Disease (AD) dementia and stroke.
Methods: This prospective cohort study were based on the US community-based Framingham Heart Study (FHS). Sugar in beverage was assessed between 1991 and 1995 (5th exam). Surveillance for incident events including dementia and stroke commenced at examination 9 through 2014 and continued for 15-20 years.
Results: At baseline, a total of 1865 (63%) subjects consumed no sugar in beverage, whereas 525 (18%) subjects consumed it in 1-7 servings/week and 593 (29%) in over 7 servings/week. Over an average follow-up of 19 years in 1384 participants, there were 275 dementia events of which 73 were AD dementia. And 103 of 1831 participants occurred stroke during the follow-up nearly 16 years. After multivariate adjustments, individuals with the highest intakes of sugar in beverage had a higher risk of all dementia, AD dementia and stroke relative to individuals with no intakes, with HRs of 2.80(95%CI 2.24-3.50) for all dementia, 2.55(95%CI 1.55-4.18) for AD dementia, and 2.11(95%CI 1.48-3.00) for stroke. And the same results were shown in the subgroup for individuals with median intakes of sugar in beverage.
Conclusion: Higher consumption of sugar in beverage was associated with an increased risk of all dementia, AD dementia and stroke.

Key words: Sugar in beverage, dementia, Alzheimer disease, stroke, Framingham Heart Study.



Dementia is a major cause of disability among the elderly and its medical and economic burdens on society have been increasing worldwide (1). Recent population-based studies have reported that lifestyle-related diseases such as hypertension (2), diabetes as well as lifestyle factors such as smoking habits (3, 4), alcohol consumption and dietary patterns are associated with the risk of incidence of dementia and AD (5, 6). However, the influence of particular individual dietary components at habitual intake amounts on dementia, especially Alzheimer disease (AD) dementia and stroke which share common risk factors and etiologies with dementia, remains incompletely understood (5).
Sugar consumption is excessive worldwide nowadays, which is widely recognized as key dietary risk factors for obesity and poor cardiometabolic health (7, 8). And longitudinal studies indicate that frequent sugar-sweetened beverages consumption is associated with a higher risk of type 2 diabetes and cardiovascular disease and raises the possibility that higher concentrations of ceramide which is associated with adverse metabolic health (9). Moreover, to our knowledge, studies yet showed that higher intake of sugary beverages was associated cross-sectionally with markers of preclinical AD and the risk of incident dementia (10, 11). Generalizability of these findings was limited due to the less accurate for sugar in beverage and less comprehensive outcomes only including dementia or preclinical AD without stroke.
Accordingly, the primary aim of this study was to examine whether sugar in beverage were associated with the 15-year risks of incident dementia and stroke in the community-based Framingham Heart Study (FHS) with detailed review of all medical records and autopsies. And we aimed to further investigate the underlying mechanisms by taking into account socio-demographic characteristics, medical history, and lifestyle factors.



Study Design

This study was carried out as a secondary analysis of data from the population-based FHS. The FHS is a longitudinal multi-generational study conducted in Framingham, MA, USA (12). The original cohort started in 1949 and consisted initially of 5209 respondents of a random sample of two-thirds of the adult population characterized by absence of atherosclerotic cardiovascular disease. The ‘Framingham Offspring Study’ was initiated in 1971 as a sample (n=5 5124, age = 12–58 years) consisting of the surviving descendants of the original cohort participants and spouses of those descendants.[13] Since their recruitment, participants from the Offspring cohort have had 9 serial examinations including standardized interviews, physician examinations, and laboratory testing. The characteristics and study protocol of both cohorts have been published elsewhere. For the present investigation, we included participants of the offspring cohort who were examined between 1990 and 1994 and for whom data of a detailed pain examination were available. This study complied with the Declaration of Helsinki; written informed consent was obtained from all study participants. The National Heart, Lung, and Blood Institute (NHLBI) of National Institutes of Health (NIH) has approved the study protocol.

Sugar in beverage assessment

Over the study course, Participants completed the Harvard semi quantitative FFQ at examination 5. For further details, see This FFQ is designed to measure dietary intake over the past year and has been validated extensively. Participants were asked how often they consumed one glass, bottle, or can of each sugary beverage item, on average, across the previous year. Each item was scored according to nine responses spanning from “never or almost never” to “61 per day.” Participants were presented with three items on sugar-sweetened soft drink (“Coke, Pepsi, or other cola with sugar,” “caffeine-free Coke, Pepsi, or other cola with sugar, “and “other carbonated beverages with sugar”), four items on fruit juice (“apple juice,” “orange juice,” “grapefruit juice,” and “other juice”), one item on sugar-sweetened fruit drinks and three items on diet soft drink (“low-calorie cola with caffeine,” “low-calorie caffeine-free cola,” and “other low-calorie beverages”). The individual items were summed to create exposure variables reflecting intake of (I) total sugary beverages (excluding diet soft drinks), (II) fruit juice, (III) sugar-sweetened soft drinks (all sugary beverages excluding diet soft drink, fruit juice, and fruit drinks), and (IV) diet soft drinks. Total sugary in beverage consumption was examined as 0-140 servings per week and was further categorized into a three-level variable: none intake (0 serving/week), 1-7 servings/week and >7 servings/week. Moreover, intake of sugary in beverages ascertained via the FFQ is reliable when measurements are repeated after 12 months, with correlation coefficients ranging from 0.85 for cola and 0.86 for fruit juice. Details were further described in FHS coding manual (http: // manuals/vr_ffreq_ex05_1_0575d_coding_manual.pdf).

Ascertainment of dementia and AD

All FHS participants were under ongoing continuous surveillance for onset of cognitive impairment and clinical dementia. We related beverage consumption to the 15-year risk of dementia. Surveillance commenced from examination cycle 5 to the time of incident event over a maximum of 41 years or until last known contact with the participant. A diagnosis of dementia was made in line with the Diagnostic and Statistical Manual of Mental Disorders, 4th edition. A diagnosis of Alzheimer’s disease (AD) dementia was based on the criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the AD and Related Disorders Association for definite, probable, or possible AD. See the online-only data supplement for complete details on our methods of surveillance, diagnosis, and case ascertainment on dementia and AD (14, 15). (eMethods in the Supplement).

Ascertainment of stroke

Stroke incidence was assessed through the continuous monitoring of hospital admissions in Framingham and by reviewing all available medical records and results (16). Stroke was defined as focal neurological symptoms of rapid onset and presumed vascular origin, lasting >24 hours or resulting in death within 24 hours. A committee comprising of least 3 FHS investigators, including at least 2 neurologists, adjudicated stroke diagnosis. The committee considered all available medical records, brain imaging, cerebrovascular imaging, and the assessment of the study neurologist who visited the participant. See more details on the FHS code manual (http: //


The most recent covariate information during the period of sugar in beverage assessment (Exam 5). We calculated nutrient, energy intake and alcohol consumption from the aforementioned FFQ. Hypertension was defined as systolic blood pressure>140 mm Hg, diastolic blood pressure>90 mm Hg, or use of antihypertensive medications. Individuals with fasting plasma glucose concentrations≥7 mmol/L or who self-reported use of antidiabetic medications were considered diabetic. BMI was calculated as kg/m2. Waist circumference (in inches) was measured at the level of the umbilicus. Current smokers were defined as participants who smoked regularly in the year preceding examination cycle 5.The educational level were assessed by medical interview. Total cholesterol, low-density lipoprotein cholesterol and glucose were measured after an overnight (>10 hours) fast.

Statistical Analysis

Descriptive statistics was performed for the 3 subgroups: participants with no intake for sugar in beverage, 1-7 servings/week and >7 servings/week. For continuous variables mean, standard deviation and range were calculated for approximately normally distributed data, otherwise median and range were used. For discrete data, absolute and relative frequencies were computed. Follow-up for dementia and stroke was from the baseline examination to the time of incident event. And for persons with no incident events, follow-up was censored at the time of death or the date the participant was last known to be dementia free. For survival analysis, Cox proportional hazards modelling was applied and the following covariates were included: crude analysis; Model 1 adjusted for age and sex; Model 2 adjusted for age, sex, and the following dementia risk factors: hypertension, smoking, diabetes and body mass index. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. The same analysis was carried out for Alzheimer’s disease and stroke. Analyses were performed using SAS software, version 9.4 (SAS Institute Inc, Cary, North Carolina) and Stata statistical software, version 15 (Stata Corporation, College Station, Texas, USA). A 2-sided with P<0.05 was considered statistically significant.




Of the 3702 participants of the FHS Offspring Cohort attending the 5th exam, 2983 subjects underwent the sugar in beverage examination and thus were included in this study. Thereof, 770 subjects were excluded because 730 subjects lacked assessment on dementia, 4 subjects were diagnosed with dementia at baseline and 36 subjects were lost follow-up. Thus, 2664 subjects could be included in the final analyses for dementia, with 1865(63%) subjects with no intake for sugar in beverage, 525(18%) subjects with 1-7 servings/week and 593(29%) subjects with over 7 servings/week. And 2916 subjects were available for incident stroke analysis, since 40 lacked assessment on dementia and 13 were lost follow-up. See Figure 1.

Figure 1. Selection of Study Participants in the Framingham Heart Study

Descriptive data are presented in Table 1. The initial health status was comparable between both groups. The three groups descriptively differed with regard to age, sex, BMI, total caloric intake, saturated fat, dietary fiber, smoking status and history of diabetes. See Table 1.

Table 1. Characteristics of the study sample of the Framingham Heart Study at baseline sugar in the beverage assessment

Abbreviations: BMI, body mass index; WC, waist circumference. LDL, low-density lipoprotein, TC: total cholesterol.


Sugar in beverage and Risk for Dementia and stroke

The mean follow-up time for dementia was 19 years (interquartile range, 18-20years) and for stroke was 16 years (interquartile range, 15-18 years).The absolute number of all dementia events over the follow-up was 553 including 142 for AD dementia. And 203 stroke occurred during the follow-up.
When adjusting for age, sex, diabetes, hypertension, smoking and BMI using Cox regression analysis, subjects with more than 7 servings/week of sugar in beverage showed a higher risk of all dementia, AD dementia and stroke [HR 2.80 (2.24-3.50) for all dementia; 2.55(1.55-4.18) for AD dementia; 2.11(1.48-3.00) for stroke] than subjects with no intake for sugar in beverage. And the results were comparable for participants with 1-7 servings/week of sugar in beverage [HR 2.65(2.14-3.30) for all dementia; 2.49(1.55-3.99) for AD dementia; 1.94(1.38-2.72) for stroke] (Table 2.).And the results were comparable when limiting the age to over 65 years for incidence of dementia and AD dementia, but not for the stroke which may be resulted from the rare stroke events occurred (eTable 1 in supplement). Figure 2 shows the cumulative incidence curves for all dementia, AD dementia and stroke stratified by groups with different levels of sugar in beverage intake after adjusting for age and sex.

Table 2. Cumulative hazards based on sugar in beverage intake

Abbreviations: AD, Alzheimer’s disease; HR, hazard ratio; CI, confidence interval; Model 1 Sex and Age; Model 2 in addition for hypertension, smoking, diabetes and body mass index.

Figure 2. Adjusted cumulative incidence of dementia and stroke based on sugar in beverage intake



In our longitudinal analysis of a large community-based sample, it is found that higher consumption of sugar in beverage was associated with an increased risk of all dementia, AD dementia and stroke at the same time. These findings were striking given that they were evident in a middle-aged sample and were observed even after statistical adjustment for numerous confounders such as sex, age, hypertension, diabetes and BMI.
Excess intake of sugar in beverage is known to be associated with cardiovascular disease and metabolic disease, which, in turn, is associated with vascular brain injury (17, 18). Although this suggests a possible link between sugary beverage consumption and vascular brain injury, we observed more striking associations between sugar in beverage and incidence of dementia and AD dementia. The Nurses’ Health Study and Health Professionals Follow-Up Study reported that higher consumption of sugar and artificially sweetened soft drinks was each independently associated with a higher risk of incident stroke over 28 years of follow-up for women and 22 years of follow-up for men (19). On the contrast, a prospective cohort study including 2888 participants reported an association between daily intake of artificially sweetened soft drink and an increased risk of both all-cause dementia and dementia because of AD, but not sugar-sweetened (10). Stephan and colleagues speculated that high fructose intake is a risk factor for dementia and that increasing consumption of fructose in the U.S. population could lead to greater dementia risk, but there is little persuasive evidence in humans at typical intake levels (20). In a cross-sectional observations in a large community-based sample, higher sugary beverage intake was associated with markers of preclinical AD, including brain atrophy and poorer episodic memory (11). To our knowledge, our study is the first to demonstrate accelerated probability of developing all dementia, AD dementia and stroke year-on-year at a population level and much more accurate assessments for sugar in beverage. And it is not a cross-sectional observations, but a long-term implications, especially for more than 5-year follow-up. In detail, it is like a fraud that excessive consumption of sugar in beverage is safe in the short term like less than 5-year follow-up, but would increase the risk of dementia, AD dementia and stroke in the long term. And only in a prospective study the truth can be told.
First, the metabolic changes followed by intake of sugar in beverage may harm cognitive function and lead to stroke, although the mechanisms are incompletely understood, and inconsistent findings have been reported. Sugar in beverage leading to a rapid rise in blood glucose and insulin, providing a plausible mechanism to link the sugar in beverage consumption to the development of dementia and stroke risk factors (21). A study shown that long term consumption of sucrose-sweetened water led to increased body weight, glucose intolerance, insulin resistance, and hypercholesterolemia and were associated with exacerbation of spatial learning and memory impairment and cerebral Aβ deposition in transgenic mouse model of AD (22). Also in animal models, high fat refined sugar diets increase Aβ aggregation and plasma total tau level, hippocampal atrophy, and reduce levels of brain-derived neurotropic factor (BDNF) which could lead to impaired memory performance and reduced synaptic plasticity within the hippocampus (11, 23, 24). Secondly, the association may be mediated by lifestyle factors related to high intake of sugar in beverage. A study recruiting 640 adolescents suggested that sugar-sweetened beverage intake may attenuate the beneficial effects of physical activity on skeletal muscle mass and lead to obesity (25). A survey demonstrated that individuals who consumed diet beverages were simply more likely to smoke regularly like what is showed in our baseline characters (26). And it is confirmed that low physical activity and muscle mass, obesity and smoking regularly were risky for incident dementia (4, 27-29). In a large, dementia-free community-dwelling cohort, our results provide further evidence that sugar in beverage consumption is associated with increased risk of all dementia, AD dementia and stroke after adjustment for confounders including sex, age, BMI, diabetes and hypertension. These findings are highly relevant given that dementia and stroke are two of the largest global public health challenge interfering with a person’s daily living activities facing our aging population.
The main strength of our study was the use of a large and well-characterized community-based sample free of clinical stroke and dementia based on FHS which collected detailed dietary, lifestyle, clinical data and accurate assessment for sugar in beverage. Limitations of our study include the observational nature of the study, which precludes conclusions about causality and the temporal associations between sugar in beverage intake and dementia, AD dementia and stroke. Second, the use of a self-report FFQ to quantify sugar in beverage intake data may be subject to recall bias. Third, there is multifactor affecting the relationship between sugar in beverage and cognitive decline. Although we addressed confounding in numerous ways, we cannot exclude the possibility of residual confounding.



In conclusion, the findings of this research demonstrate that sugar in beverage consumption was associated with an increased risk of all dementia, AD dementia and stroke independent from multiple demographic factors. These finding may help to further advise people consume less sugar in beverage to keep better health. Future research is needed to replicate our findings and to investigate the mechanisms underlying the reported associations.


Data Availability: Data described in the manuscript, code book, and analytic code will not be made available because the authors are prohibited from distributing or transferring the data and codebooks on which their research was based to any other individual or entity under the terms of an approved NHLBI Framingham Heart Study Research Proposal and Data and Materials Distribution Agreement through which the authors obtained these data.

Conflicts of Interest: The authors declare that they have no conflicts of interest.

Acknowledgments: The authors thank the National Heart, Lung, and Blood Institute Framingham Heart Study, Framingham, MA, Chongqing Mental Health Center and Jiangjin Central Hospital of Chongqing.

Grant Support: None. All authors have read the journal’s authorship agreement and that the manuscript has been reviewed by and approved by all named authors.

Ethical Standards: The study procedures followed were in accordance with the ethical standards of the Institutional Review Board and the Principles of the Declaration of Helsinki.



1. Gao S, Burney HN, Callahan CM, Purnell CE, Hendrie HC. Incidence of Dementia and Alzheimer Disease Over Time: A Meta-Analysis. J Am Geriatr Soc 2019;67, 1361-1369.
2. Turana Y, Tengkawan J, Chia YC, Hoshide S, Shin J, Chen CH, Buranakitjaroen P, Nailes J, Park S, Siddique S, Sison J, Ann Soenarta A, Chin Tay J, Sogunuru GP, Zhang Y, Wang JG, Kario K. Hypertension and Dementia: A comprehensive review from the HOPE Asia Network. J Clin Hypertens (Greenwich) 2019;21, 1091-1098.
3. Gudala K, Bansal D, Schifano F, Bhansali A. Diabetes mellitus and risk of dementia: A meta-analysis of prospective observational studies. J Diabetes Investig 2013;4, 640-650.
4. Otuyama LJ, Oliveira D, Locatelli D, Machado DA, Noto AR, Galduroz JCF, Prince MJ, Ferri CP. Tobacco smoking and risk for dementia: evidence from the 10/66 population-based longitudinal study. Aging Ment Health, 2019;1-11.
5. Nagata T, Ohara T, Hata J, Sakata S, Furuta Y, Yoshida D, Honda T, Hirakawa Y, Ide T, Kanba S, Kitazono T, Tsutsui H, Ninomiya T. NT-proBNP and Risk of Dementia in a General Japanese Elderly Population: The Hisayama Study. J Am Heart Assoc 2019;8, e011652.
6. Matsushita S, Higuchi S.[Alcohol and the Risk of Dementia]. Brain Nerve 2016;68, 819-827.
7. Mozaffarian D. Dietary and Policy Priorities for Cardiovascular Disease, Diabetes, and Obesity: A Comprehensive Review. Circulation 2016;133, 187-225.
8. Huang Y, Pomeranz J, Wilde P, Capewell S, Gaziano T, O’Flaherty M, Kersh R, Whitsel L, Mozaffarian D, Micha R. Adoption and Design of Emerging Dietary Policies to Improve Cardiometabolic Health in the US. Curr Atheroscler Rep 2018;20, 25.
9. Walker ME, Xanthakis V, Moore LL, Vasan RS, Jacques PF. Cumulative sugar-sweetened beverage consumption is associated with higher concentrations of circulating ceramides in the Framingham Offspring Cohort. Am J Clin Nutr 2010;111, 420-428.
10. Pase MP, Himali JJ, Beiser AS, Aparicio HJ, Satizabal CL, Vasan RS, Seshadri S, Jacques PF. Sugar- and Artificially Sweetened Beverages and the Risks of Incident Stroke and Dementia: A Prospective Cohort Study. Stroke 2017;48, 1139-1146.
11. Pase MP, Himali JJ, Jacques PF, DeCarli C, Satizabal CL, Aparicio H, Vasan RS, Beiser AS, Seshadri S. Sugary beverage intake and preclinical Alzheimer’s disease in the community. Alzheimers Dement 2017;13, 955-964.
12. Dawber TR, Meadors GF, Moore FE, Jr. Epidemiological approaches to heart disease: the Framingham Study. Am J Public Health Nations Health 1951;41, 279-281.
13. Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring Study. Design and preliminary data. Prev Med 1975;4, 518-525.
14. American Psychatric Association.Arlington V. Diagnostic and Statistical Manual of Mental Disorders.4th ed, American Psychiatric Publishing., 2000
15. Macfarlane GJ, Barnish MS, Jones GT. Persons with chronic widespread pain experience excess mortality: longitudinal results from UK Biobank and meta-analysis. Ann Rheum Dis 2017;76, 1815-1822.
16. Weinstein G, Beiser AS, Preis SR, Courchesne P, Chouraki V, Levy D, Seshadri S. Plasma clusterin levels and risk of dementia, Alzheimer’s disease, and stroke. Alzheimers Dement (Amst) 2016;3, 103-109.
17. Hirahatake KM, Jacobs DR, Shikany JM, Jiang L, Wong ND, Steffen LM, Odegaard AO. Cumulative intake of artificially sweetened and sugar-sweetened beverages and risk of incident type 2 diabetes in young adults: the Coronary Artery Risk Development In Young Adults (CARDIA) Study. Am J Clin Nutr 2019;110, 733-741.
18. Chun S, Choi Y, Chang Y, Cho J, Zhang Y, Rampal S, Zhao D, Ahn J, Suh BS, Pastor-Barriuso R, Lima JA, Chung EC, Shin H, Guallar E, Ryu S. Sugar-sweetened carbonated beverage consumption and coronary artery calcification in asymptomatic men and women. Am Heart J 2016;177, 17-24.
19. Bernstein AM, de Koning L, Flint AJ, Rexrode KM, Willett WC. Soda consumption and the risk of stroke in men and women. Am J Clin Nutr 2012;95, 1190-1199.
20. Stephan BC, Wells JC, Brayne C, Albanese E, Siervo M. Increased fructose intake as a risk factor for dementia. J Gerontol A Biol Sci Med Sci 2010;65, 809-814.
21. Tey SL, Salleh NB, Henry J, Forde CG. Effects of aspartame-, monk fruit-, stevia- and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake. Int J Obes (Lond) 2017;41, 450-457.
22. Orr ME, Salinas A, Buffenstein R, Oddo S. Mammalian target of rapamycin hyperactivity mediates the detrimental effects of a high sucrose diet on Alzheimer’s disease pathology. Neurobiol Aging 2014;35, 1233-1242.
23. Stranahan AM, Norman ED, Lee K, Cutler RG, Telljohann RS, Egan JM, Mattson MP. Diet-induced insulin resistance impairs hippocampal synaptic plasticity and cognition in middle-aged rats. Hippocampus 2008;18, 1085-1088.
24. Molteni R, Barnard RJ, Ying Z, Roberts CK, Gomez-Pinilla F. A high-fat, refined sugar diet reduces hippocampal brain-derived neurotrophic factor, neuronal plasticity, and learning. Neuroscience 2002;112, 803-814.
25. Hao G, Pollock NK, Harris RA, Gutin B, Su S, Wang X. Associations between muscle mass, physical activity and dietary behaviour in adolescents. Pediatr Obes 2019;14, e12471.
26. Wolkenstein P, Misery L, Amici JM, Maghia R, Branchoux S, Cazeau C, Voisard JJ, Taieb C. Smoking and dietary factors associated with moderate-to-severe acne in French adolescents and young adults: results of a survey using a representative sample. Dermatology 2015;230, 34-39.
27. Stephen R, Hongisto K, Solomon A, Lonnroos E. Physical Activity and Alzheimer’s Disease: A Systematic Review. J Gerontol A Biol Sci Med Sci 2017;72, 733-739.
28. Moon Y, Choi YJ, Kim JO, Han SH. Muscle profile and cognition in patients with Alzheimer’s disease dementia. Neurol Sci 2018;39, 1861-1866.
29. Arnoldussen IA, Kiliaan AJ, Gustafson DR. Obesity and dementia: adipokines interact with the brain. Eur Neuropsychopharmacol 2014;24, 1982-1999.



C. Udeh-Momoh1, G. Price1, M.T. Ropacki2, N. Ketter3, T. Andrews5, H.M. Arrighi2, H.R. Brashear3, C. Robb1, D.T. Bassil1, M. Cohn1,6, L.K. Curry1, B. Su1, D. Perera1, P. Giannakopoulou1, J. Car6, H.A. Ward7, R. Perneczky1,8,9, G. Novak4, L. Middleton1


1. Imperial College London, Neuroepidemiology and Ageing Research, School of Public Health, London, UK; 2. Janssen Research & Development, LLC, Fremont, California, USA; 3. Janssen Alzheimer Immunotherapy Research & Development, LLC, South San Francisco, California, USA; 4. Janssen Research & Development, LLC, Titusville, USA; 5. Central and North West London NHS Foundation Trust, London, UK; 6. Imperial College London, Department of Primary Care and Public Health, School of Public Health, London, UK; 7. Imperial College London, Department of Epidemiology and Biostatistics, School of Public Health, London, UK; 8. University Hospital, LMU Munich, Department of Psychiatry and Psychotherapy, München, Germany; 9. German Center for Neurodegenerative Diseases (DZNE) Munich, München, Germany

Corresponding Author: Gerald Novak, MD, Janssen Research & Development, LLC, 1125 Trenton-Harbourton Rd., Titusville, NJ 08560, USA, Tel.:+1 609 730 4416, Fax: +1 908 730 2069, Email:

J Prev Alz Dis
Published online July 10, 2019,



Background: The CHARIOT PRO Main study is a prospective, non-interventional study evaluating cognitive trajectories in participants at the preclinical stage of Alzheimer’s disease (AD) classified by risk levels for developing mild cognitive impairment due to AD (MCI-AD).
Objectives: The study aimed to characterize factors and markers influencing cognitive and functional progression among individuals at-risk for developing MCI-AD, and examine data for more precise predictors of cognitive change, particularly in relation to APOE ε4 subgroup.
Design: This single-site study was conducted at the Imperial College London (ICL) in the United Kingdom.  Participants 60 to 85 years of age were classified as high, medium (amnestic or non-amnestic) or low risk for developing MCI-AD based on RBANS z-scores. A series of clinical outcome assessments (COAs) on factors influencing baseline cognitive changes were collected in each of the instrument categories of cognition, lifestyle exposure, mood, and sleep. Data collection was planned to occur every 6 months for 48 months, however the median follow-up time was 18.1 months due to early termination of study by the sponsor.
Results: 987 participants were screened, among them 690 participants were actively followed-up post baseline, of whom 165 (23.9%) were APOE ε4 carriers; with at least one copy of the allele. The mean age was 68.73 years, 94.6% were white, 57.4% were female, and 34.8% had a Family History of Dementia with a somewhat larger percentage in the APOE ε4 carrier group (42.4%) compared to the non-carrier group (32.4%). Over half of the participants were married and 53% had a Bachelor’s or higher degree.  Most frequently, safety events typical for this population consisted of upper respiratory tract infection (10.4%), falls (5.2%), hypertension (3.5%) and back pain (3.0%).
Conclusion (clinical relevance): AD-related measures collected during the CHARIOT PRO Main study will allow identification and evaluation of AD risk factors and markers associated with cognitive performance from the pre-clinical stage. Evaluating the psycho-biological characteristics of these pre-symptomatic individuals in relation to their natural neurocognitive trajectories will enhance current understanding on determinants of the initial signs of cognitive changes linked to AD.

Key words: CHARIOT, aging registry, cognitive health, pre-clinical, Alzheimer Disease.



An increasing body of scientific evidence suggests that, in the field of Alzheimer’s disease (AD), the optimal time to intervene with disease-modifying therapies is prior to the emergence of clinical symptoms of mild cognitive impairment (MCI) or AD dementia (1).  The term “asymptomatic at-risk state for AD” (ARAD) (2) has been proposed as a descriptor for asymptomatic individuals with evidence of cerebral amyloid-β (Aβ) burden.  Such individuals are at increased risk of progression to clinically symptomatic AD (3) and hence, potentially, good candidates for trials of preventative interventions.
Amongst cognitively normal (CN) individuals, subtle deficits or decreases in cognitive performance over time (even within the range of “normal” values) have been found to be associated with higher Aβ burden and/or carriage of the apolipoprotein (APOE) epsilon 4 allele with subsequent cognitive decline (3) and progression to MCI or dementia due to AD (4, reviewed in 5).  This suggests that evolution of certain cognitive profiles may be sufficient in themselves (even in the absence of supporting biomarker information) to constitute an ARAD.
However, a putative ARAD cognitive profile has not yet been clearly identified or specified.  Whereas many publications have focused on what may be termed ‘late’ MCI defined as individuals most likely to transition to dementia within a year or two, information is limited on cognitive characterization of memory and other cognitive domains and how these manifest and change, in cognitively healthy individuals in the “pre-clinical” stages. Such information may indeed improve our understanding of the natural evolution of AD over time both cognitively and functionally, and help to identify opportunities for intervention.
The goals of this non-interventional cohort study were (a) to prospectively collect information on cognitively healthy individuals to determine the value of biomedical, lifestyle and neuropsychological markers in predicting clinical progression or cognitive decline consistent with AD; and (b) to develop a well-characterized longitudinally followed, prospective readiness cohort, asymptomatic yet at risk for AD for future clinical trials. Here, we report on the methods employed for the extensive phenotyping of the study participants, as well as the population characteristics of the sample at baseline.



Study design

The CHARIOT (Cognitive Health in Ageing Register: Investigational, Observational and Trial Studies in dementia Research) PRO (Prospective Readiness cOhort) Main Study was a prospective, single center, non-interventional study conducted at the Imperial College London (ICL) in the United Kingdom. The Main Study recruited participants between the ages of 60-85 years from the CHARIOT Register at ICL, or self-referred.  The CHARIOT Register is a community-based research register of older individuals without a diagnosis of dementia in the United Kingdom, who have provided informed consent to be invited to interventional and non-interventional studies for the prevention of AD and other age-related neurodegenerative diseases. The Register is managed by physicians, investigators, and staff of the School of Public Health (SPH) at ICL.  Established in 2011, the Register is based on a collaboration between SPH and General Practitioner surgeries in central and west London, and now consists of ~ 30,000  consented volunteers (6).
The planned sample size of enrolled participants for the CHARIOT PRO Main study was 700.  In order to yield sufficient likelihood of detecting a rare event (e.g., progression to MCI-AD), 630 participants should be enrolled with consideration of 10% overall dropout rate.
Table 1S of Supplementary Material presents the precision estimates with a sample size of 630 for varying rates (proportions) of rare event of interest.  For example, with a sample size of 630 participants, the probability of detecting at least one event with a true event rate of 0.001 is 46.8%, and for all others it is greater than this.  In total, 987 participants were screened, and 712 were eligible for follow-up, with 690 actively followed-up post baseline.
The Main Study was conducted in accordance with Good Clinical Practice (GCP) Guidelines, Guidelines for Good Pharmacovigilance Practices (GPP) issued by the International Society for Pharmacoepidemiology (ISPE), applicable national guidelines, and to the Declaration of Helsinki.  An independent ethics committee approved participant written informed consent forms before enrollment collected during the baseline clinic visit.


The investigations were aimed at better understanding the natural history of cognitive changes in asymptomatic participants that may precede the occurrence of clinically overt MCI or dementia due to AD. In addition, the study aimed to evaluate the sensitivity of baseline neuropsychological, biological and lifestyle measures for predicting longitudinal AD-related cognitive decline, in order to improve screening of individuals at risk for developing AD for future clinical trials.

Study population and selection criteria

Individuals aged 60 to 85 years without dementia were recruited and screened from the CHARIOT Register or self-referred. Concomitant therapies for treatment of stable medical conditions known in older population were permitted.
Participants were not eligible for enrollment if they met any of the following exclusion criteria: a previous diagnosis of dementia, MCI or other neurological disease or condition (such as Parkinson’s disease); met criteria for AD dementia (per National Institute on Aging-Alzheimer’s Association) at baseline; a history of traumatic brain injury, stroke or evidence of transient ischemic attack (TIA); epileptic seizures – excluding febrile seizures in childhood; significant psychiatric illness; hydrocephalus at any time; uncontrolled hypothyroidism or hyperthyroidism; any clinically significant unstable illness, metabolic problems or nutritional deficiencies; a clinically significant infection within 30 days of study entry; HIV positivity; history of alcohol or drug dependence or abuse; used memantine or cholinesterase inhibitors; chronically used medications known to impair cognition such as sedatives, anticonvulsants, or pain medications; had significant sensory or motor dysfunction; any physical disability that would prevent completion of study procedures or assessments; concurrent participation in an interventional or non-interventional trial (with exceptions, also based on PI judgement).  Following baseline assessment, participants were excluded from follow-up if their age- and education-adjusted cognitive performance (z-score) on any Index of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) fell more than 1.5 standard deviations below normal (unless adjudicated for inclusion by the sponsor’s medical monitor).

Sample size, schedule of events

The study aimed to enroll 700 participants.  Over a two-year period, 987 participants were screened and 712 were enrolled.  All enrolled participants were genotyped for apolipoprotein ε4 (APOE ε4) allele carrier or non-carrier status, and both participants and study investigators were blinded to APOE genotype status.  Screening and baseline assessments were performed by study investigators, including trained psychometricians across one or two visits within 30 days.  Enrolled participants were evaluated across the large number of study instruments every six months from baseline, for four years. The trial, however, was terminated early by the sponsor such that median follow-up time reached 18.1 months.
Outreach to participants who failed to attend their regularly scheduled bi-annual visit followed a 2-step approach. First, three attempts were made to contact the participant (via email or telephone within 1 week), and if needed a second step involved contact by a regular mail letter with delivery confirmation sent to participant’s home.  If the participant failed to respond to all outreach attempts, they were considered lost to follow-up.

Evaluations and Outcome Measures

Data collected at each time point per schedule of assessments is shown in Table 2S of Supplementary Materials.  These included evaluations of medical status, vital signs, anthropometrics, cognitive function, mood, sleep, diet, physical and leisure activity, functional activity and biological sample collection (urine, saliva and blood). Use of medications known to impair cognition was prohibited within 48 hours or 4 times the half-life (whichever longer) before baseline cognitive assessments.  All outcome measures were administered by or under the supervision of a qualified health professional or psychologist as appropriate.
At the time of initial enrollment, participants were classified as hypothetically at high (67, 9.4%), medium (91, 12.8%) or low (554, 77.8%) risk for developing MCI-AD, based on the participant’s age and education-adjusted baseline cognitive performance on the RBANS Indices, as shown in Table 1.

Table 1. Participant Risk Group Classification

Table 1. Participant Risk Group Classification

RBANS = Repeatable Battery for the Assessment of Neuropsychological Status, SD = standard deviation



All participants completed the RBANS, Mini Mental State Examination (MMSE), the Memory and Executive Function modules of the Neuropsychological Assessment Battery (NAB), and the National Adult Reading Test (NART). In addition, three supplemental cognitive assessments were administered: the CogState Brief Battery (CBB), the Cognitive Drug Research Assessment System (CDR-AS), or the Trail-Making and Verbal Fluency subtests of the Delis Kaplan Executive Function System (DKEFS). To minimise participant burden and fatigue, each participant was randomly allocated to undertake only one of these three assessments.  Randomization was stratified by RBANS risk classification.  Once randomized at baseline, the participant retained the same allocation at all follow-up visits.

Mini-Mental State Examination (MMSE)

The MMSE is a brief 11-item face-to-face examination used to screen for cognitive impairment, to estimate the severity of cognitive impairment at a given point in time, to follow the course of cognitive changes in an individual over time, and to document an individual’s response to treatment.  The examination includes stimuli for comprehension, reading, writing, and drawing tasks.  It is widely translated and has shown validity and reliability in psychiatric, neurologic, geriatric and other medical populations (7).

Repeatable Battery for the Assessment of Neuropsychological Status (RBANS)

The RBANS (8) includes 12 subtests yielding five indices: the Attention Index is comprised of Digit Span and Coding, the Language Index consists of Picture Naming and Semantic Fluency subtests, the Visuospatial/Construction Index is made up of Figure Copy and Line Orientation subtests, the Immediate Memory Index is comprised of List Learning and Story Memory subtests, and the Delayed Memory Index consists of List Recall, List Recognition, Story Recall and Figure Recall subtests. This face-to-face assessment takes approximately 25 minutes to complete.

Neuropsychological Assessment Battery (NAB)

NAB is a comprehensive, modular neuropsychological test battery designed to assess a range of cognitive skills and functions of adults from 18 to 97 years old. The specific cognitive domains of attention/concentration, language, memory, visuospatial and executive functioning are measured by specific co-developed and normed modules within the NAB.  For the purposes of this investigation, the Memory and Executive Function modules were administered as these functions are considered to be most impacted early in the AD disease course (9). The NAB is administered face-to-face, and assessment time depends upon disease severity, with more impaired participants completing it faster.  Specifically, the NAB Memory module takes approximately 45 minutes, whereas the Executive Function module is approximately 30 minutes in duration.

National Adult Reading Test (NART)

NART is a widely used measure of word reading that assesses pronunciation of 50 English words with irregular grapheme-phoneme and stress rules.  Reading tests, such as the NART, have been shown to provide a good estimate of premorbid intellectual functioning, including in patients with neurodegenerative disorders (10). The NART is administered face-to-face and takes approximately 10 minutes to complete.

CogState Brief Battery (CBB)

The CogState Brief Battery is an approximately 15 minute computerized battery with demonstrated reliability, validity, and short term stability (11), developed expressly for maximal sensitivity to detect change.  Employing playing cards as stimuli to assure cross-cultural acceptability, CogState consists of four tasks that respectively measure the functions of attention, processing speed, visual learning, and working memory. CogState employs standard psychometric paradigms (i.e., simple and choice reaction time, n-back and pattern separation learning), and has been validated for detection of dementia in both clinical and community based screening samples.  Change over time (6-18 months) on the pattern separation learning task has been seen in healthy older adults testing positive for amyloid compared with those negative for amyloid (12). CogState can be administered via the internet or on a stand-alone computer and is available in over 50 languages.

Cognitive Drug Research Assessment System (CDR-AS)

The Cognitive Drug Research Assessment System (13) is an approximately 20 minute computerized battery designed to reliably measure changes in cognitive function in clinical trial situations.  The fully automated system includes tests of episodic memory, working memory, attention and reaction time.

Delis Kaplan Executive Function System (DKEFS)

The DKEFS is a paper and pencil measure of verbal and nonverbal executive functions and has been normed and validated for children and adults from 8 to 89 years of age.  The measure consists of nine subtests. For the purposes of this study, the Trail Making Test (TMT) and Verbal Fluency subtests were used.  These two paradigms have a long history of frequent use in AD research (14). Total time to complete these two subtests of the DKEFS is approximately 20 minutes.


Geriatric Depression Scale (GDS)

GDS is a basic self-reported screening test used to identify depression in older adults (15).  The 15-question version asks participants how they felt over the past week, and uses a Yes/No response format to enable the questionnaire to be used with moderately cognitively impaired individuals.

State-Trait Anxiety Scale

The Spielberger State-Trait Anxiety Inventory (STAI, forms Y-1 and Y-2) is a 40-item self-report instrument to assess current state anxiety and general anxiety levels (16). The STAI includes twenty items to assess the presence or absence of current (state) anxiety and twenty to assess general (trait) predisposition to anxiety, with each item scored from 1 to 4 according to intensity or frequency.

Patient Reported Outcome Measures

Revised Perceived Deficits Questionnaire (PDQ)

The PDQ was originally designed to capture decline in cognitive function most often caused by multiple sclerosis. In recent years, the PDQ has been used in at least one study of MCI (17).  The PDQ is a 20-item questionnaire that covers four domains of cognitive function from the participant’s perspective: attention/concentration, retrospective memory, prospective memory, and planning/organization. The response options are answered on a five-point Likert scale, with 0 = never, 1 = rarely, 2 = sometimes, 3 = often, and 4 = almost always. Subscales can be calculated by summing raw scores for the relevant five items (subscale range is 0 to 20), and the total score is calculated by summing raw scores for all of the PDQ items (scale range is 0 to 80). A higher score indicates greater perceived cognitive impairment.

Work Productivity and Activity Impairment (WPAI)

The WPAI Questionnaire is an instrument to measure impairments in both paid work and unpaid work, yet un-validated within an older population, some of whom remain in employment including volunteer work. The scale consists of 6 questions regarding work and activity impairment due to health problems.  The WPAI elicits data on hours worked, hours missed due to the target condition, hours missed due to other health problems and hours missed for any other reasons.  Hours missed for «other reasons» is not used in the scoring, but only as a prompt to the respondent to exclude those hours from the count of actual hours worked. The WPAI yields four types of scores: (1) absenteeism (work time missed), (2) presenteeism (impairment at work / reduced on-the-job effectiveness), (3) work productivity loss (overall work impairment/absenteeism plus presenteeism), and (4) activity impairment. The sum of specific health problem impairment and impairment due to other health reasons is equal to impairment due to all health reasons.  WPAI outcomes are expressed as impairment percentages, with higher numbers indicating greater impairment and less productivity, that is, worse outcomes (18).

Health Utilities Index Mark 3 (HUI3)

The HUI3 is a generic, preference-weighted, health status assessment completed by the participant that measures health status and health-related quality of life and allows the computation of utility scores.  The 15-item questionnaire (15Q) is designed for self-completion, includes 15 multiple-choice HUI3 questions plus one global health question (Q16) common in many health surveys (19).

Lifestyle Measures

Imperial Lifestyle Questionnaire (ILQ)

Participants were asked to complete a self-reported questionnaire to address a wide range of health and lifestyle characteristics: demographics (age, marital status, ethnicity); socioeconomic status (education, income, employment status, type of occupation); activities of daily living (assessed by the Lawton scale (20)); occupational and leisure time physical activity (the Physical Activity for the Elderly Scale, PASE(21)) and the short form of the International Physical Activity Questionnaire, IPAQ(22)); leisure activities (frequency of social visits, reading, musical and artistic pastimes, speaking a second language, solo recreational mental activities e.g. puzzles); midlife experiences (occupation, physical and leisure activities, travel, training); smoking (type, quantity, duration; details of smoking at age 40, time since cessation; second-hand smoke exposure); health history (diagnosed conditions, related treatments, Rose angina questionnaire, surgical procedures, multivitamin use, weight and dietary change over time, use of medical services, family history); and female reproductive history (menstruation, childbearing, breastfeeding, hormone replacement therapy use).
A follow-up version of the above-listed content was administered every six months after baseline.  The follow-up questionnaire was as above, except for factors that would not have changed since baseline: subsets of the questions on demographics (ethnicity, early education), the series on mid-life experiences, history of surgical procedures, previously diagnosed medical conditions, history of weight and dietary changes, history of use of medical services, family health history, childbearing and HRT use).

Scottish Collaborative Group Food Frequency Questionnaire

The semi-quantitative Scottish Collaborative Group food frequency questionnaire (SCG-FFQ) is a 150-item instrument designed to assess the habitual diet of United Kingdom residents over the previous 3 months.  The SCG-FFQ is derived from the dietary questionnaires used in the Scottish Heart Health/ MONICA Study. The SCG-FFQ provides quantitative estimates of the intake of food and nutrients and is appropriate for ranking individuals into broad categories of intake (e.g., high, medium, and low) as opposed to absolute levels of intake. The SCG-FFQ has been validated among older adults in the United Kingdom (23).


Willing participants were requested to wear actigraph wristwatch-like device that monitored rest and activity cycles for a prescribed time period to assess job-related, transportation-related, and recreation, sport and leisure-time physical activities. The measure also captures and quantifies periods of inactivity (e.g., sitting).


Berlin Questionnaire

The Berlin Questionnaire is a simple sleep apnea screening questionnaire (10 items) used to quickly identify the risk (low to high) of sleep disordered breathing. The questionnaire consists of 3 categories and risk is based on the responses to individual items and overall scores in the symptom categories (24).

Pittsburgh Sleep Quality Index (PSQI)

The PSQI is a self-rated questionnaire which assesses sleep quality and disturbance over a 1 month time period. Nineteen individual items generate seven “component” scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication and daytime dysfunction. The sum of scores for the 7 components gives 1 global score ranging from 0 (better) to 21 (worse). A total score of >5 is associated with poor sleep quality (25).

Safety Evaluation

Study events, whether serious or non-serious, were recorded throughout the study period from the time of informed consent until completion of the participant’s last study-related procedure.  A serious study event meets one or more of the following parameters: fatal; immediately life-threatening; requires hospitalization or prolongs existing hospitalization; permanently (or significantly) disabling; a congenital anomaly or birth defect (in an offspring); or medically significant.  All serious study events were reported to the sponsor by study-site personnel within 30 days of their knowledge of the event. Each suspected adverse event included reporting of description (e.g. signs and symptoms or diagnosis), seriousness criteria, severity rating, duration (onset and resolution date), actions taken and outcome.



Sample disposition and baseline characteristics

The study started in February 2014 and truncated in December of 2016 with a median follow-up time of 18.1 months.  The early discontinuation was due to introduction of a follow-on ongoing substudy that enrolled participants from the Main study and the Register and was designed to enhance the scientific strength of the main study objectives, through the addition of more detailed AD-related assessments, including biomarker evaluation of participants’ Aβ status (positron emission tomography and/or cerebrospinal fluid protein analysis), alongside brain structural and functional explorations via Magnetic Resonance Imaging (MRI).
For the main study, a total of 690 participants who met the primary analysis set criteria at baseline were analyzed.
Participant disposition is shown in Figure 1S of Supplementary Materials. The overall study attrition rates were 28% screen failure (275 out of 987), and 13% post baseline (91 out of 712).
No participant completed the study as it was terminated early. Discontinuations post baseline were primarily due to early study termination by sponsor (486, 83.1%), secondly due to withdrawal by the participant and lost to follow-up (80, 13.7%), and thirdly due to other reasons such as physician decision and protocol deviation (19, 3.2%). The participant overall estimated median time in study was 18.1 months, with 21.6 months for APOE ε4 carriers, and 17.8 months for non-carriers. The annual attrition due to participant dropout was within the expected range of approximately 10% per year. 72.5% of enrolled participants completed their 12 month visit, 51.4% completed their 18 month visit, 29.7% and 9.9 % completed their 24 month and 30 month visits respectively.  Study participation rate by APOE ε4 status is shown in Figure 1S.  The attrition rate was lower amongst APOE ε4 carriers than amongst non-carriers at all follow-up time-points, but due to incomplete follow-up and small sample sizes, the difference was not tested for statistical significance.
Demographic and baseline disease characteristics by participant APOE ε4 status are shown in Table 2.  The mean (SD) age was 68.73 (3.757) years and was similar between APOE ε4 carriers and non-carriers. This age range is younger than might be expected in AD interventional trials as the initial inclusion criteria for age range was 60-75 years, but was increased to 85 years in a much later amendment to more closely reflect expected age of participants in interventional trials.  The majority of study participants were white (94.6%) and 57.4% were female with very similar percentages in the two sub-groups.  Of the 690 participants followed up post baseline, 165 (23.9%) were APOE ε4 carriers (the vast majority with 1 copy of the allele (97.0%)), and 525 (76.1%) were non-carriers. The percent of participants with family history of dementia of any type was 34.8%, with a somewhat larger percentage in the carrier group (42.4%) compared to the non-carrier group (32.4%).  Over half of the participants (58.6%) were married, and 53% had a Bachelor’s degree or higher- level education reflecting a high socioeconomic status. Summary statistics of baseline cognitive measures for participant by APOE ε4 status is shown in Table 3. The baseline values for the primary cognition outcome measures were numerically comparable between APOE ε4 carriers and noncarriers. There were no meaningful patterns of difference in performance for any of the assessment scale baseline values across the two subgroups.

Table 2. Participant demographic characteristic by APOE ε4 status

Table 2. Participant demographic characteristic by APOE ε4 status

Notes: ApoE4 = apolipoprotein E (E4 allele), a Family history of AD or dementia includes first-degree relative, parents, or siblings, percentages are calculated with the number of all enrolled subjects with a given demographic or disease characteristic in each column as the denominators.

Table 3. Summary of Baseline Cognition Outcome Measures by ApoE4 Status

Table 3. Summary of Baseline Cognition Outcome Measures by ApoE4 Status

Key:  ApoE4 = apolipoprotein E (E4 allele), CDR = Cognitive Drug Research, Assessment System, DKEFS = Delis Kaplan Executive Function System; MMSE = Mini-Mental State Examination, NAB = Neuropsychological Assessment Battery; a. Higher scores indicate better performance; b. Lower scores indicate better performance; Score ranges: MMSE (raw): 0 to 30, DKEFS scaled scores: 0 to 19, NAB standard scores: mean=100, SD = 15; Percentages are calculated with the number of all enrolled subjects with a given cognition outcome measurement in each column as the denominators; The RBANS Index scores and Total Scale were calculated using Age based norms



As this was a non-interventional study, safety analyses focused on safety events that referred to occurrence of any untoward medical event such as any unfavorable and unintended sign, symptom, syndrome, or disease.  The incidence of any safety event was 64.8% for APOE ε4 non-carriers and 71.5% for carriers. The most frequently occurring safety events with incidence of 3% or more were upper respiratory tract infection (10.4% overall) and fall (5.2%). The incidence of serious safety events was 4.1% overall, which was numerically slightly higher among non-carriers (4.6%) than carriers (3.4%).  The most frequently occurring serious safety events were prostate cancer, renal cell carcinoma, and Parkinson’s disease, occurring in two participants each (0.3%), and all others occurred in 1 participant each.



We report on baseline characteristics of 690 cognitively healthy participants, who were prospectively evaluated every 6 months for a period of 30 months (or early termination) for changes in performance on neuropsychological test measures from baseline. Data from this study, though short in duration, will allow for examination of biological, genetic, health, and lifestyle factors and markers that influence cognitive progression among individuals at-risk for developing MCI-AD.
This study assessed APOE genotype status for all enrolled participants. Though prevalence of APOE ε4 carriers in our cohort was as expected within a cognitively normal population (~25% of enrolled participants), we noted a paucity of APOE ε4 homozygotes (~1% of enrolled participants). Observed prevalence of APOE ε4 homozygosity markedly differs from frequency reported by other cohorts that include cognitively healthy older adults. For instance, prevalence of dual ε4 alleles in the cognitively normal participant group ranged from 3.6% in the Uniform Data Set of the Alzheimer’s Disease Centres (UDS) program, to 6.1% in the Australian Imaging, Biomarkers and Lifestyle Flagship Study of Ageing (AIBL) project up to 9.2% in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) study (26). Notably these studies used total MMSE scores (24-30) to define normal cognition, as opposed to a more stringent cognitive assessment tool for designation of ‘cognitive normal’ status .
Thus, a probable explanation for the reduced homozygosity (ε4+/+) observed in the CPRO main study cohort could be the more sensitive test of cognition (RBANS) used for defining cognitive normal status during study screening and subsequent enrolment.  There was no marked APOE ε4 genotype-related difference in cognitive test performance at baseline. This was to be expected since all participants were to be cognitively healthy at baseline and performed within age-matched population norms. Modelling of the longitudinal data will inform on which of these assessments could potentially detect the very earliest cognitive changes.
Our study has important caveats and limitations that should be discussed.  The study lacked a biomarker assessment of the participants’ amyloid pathology, an important predictor of clinical progression.  In addition, no participant progressed to MCI during the study. Despite these lacks, the study included proxy measures of Aβ pathology and information on established risk factors for AD-related cognitive deterioration such as the major genetic risk determinant – APOE genotype status, as well as ample collection of relevant demographic information including age, sex, family history of dementia and subjective cognitive complaints (3, 27, 28).
Due to the relatively young age-range of the participants and the early termination of the study, the cohort may have included a significant proportion of cognitively high-functioning individuals who were unlikely to demonstrate clinical progression over the relatively short duration of follow-up.  Furthermore, the early termination of the study reduced availability of data points due to low participation rates at later time points. Nevertheless, we do not expect this attribute to limit the ability to perform longitudinal modeling and analysis, in view of related studies that have reported cognitive changes even within such limited time frames (29, 30).
The CHARIOT PRO Main Study data will be useful for assessing impact of available AD-related measures, including lifestyle exposures and biological factors, on cognitive trajectories from the pre-symptomatic stage. Such analyses may contribute towards better understanding of risk-resilience, and maintenance of cognitive function that may be evident at the preclinical AD stage, in high risk individuals.


Acknowledgements: The authors are most grateful to the study participants for their contributions and the investigational site staff for their work on the study.  The authors acknowledge Bradford Challis (Janssen Research & Development, LLC) for assistance in preparation and editorial support of this manuscript.

Study funding: Funded by Janssen Research & Development, LLC. The sponsor also provided a formal review of this manuscript.

Contributions of Authors: Dr. Udeh-Momoh served as the lead author, Drs. Udeh-Momoh, Car, Perneczky, Price, Andrews and Ward were co-investigators at School of Public Health, Imperial College London, and all contributed to the study design, coordination, data acquisition, and participated in data interpretation, development and critical review of the manuscript. Drs. Bassil, Su, Cohn, Giannakopoulou and Ms Robb, Perera and Curry were study investigators at School of Public Health, Imperial College London, and contributed to data acquisition and review of the manuscript. Dr. Ropacki contributed to the study conception, protocol design, and development of the manuscript. Drs. Novak, Arrighi, Ketter, and Brashear, contributed to study design and were responsible for data review, interpretation, and development of the manuscript. In addition, Dr. Arrighi was project pharmaco-epidemiologist and Dr. Raghavan served as project biostatistician, responsible for aspects of study design, statistical data analysis, statistical input and interpretation of data. Dr. Middleton served as the study site principal investigator, contributed to study conception and design, data review and interpretation and critical review of the manuscript. All authors met ICMJE criteria and all those who fulfilled those criteria are listed as authors. All authors had access to the study data, provided direction and comments on the manuscript, made the final decision about where to publish these data and approved the final draft and submission to this journal.

Conflict of interest/disclosures: Janssen Research & Development, LLC, funded the study. Michael T. Ropacki was formerly an employee of Janssen and is an industry consultant. Gerald. Novak, H. Michael Arrighi, and Nandini Raghavan are employees of Janssen Research & Development, LLC and own stock/stock options in the company. Nzeera Ketter is a former employee, and H. Robert Brashear is an employee of Janssen AI Research & Development, LLC and both own stock/stock options in the company. Jianing Di is an employee of Janssen China Research and Development Center and owns stock/stock options in the company. Lefkos Middleton served as principal study investigator at Imperial College of London (ICL), has a consultancy agreement with Eli Lilly, Astra Zeneca and Takeda and is National Coordinator for the TOMMORROW, Amaranth and Generation Clinical Studies; and does not hold any agreement with any of the funders in relation to patents, products in development relevant to this study or marketed products. Chinedu Udeh-Momoh, Josip Car, Robert Perneczky, Geraint Price, Tresa Andrews and Heather Ward served as co-principal study investigators at ICL for Janssen Research & Development, LLC and all declare no conflict of interest. Catherine Robb, Darina Bassil, Martin Cohn, Parthenia Giannakopoulou, Dinithi Perera, Lisa Curry and Bowen Su were study investigators at ICL and declare no conflict of interest.

Ethical standards: To ensure the quality and integrity of the research, this study was conducted in accordance with Good Clinical Practice (GCP) Guidelines, Good Pharmacoepidemiology Practices (GPPs) issued by the International Society for pharmaceutical Engineering (ISPE), applicable national guidelines, and to the Declaration of Helsinki 2013, as modified by the 52nd World Medical Assembly, Edinburgh, Scotland, 2000, and clarified by the World Medical Assembly (WMA) General Assembly, Washington 2002 and Tokyo 2004. The Chariot-Pro Main study has received National Research Ethics Services approval (15/L0/0711) and internal Imperial College London (ICL) Research Ethics, Joint Research Compliance Office approval (JRCO:15/1C/2791). Prior to consenting onto the Chariot-Pro Main study, participants were provided with a detailed study information sheet outlining study procedures, as well as risks and benefits associated with participation.  Participants were provided with a minimum of 48-hours to consider the information provided, and fully understand the study requirements prior to discussing further with study staff, where necessary; after which signed informed consent was obtained prior to undertaking any study procedures at the screening visit.





1.    Sperling RA, Aisen PS, Beckett LA, Bennett DA, Craft S, Fagan AM, et al. Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2011;7(3):280-92.
2.    Dubois B, Feldman HH, Jacova C, Cummings JL, Dekosky ST, Barberger-Gateau P, et al. Revising the definition of Alzheimer’s disease: a new lexicon. Lancet Neurol. 2010;9(11):1118-27.
3.    Donohue MC, Sperling RA, Petersen R, Sun CK, Weiner MW, Aisen PS. Association Between Elevated Brain Amyloid and Subsequent Cognitive Decline Among Cognitively Normal Persons. JAMA. 2017;317(22):2305-16.
4.    Albert M, Zhu Y, Moghekar A, Mori S, Miller MI, Soldan A, et al. Predicting progression from normal cognition to mild cognitive impairment for individuals at 5 years. Brain. 2018.
5.    Vellas B. Update on Prevention Trials in Alzheimer’s Disease. J Prev Alzheimers Dis. 2014;1(3):168-75.
6.    Larsen ME, Curry L, Mastellos N, Robb C, Car J, Middleton LT. Development of the CHARIOT Research Register for the Prevention of Alzheimer’s Dementia and Other Late Onset Neurodegenerative Diseases. PloS one. 2015;10(11):e0141806.
7.    Folstein MF, Folstein SE, McHugh PR. «Mini-mental state». A practical method for grading the cognitive state of patients for the clinician. Journal of psychiatric research. 1975;12(3):189-98.
8.    Randolph C, Tierney MC, Mohr E, Chase TN. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): preliminary clinical validity. J Clin Exp Neuropsychol. 1998;20(3):310-9.
9.    Caselli RJ, Locke DE, 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.
10.    Carlozzi NE, Stout JC, Mills JA, Duff K, Beglinger LJ, Aylward EH, et al. Estimating premorbid IQ in the prodromal phase of a neurodegenerative disease. The Clinical neuropsychologist. 2011;25(5):757-77.
11.    Lim YY, Jaeger J, Harrington K, Ashwood T, Ellis KA, Stoffler A, et al. Three-month stability of the CogState brief battery in healthy older adults, mild cognitive impairment, and Alzheimer’s disease: results from the Australian Imaging, Biomarkers, and Lifestyle-rate of change substudy (AIBL-ROCS). Archives of clinical neuropsychology : the official journal of the National Academy of Neuropsychologists. 2013;28(4):320-30.
12.    Wesnes KA, Annas P, Basun H, Edgar C, Blennow K. Performance on a pattern separation task by Alzheimer’s patients shows possible links between disrupted dentate gyrus activity and apolipoprotein E in4 status and cerebrospinal fluid amyloid-beta42 levels. Alzheimers Res Ther. 2014;6(2):20.
13.    Wesnes KA, McNamara C, Annas P. Norms for healthy adults aged 18-87 years for the Cognitive Drug Research System: An automated set of tests of attention, information processing and memory for use in clinical trials. J Psychopharmacol. 2016;30(3):263-72.
14.    Faria CA, Alves HVD, Charchat-Fichman H. The most frequently used tests for assessing executive functions in aging. Dement Neuropsychol. 2015;9(2):149-55.
15.    Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al. Development and validation of a geriatric depression screening scale: a preliminary report. Journal of psychiatric research. 1982;17(1):37-49.
16.    Spielberger C, Luschene P, Vagg P, Jacobs A. Manual for the State Trait Anxiety Inventory (Form Y) Consulting Psychologists Press: Palo Alto. SpielbergerManual for the State-Trait Anxiety Inventory1983. 1983.
17.    Doody RS, Ferris SH, Salloway S, Sun Y, Goldman R, Watkins WE, et al. Donepezil treatment of patients with MCI: a 48-week randomized, placebo-controlled trial. Neurology. 2009;72(18):1555-61.
18.    Reilly MC, Zbrozek AS, Dukes EM. The validity and reproducibility of a work productivity and activity impairment instrument. PharmacoEconomics. 1993;4(5):353-65.
19.    Horsman J, Furlong W, Feeny D, Torrance G. The Health Utilities Index (HUI): concepts, measurement properties and applications. Health and quality of life outcomes. 2003;1:54.
20.    Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9(3):179-86.
21.    Washburn RA, McAuley E, Katula J, Mihalko SL, Boileau RA. The physical activity scale for the elderly (PASE): evidence for validity. J Clin Epidemiol. 1999;52(7):643-51.
22.    Booth M. Assessment of physical activity: an international perspective. Res Q Exerc Sport. 2000;71(2 Suppl):S114-20.
23.    Jia X, Craig LC, Aucott LS, Milne AC, McNeill G. Repeatability and validity of a food frequency questionnaire in free-living older people in relation to cognitive function. The journal of nutrition, health & aging. 2008;12(10):735-41.
24.    Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Annals of internal medicine. 1999;131(7):485-91.
25.    Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry research. 1989;28(2):193-213.
26.    Heffernan AL, Chidgey C, Peng P, Masters CL, Roberts BR. The Neurobiology and Age-Related Prevalence of the epsilon4 Allele of Apolipoprotein E in Alzheimer’s Disease Cohorts. J Mol Neurosci. 2016;60(3):316-24.
27.    Liu CC, Zhao N, Fu Y, Wang N, Linares C, Tsai CW, et al. ApoE4 Accelerates Early Seeding of Amyloid Pathology. Neuron. 2017;96(5):1024-32 e3.
28.    Vogel JW, Varga Dolezalova M, La Joie R, Marks SM, Schwimmer HD, Landau SM, et al. Subjective cognitive decline and beta-amyloid burden predict cognitive change in healthy elderly. Neurology. 2017;89(19):2002-9.
29.    Ngandu T, Lehtisalo J, Solomon A, Levalahti E, Ahtiluoto S, Antikainen R, 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. Lancet. 2015;385(9984):2255-63.
30.    Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A. 2011;108(7):3017-22.



M. Rochoy1,2, E. Chazard1,3, S. Gautier1,2, R. Bordet1,2


1. Univ. Lille, F-59000 Lille, France; 2. INSERM, U1171-Degenerative and Vascular Cognitive Disorders, F-59000 Lille, France; 3. EA2694, Public Health Department, F-59000 Lille, France

Corresponding Author: Michaël Rochoy, 20 rue André Pantigny, 62230 Outreau, France. +33667576735,

J Prev Alz Dis 2019;2(6):108-111
Published online February 11, 2019,



INTRODUCTION: Alzheimer’s disease (AD) is the first cause of dementia. Diagnostic criteria have evolved: proposals to revise the NINCDS–ADRDA criteria were published in 2007. Our aim was to analyze the evolution in the coding of AD in the French nationwide exhaustive hospital discharge database (PMSI) between 2007 and 2017.
METHODS: We analyzed evolution of International Classification of Diseases and Related Health Problems, 10th edition (ICD-10) coding for AD and AD dementia in the PMSI database from 2008 to 2017 (285,748,938 inpatient stays).
RESULTS: We observed a 44% decrease in the number of inpatient stays with a principal diagnosis of AD or AD dementia from 2007 (46,313 inpatient stays) to 2017 (25,856 inpatient stays) in France. Over the same period, we observed a 49% increase in the number of inpatient stays with a principal diagnosis of related dementias (other organic mental disorders or other degenerative disorders). Overall, the number of inpatient stays for dementia remained stable despite the increase in the total number of inpatient stays: 95,377 in 2007 (0.409% of inpatient stays) and 99,190 in 2017 (0.344%).
CONCLUSION: We therefore note a shift from AD and AD dementia to other dementia diagnoses since 2007. This study suggests a more accurate use of AD related ICD-10 codes since the revised criteria in 2007.

Key words: Data reuse, big data, Alzheimer disease, PMSI, vascular dementia.



In France, dementia affected 1 to 1.2 million people directly and 3 million relatives indirectly in the early 2010’s (1, 2). Several predictive scenarios have been carried out, estimating a prevalence of 1.75 million in 2030  and 1.81 million in 2050 (3, 4). Alzheimer’s disease (AD) is the most common causes of dementia (5).
Dementia diagnosis is a clinical syndromic diagnosis, which requires fulfilling diagnostic criteria. Such criteria exist for frontotemporal dementia (6), vascular dementia (7 ,8), and dementia with Lewy bodies (9, 10). In 1984, the National Institute of Neurological Disorders and Stroke–Alzheimer Disease and Related Disorders working group (NINCDS–ADRDA) published criteria for probable AD (11). In 1999, Varma et al. showed that those criteria failed to differentiate AD from frontotemporal dementia (12). The elucidation of the biological basis of AD advanced greatly, and in 2007, a revision of the NINCDS-ADRD criteria was proposed (13). These proposals notably rely on a clinical core of early and significant episodic memory impairment. At least one of the following paraclinical abnormalities is also required: the presence of medial temporal lobe atrophy with MRI, a specific pattern on functional neuroimaging with PET, or an elevated cerebrospinal fluid concentration of amyloid-β or tau protein.
We hypothesized that the release of these proposed criteria could influence the identification of diagnoses of AD or AD dementia. The French PMSI database gives access to the evolution of diagnoses over all hospital inpatient stays over several successive years.
Our aim was to analyze the evolution of dementia ICD-10 coding in France after the revision of these criteria in 2007.



Study design

This study is based on a retrospective cohort, based on secondary use of the PMSI database (presented below). We included all the inpatient stays from 2007 to 2017, from the website ScanSanté (Technical agency for Hospital Information).

Data source

The PMSI database is the French nationwide exhaustive hospital discharge database (14). The database used in this study comprehends all the inpatient stays, from nonprofit and for-profit acute care hospitals (medicine, surgery and obstetrics), excluding psychiatric hospitals and rehabilitation care centers. This database includes administrative data (admission and discharge dates and modes), demographic data (age, gender, geographic area), diagnoses encoded in the International Classification of Diseases and Related Health Problems, 10th edition (ICD-10) (15) and other pieces of information (16). This information is anonymized and can be reused for research purposes (17).

Inclusion criteria

Dementia and related diseases encoding rules have been defined in 2006 (18). In accordance with those rules, the inpatient stays having one of the following codes as principal diagnosis were included (ICD-10 codes in brackets): AD (G30*, 4 codes), AD dementia (F00*, 84 codes).
In comparison, we noted the total number of degenerative disorders of the nervous system (G30*, G31*, G32*, for AD and other degenerative disorders of the nervous system) and organic mental disorders (codings from F00* to F09*, for AD dementia, vascular dementia, dementia during other diseases classified elsewhere, unspecified dementia, organic amnesia syndrome, non-induced delirium, other mental disorders, personality and organic behaviour disorders, organic mental disorder).



From 2007 to 2017, the total number of inpatient stays increased from 23,339,533 to 28,829,970 (+23%). The number of inpatient stays with a principal diagnosis of organic mental disorders or degenerative disorders of the nervous system remained stable despite the increase in the total number of inpatient stays: 95,377 in 2007 (0.409% of inpatient stays) and 99,190 in 2017 (0.344%). Among inpatient stays for organic mental disorders or degenerative disorders of the nervous system, AD or AD dementia accounted for 48.6% in 2007 and 26.1% in 2017 (Table 1).

Table 1. Number of inpatient stays for Alzheimer’s dementia in France

Table 1. Number of inpatient stays for Alzheimer’s dementia in France

AD: Alzheimer’s disease


Over 11 years, the number of inpatient stays with a principal diagnosis of AD or AD dementia decreased (-66% and -28% respectively): 46,313 in 2007 (0.198% of inpatient stays) and 25,856 in 2017 (0.090% of inpatient stays). Over the same period, inpatient stays for other degenerative disorders of the nervous system and for other organic mental disorders increased (+6% and +55% respectively): 49,064 in 2007 (0.210% of inpatient stays) and 73,334 (0.254% of inpatient stays) (Figure 1).

Figure 1. Evolution of subtypes of dementia among the principal diagnosis of inpatient stays in France between 2007 and 2017

Figure 1. Evolution of subtypes of dementia among the principal diagnosis of inpatient stays in France between 2007 and 2017



We observed a 44% decrease in the number of inpatient stays with a principal diagnosis of AD or AD dementia from 2007 (46,313 inpatient stays) to 2017 (25,856 inpatient stays) in France. Over the same period, we observed a 49% increase in the number of inpatient stays with a principal diagnosis of related dementias (other organic mental disorders or other degenerative disorders). Overall, the number of inpatient stays for dementia remained stable. We therefore note a shift from AD and AD dementia to other dementia diagnoses since 2007.
Several hypotheses could explain our results. First, a reduction in hospitalizations of patients with Alzheimer’s disease over the 7-year period would be possible. It seems unlikely : over a similar period, the population of people over 75 living in Metropolitan France grew by 9.6%, while the overall population grew by only 3.7% (19). In a study with a 15-year follow-up, the annual rate of hospitalization was 16.3 per 100 person.years (95% CI 15.0-17.7) and 47% were readmitted after their initial hospitalization (20). Second, a decline of dementia incidence in France would be possible, partly explained by improved management of modifiable risk factors (21, 22), or improvement of education (23). This hypothesis is supported by several recent studies (21, 22, 24–26). Nevertheless, the 44% decrease in the number of stays associated with AD or AD dementia in our study is far greater than the expected incidence decrease according to the literature. Third, an increased clinician caution would be possible with AD or AD dementia diagnoses since the revision of NINCDS–ADRDA criteria (13). Thus, AD coding seemed more focused prior to the revised criteria, but was probably inaccurate.
Principal strength of our study is to study all inpatient stays in France over 11 consecutive years, for a total of 285,748,938 stays.
Our study has potential limitations. We study here the number of inpatient stays; some patients may have had several stays. Besides, we interested in the evolution of coding, and there is no reason to believe that the annual number of inpatient stays of patients with AD may have changed between 2007 and 2017. Also, the inpatient stays ICD-10 codes were registered for billing purposes, and we did not validate them versus clinical records. The use of the PMSI for activity-based payment of hospital services can lead to over-coding for some pathologies that increase the payment, and especially under-coding for diseases that have no financial impact. However, the coding as the main diagnosis of vascular dementia (regardless of the precision) or Alzheimer’s dementia is associated with the same homogeneous group of stay, and with the same pricing for stays from 3 to 20 days. The pricing of the stay will increase by coding the associated diagnoses (undernutrition, decubitus ulcer, bedridden patient, etc.); thus, there is no financial inducement to use precisely one code of dementia rather than the other.
Nevertheless, use of hospital coding to define the diagnosis is a recognized way of counting dementia (27). Defining and identifying dementias in the major French administrative databases is an issue of interest: a recent review proposed several algorithms, two of which have been validated (28).



Our findings raise several issues. In terms of care, the shift from Alzheimer’s dementia to other diagnoses can result in more targeted care. In terms of research, the reliability of databases such as the PMSI for data mining can be questioned with diagnoses subject to significant fluctuations. ICD-10 encoding seems to be more cautious since the revised NINCDS-ADRDA criteria, and AD diagnoses are probably more reliable now, which is engaging for future studies.


Ethical standards: Not concerned. All hospitalized patients in France received written information on the possibility of such studies.

Conflict of interests: The authors do not have any conflict of interest to declare.



1.    L’état de santé de la population en France. InVS; 2017.
2.     Helmer C, Grasset L, Peres K, Dartigues J. Évolution temporelle des démences : état des lieux en France et à l’internationale. Bull Epidemiol Hebd. 2016;28–29:467–73.
3.     Jacqmin-Gadda H, Alperovitch A, Montlahuc C, Commenges D, Leffondre K, Dufouil C, et al. 20-Year prevalence projections for dementia and impact of preventive policy about risk factors. Eur J Epidemiol. 2013 Jun;28(6):493–502.
4.     Mura T, Dartigues J-F, Berr C. How many dementia cases in France and Europe? Alternative projections and scenarios 2010-2050. Eur J Neurol. 2010 Feb;17(2):252–9.
5.     2018 Alzheimer’s disease facts and figures. Alzheimers Dement J Alzheimers Assoc. 2018 Mar 1;14(3):367–429.
6.     Neary D, Snowden J, Mann D. Frontotemporal dementia. Lancet Neurol. 2005 Nov 1;4(11):771–80.
7.     Román GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology. 1993 Feb;43(2):250–60.
8.     Sachdev P, Kalaria R, O’Brien J, Skoog I, Alladi S, Black SE, et al. Diagnostic criteria for vascular cognitive disorders: a VASCOG statement. Alzheimer Dis Assoc Disord. 2014 Sep;28(3):206–18.
9.     McKeith IG, Galasko D, Kosaka K, Perry EK, Dickson DW, Hansen LA, et al. Consensus guidelines for the clinical and pathologic diagnosis of dementia with Lewy bodies (DLB): report of the consortium on DLB international workshop. Neurology. 1996 Nov;47(5):1113–24.
10.     McKeith IG, Boeve BF, Dickson DW, Halliday G, Taylor J-P, Weintraub D, et al. Diagnosis and management of dementia with Lewy bodies: Fourth consensus report of the DLB Consortium. Neurology. 2017 Jul 4;89(1):88–100.
11.     McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology. 1984 Jul;34(7):939–44.
12.     Varma AR, Snowden JS, Lloyd JJ, Talbot PR, Mann DMA, Neary D. Evaluation of the NINCDS-ADRDA criteria in the differentiation of Alzheimer’s disease and frontotemporal dementia. J Neurol Neurosurg Psychiatry. 1999 Feb 1;66(2):184–8.
13.     Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau P, Cummings J, et al. Research criteria for the diagnosis of Alzheimer’s disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 2007 Aug;6(8):734–46.
14.     Boudemaghe T, Belhadj I. Data Resource Profile: The French National Uniform Hospital Discharge Data Set Database (PMSI). Int J Epidemiol. 2017 01;46(2):392-392d.
15.     WHO | International Classification of Diseases (ICD) [Internet]. WHO. [cited 2013 Oct 28]. Available from:
16.     Bousquet C, Trombert B, Souvignet J, Sadou E, Rodrigues J-M. Evaluation of the CCAM Hierarchy and Semi Structured Code for Retrieving Relevant Procedures in a Hospital Case Mix Database. AMIA Annu Symp Proc. 2010;2010:61–5.
17.     Bezin J, Duong M, Lassalle R, Droz C, Pariente A, Blin P, et al. The national healthcare system claims databases in France, SNIIRAM and EGB: Powerful tools for pharmacoepidemiology. Pharmacoepidemiol Drug Saf. 2017 Aug 1;26(8):954–62.
15.     Schwarzinger M, Thiebaut SP, Rehm J. The neglected role of alcohol use disorders in early-onset dementia: a nationwide study from France. Alzheimers Dement J Alzheimers Assoc. 2016 Jul 1;12(7):P194–5.
19.     Agence technique de l’information sur l’hospitalisation (ATIH). Guides méthodologiques 2008 à 2014 [Internet]. [cited 2017 Aug 30]. Available from:
20.     Bivand R, Lewin-Koh N, Pebesma E, Archer E, Baddeley A, Bearman N, et al. maptools: Tools for Reading and Handling Spatial Objects [Internet]. 2017 [cited 2018 Apr 25]. Available from:
21.     R Core Team. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2014. Available from:
22.     Leys D, Hénon H, Mackowiak-Cordoliani M-A, Pasquier F. Poststroke dementia. Lancet Neurol. 2005 Nov 1;4(11):752–9.
23.     Launer LJ. Counting dementia: There is no one “best” way. Alzheimers Dement J Alzheimers Assoc. 2011 Jan;7(1):10–4.
24.     Grasset L, Brayne C, Joly P, Jacqmin-Gadda H, Peres K, Foubert-Samier A, et al. Trends in dementia incidence: Evolution over a 10-year period in France. Alzheimers Dement J Alzheimers Assoc. 2016 Mar 1;12(3):272–80.
25.     Jagger C, Andersen K, Breteler MM, Copeland JR, Helmer C, Baldereschi M, et al. Prognosis with dementia in Europe: A collaborative study of population-based cohorts. Neurologic Diseases in the Elderly Research Group. Neurology. 2000;54(11 Suppl 5):S16-20.
26.     Rudolph James L., Zanin Nicole M., Jones Richard N., Marcantonio Edward R., Fong Tamara G., Yang Frances M., et al. Hospitalization in Community-Dwelling Persons with Alzheimer’s Disease: Frequency and Causes. J Am Geriatr Soc. 2010 Aug 3;58(8):1542–8.
27.     Schrijvers EMC, Verhaaren BFJ, Koudstaal PJ, Hofman A, Ikram MA, Breteler MMB. Is dementia incidence declining?: Trends in dementia incidence since 1990 in the Rotterdam Study. Neurology. 2012 May 8;78(19):1456–63.
28.     Qiu C, von Strauss E, Bäckman L, Winblad B, Fratiglioni L. Twenty-year changes in dementia occurrence suggest decreasing incidence in central Stockholm, Sweden. Neurology. 2013 May 14;80(20):1888–94.
29.     EClipSE Collaborative Members, Brayne C, Ince PG, Keage HAD, McKeith IG, Matthews FE, et al. Education, the brain and dementia: neuroprotection or compensation? Brain J Neurol. 2010 Aug;133(Pt 8):2210–6.
30.     Matthews FE, Arthur A, Barnes LE, Bond J, Jagger C, Robinson L, et al. A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. Lancet Lond Engl. 2013 Oct 26;382(9902):1405–12.
31.     Doblhammer G, Fink A, Fritze T. Short-term trends in dementia prevalence in Germany between the years 2007 and 2009. Alzheimers Dement J Alzheimers Assoc. 2015 Mar;11(3):291–9.



J. Luo1,2,3*, H. Weng3*, J.C. Morris4,5,6, C. Xiong3,4


1. Division of Public Health Sciences, Department of Surgery, Washington University School of Medicine, St. Louis, MO, 63110, USA; 2. Siteman Cancer Center Biostatistics Core, Washington University School of Medicine, St. Louis, MO, 63110, USA; 3. Division of Biostatistics, Washington University School of Medicine, St. Louis, MO; 4. Knight Alzheimer Disease Research Center,  Washington University School of Medicine, St. Louis, MO, 63110, USA; 5. Departments of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, 63110, USA; 6. Department of Neurology,  Washington University School of Medicine, St. Louis, MO 63110, USA; * These authors contribute equally.

Corresponding Author: Chengjie Xiong, Division of Biostatistics, Campus Box 8067, 4523 Clayton Ave., St. Louis, MO, 63110-1093, Phone: 314-362-3635; Fax: 314-362-2693, Email:

J Prev Alz Dis 2018;5(2):110-119
Published online March 16, 2018,



Background: Clinical trials of investigational drugs for Alzheimer disease (AD) increasingly focus on the prodromal (symptomatic) stage of the illness and now its preclinical (asymptomatic) stage. Sensitive and specific cognitive and functional endpoints are needed to track subtle cognitive and functional changes in the early and preclinical stages to minimize sample sizes in these trials.
Objectives: To identify informative items in a standard clinical assessment protocol and a psychometric battery that are predictive of onset of dementia symptom.
Design: Longitudinal retrospective study.
Setting: Washington University (WU) Knight Alzheimer Disease Research Center (ADRC).
Participants: A total of 735 individuals at least 65 years old and cognitively normal at baseline from a longitudinal clinical cohort at the WU Knight ADRC.
Measurements: The annual clinical assessment included a wide spectrum of functional and cognitive domains; a comprehensive psychometric battery was completed about 2 weeks after the clinical evaluation. Psychometricians are blinded to the results of the clinical evaluation and to the prior performance of the participants on the psychometric tests.
Results: The mean age at baseline of the 735 participants was 74.30 and 62.31% were female. 240 individuals developed prodromal dementia symptoms (consistent with mild cognitive impairment due to AD and with very mild AD dementia) during longitudinal follow-up (mean follow-up=6.79 years). Among a total of 562 items in the clinical and cognitive assessments under analysis, 292 (52%) were identified as informative because their longitudinal changes were predictive of symptomatic onset. When these items were used to form the functional and cognitive composites, the longitudinal rates of changes were free of a learning effect and captured subtle longitudinal progression prior to symptomatic onset. The rates of change were much greater right after the symptomatic onset than those from the functional and cognitive composites formed using non-informative items. Although the sample sizes for prevention trials (prior to symptomatic onset) using the informative items still yield large numbers, the sample sizes for early treatment trial (after symptomatic onset) was much smaller than those derived from all the items or from the non-informative items alone.
Conclusions: The antecedent longitudinal changes in nearly half of the items in a clinical assessment protocol and a comprehensive cognitive battery did not show statistically significant ability to predict the dementia symptom onset, and hence may be non-informative to track the preclinical functional and cognitive progression of AD. The remaining items, on the other hand, captured some of the preclinical changes prior to the symptom onset, but performed much better right after the symptom onset. Currently ongoing prevention trials on preclinical AD of elderly individuals may need to re-assess the sample sizes and statistical power.

Key words: Age of symptom onset, Alzheimer disease, prevention trials, treatment trials, informative items, power.



Alzheimer disease (AD) is a neurodegenerative disorder characterized by the pathophysiological process of formation and accumulation of senile plaques and neurofibrillary tangles in the brain (1) and phenotypical process of progressive impairment of cognition, function and behavior. Considerable evidence accumulated in the past decade through mostly  cross-sectional studies suggests that individuals who develop symptomatic AD exhibit cognitive deficits several years before the clinical diagnosis of mild cognitive impairment (MCI) due to AD or of very mild AD (2). Because AD is an irreversible neurodegenerative disease that results from neuronal loss in one or multiple brain regions, an early detection and intervention offers the optimal hope for the disease treatment. Because current symptomatic therapies are initiated only after diagnosis, their modest benefit may be partly explained by the fact that some irreversible brain damage has already occurred by the time AD is clinically recognized. Given that no pharmaceutical treatments to date have demonstrated efficacy in reversing or stabilizing dementia progression in the mild or moderate stages of AD, antecedent disease markers when individuals are still cognitively normal or at very early symptomatic stages are especially important to identify individuals at high risk for trials of putative disease-modifying therapies to allow optimal early intervention and prevention.
Although standard cognitive and functional instruments discriminate established symptomatic AD from normal aging, they are far from satisfactory in tracking the early changes of AD, partly because of the enormous ceiling and floor effects (3). Because data with significant ceiling and floor effects have limited use in tracking the longitudinal changes of the disease, their use as part of the cognitive outcome in prevention or early treatment trials may lead to large sample sizes and  represent a waste of time and precious research resources in terms of both research participants/informants and investigators. On the other hand, there is evidence in the literature suggesting even the current composite scores of these tests can help, to some degree, identify individuals at the preclinical or early stage of AD (4). This implies that many individual items from the standard cognitive battery and functional tests may be informative in identifying individuals at high risk for symptomatic AD. Because these items are buried in and scattered across different tests that were not originally designed for tracking preclinical or very early changes of AD, their potential has not been fully appreciated due to the lack of optimal tools to identify them and to integrate them.
This paper aims to investigate whether longitudinal changes from the individual items in  a standard clinical assessment protocol and a comprehensive cognitive battery used by the Washington University (WU) Knight Alzheimer’s Disease Research Center (ADRC) were associated with onset of dementia symptoms. We hypothesize that the clinical protocol and cognitive battery contain many individual items that are neither sensitive nor specific for tracking early changes of AD. We further hypothesize that identification of items which are  informative to predict symptomatic onset (called ‘informative items’) will lead to an improved estimate to the longitudinal cognitive decline, both prior to and after symptomatic onset. Finally, we  evaluate whether the functional and cognitive endpoints defined by the informative items alone in future prevention or early treatment trials improve statistical power for efficacy comparison over the endpoints derived from the entire functional and cognitive batteries (which contain both informative and non-informative items) and  those derived from the non-informative items alone.




The WU Knight ADRC has enrolled elderly individuals in a longitudinal clinical-pathologic cohort study of aging and dementia since 1979.  Participants received annual clinical and psychometric examinations. The Clinical Dementia Rating (CDR) (5) staged the presence or absence of dementia and, when present, its severity such that CDR 0 indicates cognitively normality and CDR 0.5, 1, 2, and 3 indicates very mild, mild, moderate, and severe dementia, respectively. 735 individuals at least 65 years old and cognitively normal at baseline and followed longitudinally were included in the analyses. All study participants provided written informed consent. The study was approved by the Institutional Review Board of WU School of Medicine.

Clinical and Psychometric Assessments

The clinical and psychometric assessments were conducted independently to permit the cognitive data to be evaluated without contamination and possible circularity that might result when cognitive scores were used in diagnostic classifications. The clinical assessment at the WU Knight ADRC assesses a wide spectrum of functional and cognitive domains (each with many items) on the participants, which also includes information about the participants provided by their informants. A total of 111 items from participants and 77 items about the participants from the informants were analyzed due to their availability of longitudinal item level data. Some tests/scales were administered on both research participants and informants, at least for a period of time that resulted in longitudinal item level data, including Geriatric Depression Scale (GDS) (6), the box of Judgment and Problem Solving in the CDR sum of boxes (7), and the Short Portable Mental Status Questionnaire (8). Some tests were administered to participants only, including Mini Mental State Examination (MMSE) (9), Short Blessed Test (SBT) (10), Assessment of Aphasia (11) and a Drawing test. Some other tests were administered to informants only regarding the participant’s behavioral features and functional abilities, including the neuropsychiatric inventory questionnaire (NPI-Q) (12), Functional Assessment Questionnaire (FAQ) (13), Ferman test (14), the box of personal care and community affairs in the CDR sum of boxes, orientation and daily activities with items from the Blessed Dementia Scale (BDS) (15)  and a depressive features battery.
About 2 weeks after the clinical evaluation, participants completed an approximate 2-hour battery of psychometric tests. Psychometricians were blinded to the results of the clinical evaluation and previous performance of the participant on the psychometric tests. Episodic memory was assessed by the Wechsler Memory Scale Logical Memory, including immediate and delayed tests (16), Digit Span (both forward and backward) (16), Associate Learning subtests from the Wechsler Memory Scale (WMS) (17) and the Visual Retention Test (Form C, 10-second exposure) (18). Two measures of semantic memory included the Information subset of the Wechsler Adult Intelligence Scale (WAIS) (19) and the Boston Naming Test (20). WAIS Block Design was also included for measuring visuospatial ability (19). Other tests in the psychometrics battery were Free and Cued Selective Reminding Test (21) and a mental control test of the Wechsler Memory Scale (WMS) (17). The total number of items in the clinical and the psychometric battery that were analyzed can be found in Table 2.
Since 2005, the primary clinical and cognitive assessments of the WU Knight ADRC follow that of the National Alzheimer Coordinating Center Uniform Data Set (22),  which include standard definitions and diagnostic criteria for detection of dementia and its differential diagnosis (23). Prior to and after 2005 during more than 30 years of longitudinal follow-up, some of the tests and items were discontinued while others were added. More details regarding time of inclusion/discontinuation of all these items are summarized in Supplemental Table 1.

Other Covariates

Demographics such as baseline age, sex, and years of education were recorded at baseline. APOE genotyping was dichotomized into those with at least one copy of the E4 allele (E4 positive) vs. those without an E4 allele (E4 negative).

Selection of Informative items

Each item score was first converted into a binary scale, labeled as endorsement versus non-endorsement of the item, oriented in the way across all items that non-endorsement always indicates difficulty with function or cognition. For the items with a categorical score of three or more possible levels, a stringent dichotomization was applied. For instance, for the item named int562, participants were required to draw a triangle and then given a score as 0=‘correct’, 1=‘partially correct’ or 2=‘‘incorrect’. In our analyses, this item was dichotomized as endorsement with a score of 1= ‘correct’ answer and non-endorsement with a score of 0= ‘partially correct’ or ‘incorrect’ answer. Thus, a lower item score always corresponds to non-endorsement, i.e., difficulty with cognition, across the item pool.
The longitudinal trajectory of each individual item’s score was examined for its association with the age of symptomatic onset, defined as the age with the first occurrence of CDR>0 over longitudinal follow-up. For each individual, we first computed the age of symptomatic onset, either observed within the follow-up or right censored if an individual was never rated as having CDR>0 during the entire follow-up. For each individual and each item, we then computed the age of non-endorsement for each item, defined as the age when the item was not endorsed and given a score of zero.  An item’s score usually fluctuated between endorsement and non-endorsement for some time before finally stabilizing at non-endorsement throughout the remaining time. With this type of fluctuation under consideration, the item-specific age of non-endorsement for each individual was treated as interval censored. The left side of the interval was the first age in the follow-up when the item was not endorsed, and the right side of the interval was the age at the first occurrence of non-endorsement after which the item remained not endorsed throughout the remaining follow-up. If non-endorsement of an item was observed right at baseline, the left side of the age interval of non-endorsement was defined as zero. If an individual endorsed an item during the entire follow-up, the left side of the interval would be set as the age of last assessment and thus the age of non-endorsement became right-censored. For each item, the item-specific age of non-endorsement and the age at symptomatic onset were correlated across all participants. A good item was expected to render a significant concordance correlation between the age of non-endorsement and the age of symptomatic onset, as measured  by the  Kendall’s coefficient of concordance. This concordance correlation between the two interval censored variables was estimated through a bivariate smoothing of the joint density of the two variables (in logarithm scale) using a mixture of Gaussian densities fixed on a grid with weights determined by a penalized likelihood approach (24). Items with a significant concordance correlation (p≤0.05) were thereafter called informative for tracking early disease progression.

Estimating the longitudinal rate of functional and cognitive change

A participant’s overall functional and cognitive ability, as evaluated by the clinical protocol and the psychometric battery, respectively, was calculated by the corresponding composite of z-scores from multiple components. A z-score of a component test in the clinical protocol or the psychometric battery was calculated across all the items belonging to the test, using the baseline mean and standard deviation of the component test. The average of z-scores across all components in the clinical protocol was calculated as the composite score to represent the overall functional outcome, and the average of z-scores across all components in the psychometric battery was calculated to represent the overall cognitive outcome. We also computed the composite score combining the clinical protocol and the psychometric battery by averaging the z-scores across all the components in the two protocols. We derived these composite scores in three different ways using: 1) all items, 2) informative items selected using Kendall’s coefficient of concordance, and 3) the unselected non-informative items. Longitudinal changes in a participant’s overall functional and cognitive performance were usually very subtle prior to symptomatic onset (first time occurrence of CDR>0), but the decline of the scores accelerated afterwards. In recognition of this,  a piecewise random-intercept and random-slope linear mixed effects model was fit to the time segment from baseline to the symptomatic onset and then to the time segment after symptomatic onset for each of the three functional or cognitive composite scores.  The estimated slopes (i.e., the longitudinal rate of change) both prior to and after symptomatic onset are reported.  A negative slope indicates longitudinal cognitive decline.

Powering future clinical trials on early and preclinical AD

We further examined the sample sizes required to adequately power a future randomized clinical trial (RCT) using the proposed functional and cognitive composites from the informative items as the primary efficacy outcome variable. We considered both a prevention trial on asymptomatic individuals (i.e., prior to symptom onset) and a treatment trial on early symptomatic AD (encompassing both mild cognitive impairment due to AD and very mild AD dementia). For the prevention trial, cognitively normal individuals must first be identified as having elevated risk for symptomatic AD (i.e., preclinical AD). Cerebrospinal fluid (CSF) and neuroimaging biomarkers of AD now are used in major secondary prevention RCTs, including the Anti-Amyloid Treatment in Asymptomatic Alzheimer’s (A4) trial (25), the Dominantly Inherited Alzheimer Network-Trials Unit (DIAN-TU) trials (26), and the Alzheimer’s Prevention Initiative (API) trial (27), to identify persons with preclinical AD. However, the AD biomarkers may not identify all individuals who will eventually develop symptomatic AD. Hence, we decided to use the longitudinal functional and cognitive data prior to symptomatic onset from the 240 converters who became CDR>0 during follow up to power a future prevention trial, and the longitudinal data after symptomatic onset to power a treatment trial on early symptomatic AD by estimating the corresponding longitudinal rates of functional and cognitive composites. Specifically, the estimates from the first time segment prior to symptomatic onset in the piecewise linear mixed model were used as the placebo effect for the prevention trial, and those estimated from the second time segment after symptomatic onset were used as the effect in the placebo arm for the therapeutic trial on early symptomatic AD. The sample sizes for detecting a range of effect sizes (ES) for a novel treatment with 80% power were calculated using a standard normal test (28). For comparison purpose, similar sample size calculations for the treatment trial were also done on the functional and cognitive composites using all the items as well as using the non-informative items only.

Statistical analyses

All statistical analyses were implemented in SAS® (version 9.4, SAS Institute, Cary, NC).  All tests were two-sided and statistical significance was defined at the 5% level. Main analyses were also repeated on the converters who received an etiologic diagnosis of AD for comparison.



735 individuals at least 65 years old and cognitively normal (with CDR 0) at baseline were assessed annually up to 29 years of follow-up (mean follow-up=6.79 years, SD=5.55 years). Baseline characteristics of the participants are summarized in Table 1. Over the course of follow-up, 240 individuals converted from being cognitively normal (i.e., CDR = 0) at baseline to an early dementia with CDR ≥ 0.5, and were termed ‘converters’ thereafter. The average age of all participants is 74.30 years (SD=8.86 years) at baseline. 62.31% of the participants are female. Among a total of 562 items, including 374 in the cognitive battery and 188 in the clinical protocol, only 292 items were identified as informative, with 174 of them in the cognitive battery and 118 in the clinical protocol. Table 2 shows the number of informative items from each of the tests in the two instruments. Additional analyses restricted to the converters who received a etiologic diagnosis of AD resulted in largely consistent findings. Supplemental Table 1 in the Appendix lists all individual items that were found to be informative, i.e., predicting the age of symptom onset.

Table 1. Baseline Characteristics of the cohort (N = 735, mean/SD follow-up time = 6.79/5.55 y)

Table 1. Baseline Characteristics of the cohort (N = 735, mean/SD follow-up time = 6.79/5.55 y)

* The P-values for age, educ were from two-sided two-sample t-test and others from two-sided Fisher’s exact test.

Table 2. Numbers of total and informative items in the WU Knight ADRC clinical protocol and cognitive battery (tests in italic are no longer in use)

Table 2. Numbers of total and informative items in the WU Knight ADRC clinical protocol and cognitive battery (tests in italic are no longer in use)


For each composite, the piecewise linear mixed effects models resulted in two estimates to the slopes (i.e., longitudinal rates of change) along with their associated standard error (SE):  the preclinical rate of change (Slope 1) from baseline to symptomatic onset, and the rate of disease progression (Slope 2) after symptomatic onset, as presented in Table 3.  Results in Table 3 indicate that the estimated longitudinal rates of change in both the functional composite and cognitive composite, using the items identified as non-informative, are positive prior to symptomatic onset, indicating that asymptomatic individuals exhibited certain degree of learning over repeated testing of these items. When the informative items alone were used to form the functional and cognitive composites, however, their longitudinal rates of changes prior to the dementia symptom onset were negative. This implies that the informative items started to capture some of the subtle preclinical disease progression prior to symptomatic onset. Interestingly, the estimated longitudinal rates of change in both the functional composite and cognitive composites, using the items from the entire battery with both informative and non-informative ones, were also positive prior to symptomatic onset. This suggests that a large portion of the non-informative items in the battery easily overwhelmed the informative ones, leading to a collective learning effect over repeated testing for the entire batteries. Further, the longitudinal rates of change on both cognitive and functional composites after symptomatic onset using either all items in the item pool, or the informative items alone, or non-informative items alone, were all negative, suggesting that the entire clinical protocol and the psychometric battery effectively tracked progression following symptomatic onset.  More importantly, the longitudinal rates of change after symptomatic onset estimated using informative items only were larger in magnitude, both for cognitive and functional composites, than those estimated using all items, which were  in turn  larger (in magnitude) than the estimates using non-informative items.
Because cognitively normal individuals exhibited learning effects prior to symptomatic onset on the functional and cognitive composites using items identified as non-informative or using the entire item pool, these composites are hence of limited utility in designing future prevention trials on asymptomatic individuals. Given that the functional composite and cognitive composite, using only items identified as informative, did show decline longitudinally prior to the symptom onset, we explored the feasibility of using these two composites to power a future prevention trial of AD on asymptomatic individuals. In a hypothetical two-arm future prevention trial with a 1:1 sample size ratio and annual functional and cognitive assessments over 4 years, we calculated the total sample size with 80% power using the functional and cognitive composites from the informative items alone. Table 4a presents the results to detect a set of ES, expressed as percentages of improvement by the active treatment arm over the longitudinal rates of functional and cognitive progression in the placebo. After symptomatic onset, the larger magnitudes of the rate of progression using the functional and cognitive composites from the informative items (in comparison to those using either the entire item pool or the non-informative items alone) suggest that they may improve (i.e., reduce) the sample sizes for future therapeutic trials in early symptomatic AD. We again assumed a two-arm RCT treating participants with early symptomatic AD with a 1:1 sample size ratio between a novel therapeutic arm and a placebo arm and annual assessments over 4-year follow-up. Table 4b presents the total sample size required to detect a set of ES with 80% power using the functional and cognitive composites from the informative items alone, and for comparison, the total sample sizes required using the functional and cognitive composites from the non-informative items alone, as well as from the entire pools of items available. Given the same ES (in %)  to be detected and the same statistical power, the sample sizes using the functional composite of the informative items alone from the clinical protocol as the efficacy outcome are less than a third of those using the non-informative items. For the cognitive composite as the primary efficacy endpoint, using informative items in the cognitive battery also leads to a dramatic reduction in the sample size when compared to using the non-informative items.  For example, for an ES of 50%, the RCT with the functional endpoint can be adequately powered with a total 940 participants using informative items in the clinical protocol alone, a more than 70% reduction or about 7% reduction to the sample sizes using non-informative items alone (n=3174 participants) or the entire item pool (n=1010 participants), respectively.  For the cognitive endpoint with  an ES of 50%, the RCT can be adequately powered with a total of 1250 participants using informative items alone, only 11.5% of the sample size using non-informative items alone (n=10830 participants) and 66% of the sample size using the entire item pool (n=1884 participants).

Table 3. Estimates to the longitudinal rate of change prior to and after symptomatic onset, labeled as “Slope 1” and “Slope 2”, respectively, using all items in the item pool, informative items only, and non-informative items only

Table 3. Estimates to the longitudinal rate of change prior to and after symptomatic onset, labeled as “Slope 1” and “Slope 2”, respectively, using all items in the item pool, informative items only, and non-informative items only

*  Slope 1: slope before dementia onset with the first CDR>0 diagnosis; Slope 2: slope after symptomatic  onset.


Table 4 (a). Total sample sizes needed for a future prevention trial of AD to detect the effect sizes (% improvement) in the rate of disease progression prior to symptomatic onset, using the functional and cognitive composites formed by the informative items alone from the clinical protocol and the psychometric battery, respectively. (sample size results were only provided using the functional and cognitive composites from the informative items alone, because those using all items in the entire battery and the non-informative items alone yielded positive estimates to the longitudinal rate of change in the placebo arm, which indicated a learning effect over repeated testing).

Table 4 (a). Total sample sizes needed for a future prevention trial of AD to detect the effect sizes (% improvement) in the rate of disease progression prior to symptomatic onset, using the functional and cognitive composites formed by the informative items alone from the clinical protocol and the psychometric battery, respectively. (sample size results were only provided using the functional and cognitive composites from the informative items alone, because those using all items in the entire battery and the non-informative items alone yielded positive estimates to the longitudinal rate of change in the placebo arm, which indicated a learning effect over repeated testing).


Table 4 (b).Total sample sizes needed for future treatment RCTs in early symptomatic AD to detect the effect sizes (% improvement) in rate of disease progression after symptomatic onset, using the functional and cognitive composites formed by all items in the item pool (“All”), informative items alone (“Informative”), and non-informative items alone (“Non-informative”)

Table 4 (b).Total sample sizes needed for future treatment RCTs in early symptomatic AD to detect the effect sizes (% improvement) in rate of disease progression after symptomatic onset, using the functional and cognitive composites formed by all items in the item pool (“All”), informative items alone (“Informative”), and non-informative items alone (“Non-informative”)



A major paradigm shift in RCTs of investigational drugs for AD is the current focus on the preclinical or very early symptomatic stages. Major secondary prevention RCTs, including the A4 trial, the DIAN-TU trials, and the API trial, are currently ongoing. Given that the recently revised Food and Drug Administration (FDA) guidelines for RCTs for early AD mandate that treatments only be approved if they demonstrate cognitive and functional benefits, a well designed future RCT requires not only longitudinal cognitive and functional assessments, but also the linkage between the preclinical or early symptomatic stages and the rate of cognitive and functional decline in the placebo arm. A common challenge to all ongoing prevention RCTs on preclinical AD or treatment RCTs for early symptomatic AD is the optimum cognitive and functional endpoints that can best power the trials (29)]. The prevention or early treatment RCTs on AD mandate instruments that are much more sensitive and specific to the subtle early and preclinical longitudinal changes of AD than the existing ones (e.g., ADAS-cog[30]). As a matter of fact, several recent studies have failed to detect significant decline in the placebo groups of RCTs on mild cognitive impairment or even established mild to moderate AD populations with the existing cognitive and functional instruments. The situation will get even worse when it comes to designing primary preventive trials for AD. Because of the lack of highly reliable and well validated sensitive and specific cognitive and functional tests, it is difficult to establish a priori who in a population will ultimately develop AD symptoms and over what time frame. RCTs must therefore study a huge number of individuals for many years in order to guarantee that a significant number in any treatment arm will develop dementia symptoms so that meaningful statistical conclusions can be drawn. Such large and long duration studies are prohibitively costly and prone to high dropouts.
A major reason that currently used  clinical and cognitive instruments in AD research lack sensitivity and specificity to identify individuals who are at high risk of developing dementia symptoms is the ceiling and floor effects (3) as well as the learning effect due to the repeated administering of the same instruments. The fact that the A4, the DIAN-TU, and the API trials have all chosen to employ different cognitive endpoints highlights an urgent need to comprehensively analyze the longitudinal item level functional and cognitive data and to inform the RCTs with optimum cognitive and functional endpoints and adequate statistical power. We hence analyzed the longitudinal item level data from the clinical assessment protocol and the cognitive battery administered at the WU Knight ADRC to identify informative items that were most predictive of early symptomatic onset. We found that approximately half of the items among a total of 562 items were uninformative to predict symptomatic onset, likely because these items showed very little or only random changes over the preclinical time window prior to the onset of symptoms. Unsurprisingly, we found that, for both the cognitive and the functional composites that were formed using the non-informative items, the estimated longitudinal rate of change prior to symptomatic onset was positive, suggesting a learning effect during the preclinical stage of the disease. Importantly, we found that for both the cognitive and the functional composites using the informative items, the estimated longitudinal rate of change prior to symptomatic onset became negative, although the rate of changes for these composites using all items from the entire batteries was still positive. This suggests that the contamination of the batteries by non-informative items prevents adequate tracking of subtle preclinical disease progression. After symptomatic onset, as expected, we found that the functional and cognitive composites using the informative items alone rendered a larger rate of decline  (in magnitude) in comparison to the corresponding composites using only the non-informative items, as well as using the entire item pool with both informative and non-informative items.
Using these results to design a future prevention trial on asymptomatic individuals who will eventually develop symptomatic AD, we found that even with the informative items alone, the sample sizes required to adequately power such a trial remain formidable. For example, for detecting an effect size of 50% improvement as compared to placebo, the prevention RCT needs to enroll a total of 302322 participants using the functional endpoint, and a total of 6078 participants using the cognitive composite. These numbers are very hard to achieve, and much larger than the sample sizes currently estimated in some of the ongoing secondary prevention trials on elderly individuals. These large sample sizes also imply that, although the informative items from many standard cognitive tests are able to capture some of the cognitive decline prior to the symptom onset, the magnitude of decline captured is too small and the variation is too large. Hence, to best design future prevention trials on AD, completely new cognitive items and tests may need to be developed. These new items and tests should be designed in a way that will specifically target the cognitive traits and domains most vulnerable of very early change during the preclinical stage of AD, and minimize the learning effect over repeated administering. Significant resources are needed to develop such preclinical cognitive batteries and test their psychometric properties on individuals at high risk for preclinical AD.
For treatment trials on early symptomatic AD, however, sample sizes calculated using informative items only were dramatically smaller than those using non-informative items or using the entirety of the items in the batteries. In addition, these sample sizes are more feasible to achieve. For example, for detecting an effect size of 50%, the RCT with the functional composite can be adequately powered with a total 940 participants using informative items of the clinical protocol alone, and with a total of 1250 participants using the cognitive composite from informative items alone. When a meta-composite is used to combine both the cognitive and functional composites from the informatics items alone, the RCT can be adequately powered with 764 individuals with early symptomatic AD.
Major strengths of the study include a relatively large sample size of cognitively normal elderly individuals (at baseline) who were carefully characterized by annual clinical and cognitive assessments over a relatively long follow-up of up to 29 years. The relatively large number of individuals (n=240) who developed dementia symptoms during the follow-up allowed reasonably accurate estimates to the longitudinal rates of change in both function and cognition domains, both prior to (for designing prevention trials) and after symptom onset (for designing early treatment trials). Our longitudinal item analysis is also novel in the sense it directly correlated the item level changes in scores to the onset of symptoms. Realizing the vast variabilities in item-level scores over time, we analyzed the item scores over time as interval-censored variables, and used the Kendall’s coefficient of concordance to quantify the correlation between the interval censored variables and the age of symptomatic onset.
Limitations of this study include the convenience nature of the study sample, which were mostly restricted to the elder adult population in the St. Louis metropolitan area and may prevent the findings from being generalized to the more general population. Whereas the WU Knight ADRC clinical assessment protocols and cognitive battery are comprehensive, covering all major cognitive domains, they did not include some cognitive scales that are often used in the treatment trials of AD, such as the ADAS-cog, and hence the items from such scales cannot be evaluated for their utility in AD prevention trials.


Acknowledgement: The authors thank the WU Knight ADRC Clinical Core for the clinical and cognitive data used in this report.

Funding: This study was supported by National Institute on Aging (NIA) grant R01 AG034119 and R01 AG053550 (Dr. Xiong). Additional support was provided by NIA  P01 AG026276 (Dr. Morris), P50 AG005681 (Dr. Morris) and P01 AG0399131 (Dr. Morris).

Conflict of Interest Disclosure: The authors declare no competing interests.

Ethics approval and consent to the participant: The use of the WU Knight ADRC was consented by participants and approved by the Institutional Review Boards of Washington University School of Medicine.



1.    Price JL, Morris JC: Tangles and plaques in nondemented aging and «preclinical» Alzheimer’s disease. Ann Neurol 1999, 45(3):358-368.
2.    Sperling RA, Aisen PS, Beckett LA et al: Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011, 7(3):280-292.
3.    Sheehan B: Assessment scales in dementia. Ther Adv Neurol Disord 2012, 5(6):349-358.
4.    Twamley EW, Ropacki SA, Bondi MW: Neuropsychological and neuroimaging changes in preclinical Alzheimer’s disease. J Int Neuropsychol Soc 2006, 12(5):707-735.
5.    Morris JC: The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 1993, 43(11):2412-2414.
6.    JA SJaY: Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. Clinical Gerontology : A Guide to Assessment and Intervention. In. New York: The Haworth Press; 1986: 165-173.
7.    Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL: A new clinical scale for the staging of dementia. Br J Psychiatry 1982, 140:566-572.
8.    Pfeiffer E: A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc 1975, 23(10):433-441.
9.    Folstein MF, Folstein SE, McHugh PR: «Mini-mental state». A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975, 12(3):189-198.
10.    Katzman R, Brown T, Fuld P et al: Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry 1983, 140(6):734-739.
11.    Goodglass HaK, E: The Assessment of Aphasia and Related Disorders, 2nd Edition. Philadelphia: Lea & Febiger; 1972.
12.    Cummings JL: The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology 1997, 48(5 Suppl 6):S10-16.
13.    Pfeffer RI, Kurosaki TT, Harrah CH, Jr., Chance JM, Filos S: Measurement of functional activities in older adults in the community. J Gerontol 1982, 37(3):323-329.
14.    Ferman TJ, Smith GE, Boeve BF et al: DLB fluctuations – Specific features that reliably differentiate DLB from AD and normal aging. Neurology 2004, 62(2):181-187.
15.    Blessed G, Tomlinson, B.E., and  Roth, M.. : The association between quantitative measures of dementia and of senile change in the cerebral grey matter of elderly subjects. British Journal of Psychiatry 1968, 1114:797-811.
16.    Wechsler D: Wechsler Memory Scale-Revised. In. San Antonio, Texas: Psychological Corporation; 1987.
17.    Wechsler DaS, C.P: Wechsler Memory Scale. In. New York: Psychological Corporation; 1973.
18.    Benton AL: The Revised Visual Retention Test: Clinical and experimental applications. In. New York: Psychological Corporation; 1963.
19.    Goodglass HaK, E., : Boston Diagnostic Aphasia Examination Booklet. In., 3rd ed. edn. Philadelphia: Lea & Febiger; 1983.
20.    Wechsler D: Wechsler Adult Intelligence Scale. In. New York: Psychological Corporation; 1955.
21.    Grober E, Buschke H, Crystal H, Bang S, Dresner R: Screening for dementia by memory testing. Neurology 1988, 38(6):900-903.
22.    Morris JC, Weintraub S, Chui HC et al: The uniform data set (UDS): Clinical and cognitive variables and descriptive data from Alzheimer disease centers. Alz Dis Assoc Dis 2006, 20(4):210-216.
23.    American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4 edn. Washington DC: American Psychiatric Association; 2000.
24.    Bogaerts K, Lesaffre E: Estimating local and global measures of association for bivariate interval censored data with a smooth estimate of the density. Statistics in Medicine 2008, 27(28):5941-5955.
25.    Sperling RA, Rentz DM, Johnson KA et al: The A4 Study: Stopping AD Before Symptoms Begin? Science Translational Medicine 2014, 6(228).
26.    Mills SM, Mallmann J, Santacruz AM et al: Preclinical trials in autosomal dominant AD: Implementation of the DIAN-TU trial. Rev Neurol-France 2013, 169(10):737-743.
27.    Ayutyanont N, Langbaum JBS, Hendrix SB et al: The Alzheimer’s Prevention Initiative Composite Cognitive Test Score: Sample Size Estimates for the Evaluation of Preclinical Alzheimer’s Disease Treatments in Presenilin 1 E280A Mutation Carriers. J Clin Psychiat 2014, 75(6):652-660.
28.    Xiong C, Zhu, K, Yu, K, Miller, JP: Statistical Modeling in Biomedical Research: Longitudinal Data Analysis. In: Handbook of Statistics. Volume 27, edn. Edited by Rao CR MJ, Rao DC. London: Elsevier; 2007: 429-463.
29.    Ringman JM, Grill J, Rodriguez-Agudelo Y, Chavez M, Xiong C: Commentary on «a roadmap for the prevention of dementia II: Leon Thal Symposium 2008.» Prevention trials in persons at risk for dominantly inherited Alzheimer’s disease: opportunities and challenges. Alzheimers Dement 2009, 5(2):166-171.
30.    Mohs RC, Knopman D, Petersen RC et al: Development of cognitive instruments for use in clinical trials of antidementia drugs: additions to the Alzheimer’s Disease Assessment Scale that broaden its scope. The Alzheimer’s Disease Cooperative Study. Alzheimer Dis Assoc Disord 1997, 11 Suppl 2:S13-21.



K. Kahle-Wrobleski1, J.S. Andrews1, M. Belger2, W. Ye1, S. Gauthier3, D.M. Rentz4, D. Galasko5


1. Eli Lilly and Company, Indianapolis, IN, USA; 2. Lilly Research Centre, Windlesham, UK; 3. McGill Center for Studies in Aging, Douglas Mental Health Research Institute, Montreal, QC, Canada; 4. Harvard Medical School, Boston, MA, USA; 5. University of California San Diego, San Diego, CA, USA

Corresponding Author: Kristin Kahle-Wrobleski, PhD, Global Patient Outcomes and Real World Evidence, Eli Lilly and Company, Lilly Corporate Center, Indianapolis IN 46285, USA, Phone: 317-651-9881, Fax: 317-276-5791, Email:

J Prev Alz Dis 2017;4(2):72-80
Published online January 24, 2017,



Background: While functional loss forms part of the current diagnostic criteria used to identify dementia due to Alzheimer’s disease, the gradual and progressive nature of the disease makes it difficult to recognize clinically relevant signposts that could be helpful in making treatment and management decisions. Having previously observed a significant relationship between stages of functional dependence (the level of assistance patients require consequent to Alzheimer’s disease deficits, derived from the Alzheimer’s Disease Cooperative Study – Activities of Daily Living Scale) and cognitive severity, we investigated whether measures of functional dependence could be utilized to identify clinical milestones of Alzheimer’s disease progression.
OBJECTIVES: To describe the patterns of change in dependence over the course of 18 months in groups stratified according to cognitive Alzheimer’s disease dementia severity (determined using the Mini-Mental State Examination score) and to identify characteristics associated with patients showing worsening dependence (progressors) versus those showing no change or improvement (non-progressors).
DESIGN: Analysis of longitudinal data from the GERAS study.
SETTING: GERAS is an 18-month prospective, multicenter, naturalistic, observational cohort study reflecting the routine care of patients with Alzheimer’s disease in France, Germany, and the United Kingdom.
PARTICIPANTS: 1495 community-living patients, aged ≥55 years, diagnosed with probable Alzheimer’s disease dementia, and their caregivers.
MEASUREMENTS: Dependence levels, cognitive function, behavioral symptoms, caregiver burden, and cost were assessed at baseline and at 18 months.
RESULTS: Of 971 patients having both baseline and 18-month data, 42% (408) were progressors and 563 (58%) were non-progressors. This general pattern held for all three levels of baseline Alzheimer’s disease dementia severity – mild (Mini-Mental State Examination score 21–26), moderate (15–20) or moderately severe/severe (<15) – with 40–45% of each group identified as progressors and 55–60% as non-progressors. No baseline differences were seen between progressors and non-progressors in cognitive scores or behavioral symptoms, although progressors had significantly shorter times since diagnosis and showed milder functional impairment. Baseline factors predictive of increasing dependence over 18 months included more severe cognitive impairment, living with others, and having multiple caregivers. A higher level of initial dependence was associated with less risk of dependence progression. Total societal costs of care also increased with greater dependence.
CONCLUSIONS: In this large cohort, 42% of Alzheimer’s disease dementia patients at all levels of cognitive severity became more dependent within 18 months of observation while 58% did not progress. Dependence levels may be considered as meaningful interim clinical milestones that reflect Alzheimer’s disease-related functional deficits, although a time frame that extends beyond 18 months may be necessary to observe changes if used in clinical trials or other longitudinal studies. Recognition of predictors of greater dependence offers opportunities for intervention.

Key words: Alzheimer disease, dependence, observational study, ADCS-ADL, GERAS.



The core clinical criteria for making a diagnosis of dementia due to Alzheimer’s disease (AD) include declines in memory and other cognitive abilities, impairments in activities of daily living (ADL) and global functioning, as well as uncharacteristic changes in behavior or personality (1). As AD is a progressive and chronic illness, it is challenging to characterize clinically significant and discrete milestones of disease progression (2, 3). Nevertheless, defining some clinically relevant AD milestones may be useful to physicians, caregivers, and healthcare systems, as well as patients themselves, in making appropriate and timely treatment and management decisions.
Functional loss forms part of the current diagnostic criteria used to identify dementia due to AD (1); however, it is defined with less granularity than cognition, which includes standard domains such as memory and language (4). Yet, accurate measurement of functional loss can help determine the type, level, and costs associated with current and future care of a patient with AD dementia. Measures of meaningful clinical change in clinical trials of AD dementia have not been well established (5); hence, both the European Medicines Agency and the US Food and Drug Administration recommend that changes in cognition as well as function are needed for approval of a treatment for cognitive symptoms in AD dementia.
A variety of scales, developed to assess function in patients with AD dementia, have been incorporated into clinical trials and observational studies seeking to evaluate the impact of treatments or interventions. They include the Functional Activities Questionnaire; Disability Assessment for Dementia scale; Lawton scale; Global Deterioration Scale/Functional Assessment Staging system; Clinical Dementia Rating scale; Blessed Dementia Rating Scale; Amsterdam IADL (Instrumental Activities of Daily Living) Questionnaire; and the Alzheimer’s Disease Cooperative Study – Activities of Daily Living Scale (ADCS-ADL).
Assessing dependence, which refers to the level of assistance a patient requires due to AD dementia-related deficits, is an alternative, pragmatic approach to monitoring AD progression. As patients become more dependent, they face increased need for home assistance or possible institutionalization (or equivalent institutional care), along with higher healthcare, informal-care, and total-care costs (6–9). Increased dependence is accompanied by changes in family dynamics, and raises ethical and legal issues regarding financial and healthcare decision-making and guardianship (10). Dependence is thought to be influenced by – but different from – cognition, functioning, or behavior (6, 11, 12). It has been described as a distinct measurable component of dementia and an important determinant of AD-related disability (13). In the DADE Study, Jones et al. (8) observed significant associations between dependence, assessed by the Dependence Scale (DS) (13), and service use cost, patient quality of life (QoL), and caregiver perceived burden. They also suggested that, as a construct, dependence could be used to reflect the combined effects of cognitive functional, and behavioral changes seen in AD dementia into a more easily interpretable form.
Our previous work, a cross-sectional analysis of baseline data from the GERAS study (14), showed that dependence levels can be adapted from functional scales (15). Our results showed a significant relationship between assigned levels of dependence, derived from the ADCS-ADL score, and cognitive severity category in a large cohort of AD dementia patients receiving routine care. Importantly, with a greater assigned level of dependence, clinical and economic indicators showed a pattern of greater disease severity and higher costs. This work provided initial support for the use of dependence levels as appropriate interim clinical milestones that characterize the functional deficits associated with AD dementia. Our goal is to develop a system that can be used to stage disease progression in patients with AD dementia that is informative to both care providers and stakeholders, and that can also be used as a common metric in clinical trials or other research studies.
This report from the current longitudinal phase of the GERAS study expands the above-mentioned analysis with data from 18 months of evaluation. The objectives were to describe patterns of change in dependence within an 18-month period in groups stratified according to baseline cognitive AD dementia severity, to examine the correlation between dependence level and other outcome measures, and to identify characteristics that differentiated patients with AD dementia who showed worsening dependence (progressors) from those who showed no change or improved (non-progressors).



GERAS Study Design and Participants

The GERAS study was an 18-month, prospective, multicenter, naturalistic, observational, cohort study, which reflected the routine care of patients with AD dementia in France, Germany, and the United Kingdom (UK). The study was extended by a further 18-month follow-up period in France and Germany. The study design, methods, and baseline patient characteristics were previously described (14, 15). Patients were enrolled from October 2010 until September 2011. Patients and their caregivers were evaluated at baseline and during three care visits at 6-month intervals as part of their routine care. The study centers involved were mostly specialist secondary care clinics (“memory clinics”).
Briefly, patients were community-dwellers aged ≥55 years with probable AD dementia (National Institute of Neurological and Communicative Disorders and Stroke and Alzheimer’s Disease and Related Disorders Association [NINCDS-ADRDA] criteria) (16), a Mini-Mental State Examination (MMSE) score of ≤26 (17), and who presented within the normal course of care. Those with other potential causes of dementia were excluded from the study.
Each patient was also required to have a primary caregiver who agreed to participate in the study and to undertake responsibility for the patient for at least 6 months of the year. All patients (or their legal representatives) and caregivers gave written informed consent before entering the study, which was approved by ethical review boards in each country according to country-specific regulations.

Patient and Caregiver Data

Patients were stratified according to AD dementia severity, based in part on National Institute for Health and Care Excellence (NICE) guidance 217 (18), as having mild (MMSE score 21–26), moderate (15–20) or moderately severe/severe (<15) AD dementia. The study aimed to achieve approximately equal numbers of patients in the three AD dementia severity groups within each country. Functional ability was assessed using the caregiver-rated ADCS-ADL (19), with basic (BADL) and instrumental (complex) (IADL) sub-scales. Cognitive function was evaluated using the MMSE (17) in patients with mild or moderate AD dementia. The proxy version of the EuroQol-5 Dimension (EQ-5D) was used to obtain information from caregivers regarding patients’ health status (20). Behavioral symptoms were evaluated using the 12-item Neuropsychiatric Inventory (NPI-12) (21). Direct and indirect patient and caregiver resource use was measured using the Research Utilization in Dementia Scale (RUD) (22). To evaluate the impact on caregivers, total caregiver time and caregiver supervision time from the RUD were measured, as well as caregiver burden using the Zarit Burden Interview (ZBI) (23). Data were generally collected at baseline, 6, 12 and 18 months, although ADCS-ADL and EQ-5D data were measured only at baseline and 18 months.

Cost Estimations

For each country, monthly cost values were estimated by applying unit costs of services and products (2010 values) to the health and social care resource use data collected over the 18-month follow-up period. For resource use items, full details of the unit costs applied and their sources for each country have already been reported (14). All UK costs in pounds sterling (£) were converted to Euros (€) using the conversion rate £1=€1.1661, calculated as the average monthly exchange rate for 2010.
In this analysis, we looked at direct medical costs and total societal costs. Total societal costs were calculated by adding patient healthcare costs (including medications, hospitalizations, and outpatient visits), patient social care costs (including community care services, structural adaptations to the home, and extra financial support), and caregiver informal care costs (excluding caregiver direct healthcare costs).
Caregiver time, calculated as the number of hours for BADL and IADL, was capped at 24 hours/day; supervision time was excluded from cost calculations. The unit costs of caregiver time for working caregivers was the value of lost production time based on the national average wage per country population; for non-working caregivers, it was the value of lost leisure time based on 35% of the national average wage per country population.
The following imputation rules were applied for missing data: for institutionalized patients, mean monthly costs from the last visit were used for the period until institutionalization, then monthly costs of institutionalization were used from institutionalization up to 18 months; for patients who died, last observation carried forward (LOCF) was used such that costs from the last known visit were extrapolated up to the date of death (no costs after death were computed); for patients lost to follow-up, multiple imputation stratified by MMSE group and country was performed on missing costs. The factors used in the multiple imputation procedure were selected from those recently identified by Dodel et al. (24).

Categorizing Dependence Levels

Exploratory factor analysis of ADCS-ADL data was conducted in order to create subscales to aid the construction of dependence levels (4), details of which were described in our earlier analysis (15). In brief, baseline data suggested a 4-factor solution that included subdomains for specific competencies, including BADLs (eating, walking, toileting, bathing, grooming and dressing), domestic/household activities (choosing clothes, using the telephone, clearing the dishes, finding belongings, cooking, putting out the rubbish, using appliances), communication/engagement with the environment (watching television, paying attention to conversation, keeping appointments, talking about current events, reading, writing, performing hobbies), and outdoor activities (going out, shopping, paying, being left alone) (15).
The DS includes a scheme to derive levels of dependence based on the ability to perform ADLs (13), such as those represented within the ADCS-ADL (19). In an earlier study, we described six theory-driven assigned levels of dependence beginning with Level 0 (no care needed and completely independent), with dependence increasing incrementally over levels 1 through 5, which represents complete dependence, such as the needs of someone living in a nursing home (15). In the present analysis, the level of functional dependence, calculated from the ADCS-ADL, was determined for each patient at baseline and at 18 months.
A functional progressor was defined as someone whose change in functional dependence from baseline to the 18-month follow-up increased by 1 or more levels, while a non-progressor showed either no change or a decrease in 1 or more dependence levels.

Statistical Analysis

All patients and associated caregivers who provided informed consent and fulfilled the study entry criteria were included in the present analysis. All calculations were based on non-missing observations. Since dependence levels were derived from the ADCS-ADL, this analysis was limited to patients with ADCS-ADL data collected at baseline and at 18 months. Descriptive statistics (mean and standard deviation [SD] or frequency) for two groups (those for whom dependence level improved or did not change [non-progressors] and those for whom dependence level worsened [progressors]) were used to summarize baseline continuous variables (time since diagnosis, age, patient education, MMSE, ADL [total, basic, and instrumental] scores, and NPI total score and subscores) and categorical variables (gender, country, marital status, patient lives alone, any AD treatment, patient has falls, dependence level, and AD severity).
The distribution of patients at each level of dependence at baseline and 18 months for the overall patient sample and also according to cognitive severity at baseline were described. The percentage of patients with worsening dependence levels was also calculated.
Correlations between dependence levels and certain key outcome variables at baseline and at 18 months were examined using Pearson correlation coefficients. Generalized Linear Models (GLMs) with normal distribution and identity link function were used to examine whether changes in dependence level could predict changes in outcomes measures. The outcome measures were selected to reflect a range of outcomes that we expected would be related to level of dependence, including cognition (MMSE), caregiver burden (ZBI), overall caregiver time, NPI-12 total scores, patient-reported QoL (EQ-5D), patient direct medical costs, and total societal costs. GLMs were controlled for the core list of patient baseline characteristics (outcome measures, country, age, gender, cognitive severity, number of comorbidities, and total ADCS-ADL scores) and caregiver baseline characteristics (age, caregiver spouse [yes/no], and caregiver works for pay [yes/no]). In these models, changes in dependence levels and other outcome measures were considered as continuous variables to facilitate interpretation of the findings.
A stepwise logistic regression analysis was applied to the data to identify factors associated with patient progression versus non-progression. Baseline factors considered included: patient and caregiver age, gender, and country; patient cognitive MMSE severity (mild, moderate, moderately severe/severe), level of dependence, time since AD diagnosis, and NPI-12 scores (total and subdomains); indicator variables related to the patient (marital status, education, lives alone, comorbidities, receipt of AD medication, history of falling) and caregiver (lives with patient, is patient’s spouse, works for pay); number of caregivers other than the primary caregiver; NPI-12 caregiver distress score; and ZBI scores. ADCS-ADL scores were not included as the baseline dependence levels included in the model were calculated using these scores.
All data were analyzed using SAS software, version 9.2 (SAS Institute, Cary, North Carolina, USA).



Overall, 1532 patients and their primary caregivers were enrolled by 94 investigators. After excluding 35 patients who did not fulfill the study entry criteria and 2 patients following the baseline database lock, 1497 patients with probable AD dementia and caregivers (14) were included in the baseline analyses. Of these, 971 had both baseline and 18-month ADCS-ADL data. In the current analysis, 563/971 (57.98%) patients showed no change (479/971, 49.33%) or improvement (84/971, 8.65%) in dependence levels over 18 months (non-progressors) while 408/971 (42.02%) exhibited worsening of dependence levels (progressors).

Descriptive Statistics

A summary of baseline characteristics of patients and caregivers for progressors versus non-progressors is shown in Table 1. Full baseline statistics were previously reported (14, 15).
Non-progressors and progressors were similar in age, gender, years of education and marital status as well as caregiver characteristics. Most patients were married or lived with another person. Progressors had significantly shorter times since diagnosis than non-progressors (p=0.0096) and showed milder functional impairment at baseline, as indicated by total ADL (p=0.0002), BADL (p<0.0001), and IADL (p=0.0046) scores.
No differences were seen between progressors and non-progressors in cognitive (MMSE) scores or behavioral symptoms, as assessed by NPI-12 total scores or subscores. When patients were categorized according to AD dementia severity, the progressor and non-progressor groups contained approximately the same proportions of those with mild, moderate, and moderately severe/severe disease. Most patients in both groups had received treatment for AD dementia. No difference was observed between groups in the tendency to fall.
Some significant differences in dependence levels between groups were apparent at baseline (all p<0.0001). Progressors included more patients at lower dependence levels than non-progressors (e.g., 51.72% vs. 27.10% at dependence levels 0–2) and fewer at higher dependence levels (18.63% vs. 49.20% at levels 4–5, respectively).

Table 1. Mean baseline demographics (SD) for dependence non-progressors and progressors

Table 1. Mean baseline demographics (SD) for dependence non-progressors and progressors

Abbreviations: AD, Alzheimer’s Disease; ADL, activities of daily living; BADL, basic ADL; IADL, instrumental ADL; MMSE, Mini-Mental State Examination; NPI, Neuropsychiatric Inventory; SD, standard deviation. *Alzheimer’s Disease dementia severity based on MMSE scores: mild (21-26), moderate (15–20) and moderately severe/severe (<15).


Dependence Levels

Dependence level distributions for the entire cohort at baseline and 18 months are shown in Figure 1. Compared with baseline, at 18 months fewer patients in the overall patient sample were categorized as level 2 (requiring the equivalent of limited or informal home care services) and, correspondingly, more patients were found to require care equivalent to that provided by assisted living plus nursing support or placement into a nursing home (levels 4 and 5). A small proportion of patients at both time points showed either no impairments (level 0) or little impairment (level 1).

Figure 1. Dependence levels distribution at baseline and 18 months (entire cohort)

Figure 1. Dependence levels distribution at baseline and 18 months (entire cohort)

Patients with AD dementia (as determined by baseline MMSE cognitive scale) were assigned to one of six derived dependence levels, based on functional ability using the Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory (ADCS-ADL), at baseline and 18 months. The dependence levels were: level 0, no care needs; level 1, independent living with check-ins; level 2, care equivalent to limited or informal home care services; level 3, care equivalent to extensive home care with supervision or assisted living; level 4, care equivalent to that provided by assisted living plus nursing support; and level 5, care equivalent to a nursing home. Lower levels of dependence indicate better function.

Figure 2 illustrates the distribution of dependence levels according to AD dementia severity at baseline and at 18 months. The results demonstrate that both at baseline and after 18 months, patients with AD dementia at all levels of cognitive impairment had a range of dependence levels, although the distribution of patients at each dependence level differed according to severity group.


Figure 2. Dependence level distributions according to AD dementia severity at baseline and 18 months

Figure 2. Dependence level distributions according to AD dementia severity at baseline and 18 months

Patients were stratified according to disease severity by baseline cognitive function using the Mini-Mental State Examination (MMSE) as having mild (26–21), moderate (20–15) or moderately severe/severe (≤14) AD dementia. Patients with AD dementia were then assigned to one of six derived dependence levels, based on functional ability assessed using the Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory (ADCS-ADL) at baseline and again at 18 months. The dependence levels were: level 0, no care needs; level 1, independent living with check-ins; level 2, care equivalent to limited or informal home care services; level 3, care equivalent to extensive home care with supervision or assisted living; level 4, care equivalent to that provided by assisted living plus nursing support; and level 5, care equivalent to a nursing home. Note that no patients were determined to be at dependence level 0 or level 1 in the moderate AD group at 18 months or the moderately severe/severe group at baseline and at 18 months.


At baseline, 49.64% of those diagnosed with mild AD dementia required the equivalent of limited or informal home care services (dependence level 2), whereas 42.69% required at least the equivalent of extensive home care services with supervision or assisted living plus nursing support (levels 3–4). Less than 7% (6.72%) of patients with mild AD required no or little assistance (level 0 or 1) and 0.96% required the equivalent of nursing home care (level 5).
After 18 months, fewer patients (31.18%) with mild AD dementia showed the equivalent of level 2 dependence compared with baseline, while 58.75% required higher levels of assistance (levels 3–4). Almost 6% (5.76%) required the equivalent of nursing home care.
In those with moderate AD dementia, the proportion of patients requiring the equivalent of assisted living plus nursing support or placement in a nursing home (levels 4 and 5) rose from 34.50% at baseline to 57.51% after 18 months. The majority of those with severe disease needed high levels of assistance at baseline (69.71%), with a greater proportion requiring high levels of assistance at 18 months (86.72%).
The pattern of relative progression versus non-progression in dependence level observed in the overall cohort (42.02% progression/57.98% non-progression) was similar in all groups of AD dementia severity. For those with mild AD dementia (n=417), 39.57% progressed compared with baseline whereas 60.43% did not; corresponding values in the moderate AD dementia (n=313) group were 44.73% and 55.27%. Patients with severe AD dementia (n=241) manifested high levels of dependence at baseline and this pattern became more prominent at 18 months, with 42.74% of patients becoming more functionally dependent and 57.26% not progressing during the 18-month follow-up. Of the 408 progressors, 294 (72.06%) worsened by one level, 98 (24.02%) worsened by two levels and 16 (3.92%) worsened by more than two levels.

Relationship Between Dependence Levels and Other Outcome Measures

In unadjusted correlations using Pearson correlation coefficients, a number of patient clinical outcome measures correlated significantly (all p-values <0.05) with levels of dependence at baseline and at 18 months. These included negative correlations for cognition (MMSE total score) (r=−0.50 at baseline, r=−0.57 at 18 months) and patient QoL (r=−0.37, −0.46, respectively) and a positive correlation for patient neuropsychiatric disturbance (r=0.33, 0.34, respectively). From the caregiver’s perspective, increased dependence correlated with greater caregiver time (r=0.49, 0.44, respectively) and greater overall ZBI burden (r=0.36, 0.30, respectively). Although there was no statistically significant correlation with patient direct medical costs at baseline (r=0.05), dependence was positively correlated with these costs at 18 months (r=0.13), and total societal costs increased with greater dependence (r=0.35, 0.43, respectively).
Table 2 shows the degree to which a change in one dependence level is strongly associated with changes in other clinical and outcome variables. Results of the GLM regression analyses controlling for a core list of covariates for each outcome suggest that positive linear relationships between change in dependence level and change in ZBI, caregiver time, NPI-12 total score, medical costs, and total societal costs, and negative linear relationships between change in dependence level and change in MMSE and EQ-5D are statistically significant.

Table 2. Relationship between change in each outcome measure versus change in dependence levels from GLMs

Table 2. Relationship between change in each outcome measure versus change in dependence levels from GLMs

Abbreviations: AD, Alzheimer’s Disease; ADL, activities of daily living; EQ-5D, EuroQol-5 Dimension; GLM, generalized linear model; MMSE, Mini-Mental State Examination; NPI, Neuropsychiatric Inventory; ZBI, Zarit Burden Interview. Note: Δ = change from baseline to 18 months. The changes in outcomes measures from baseline to 18 months were fitted using GLM models with normal distribution with identity link function. All models were adjusted for the core list of covariates. The model is listed as follows: change in outcome measure = baseline outcome measure + change in dependence level + the core list of covariates. The core list of covariates includes AD dementia severity, country, patient age, patient gender, number of comorbidities, baseline ADL total score, caregiver age, caregiver is spouse (yes/no), and caregiver works for pay (yes/no).

Factors Associated with Progressors versus Non-progressors

Results of the logistic regression analysis shown in Figure 3 suggest that several patient baseline factors were significantly predictive of increasing dependence. They included greater cognitive impairment, such that patients with mild or moderate cognitive impairment were less likely than more severely impaired patients to have progressed a minimum of one dependency level over 18 months. However, having higher initial levels of dependence (e.g., level 2 vs. level 1) was associated with less risk of dependence progression over 18 months, as indicated by an odds ratio of 0.27 (95% CI 0.22, 0.33). Other potential predictors of increasing dependency included the patient living with someone else compared to living alone, and having multiple caregivers. Age, education, gender, other comorbidities, neuropsychiatric symptoms, and elevated caregiver burden were not found to be predictive of dependence progression.

Figure 3. Association between baseline factors and dependence progression in patients with AD dementia

Figure 3. Association between baseline factors and dependence progression in patients with AD dementia

*Note that factors with OR<1 indicate less likelihood of dependence progression, those with OR=1 have no association with dependence progression, and those with OR>1 are predictive of dependence progression.



Results of this large, observational cohort of patients with AD dementia at different stages of cognitive impairment reveal the heterogeneity of disease progression when it comes to functional abilities. After 18 months, even those with mild AD dementia required varied levels of support, with almost 60% needing the equivalent of extensive home care services with supervision/assisted living/nursing support and almost 6% the equivalent of nursing home care. As would be expected with AD disease progression, a general upward shift in dependence levels was observed over 18 months for all levels of AD dementia severity. However, individual variation was apparent and common. At all three levels of baseline AD dementia severity (mild, moderate and moderately severe/severe), over an 18-month period between 55% and 60% of AD dementia patients manifested no change or an improvement in dependence while 40% to 45% became more dependent.
Identifying early in their disease those AD dementia patients who will manifest functional deficits and require assistance is critical for patients, family members, and healthcare systems. Having identified progressors and non-progressors in our cohort, we then went back to baseline data to examine possible early differentiating factors. Interestingly, no differences were seen between progressors and non-progressors in cognition (MMSE scores) or behavioral symptoms, tendency to fall, whether they were receiving treatment, or caregiver characteristics. With logistic regression we identified baseline factors associated with dependence progression, including greater cognitive impairment, having multiple caregivers, and living with another person. Conversely, having higher initial levels of dependence was associated with less risk of dependence progression. These seemingly contradictory findings may be explained by disease and/or measurement issues. The association between greater cognitive impairment at baseline and higher likelihood of increased dependence is consistent with prior research showing a pattern of cognitive decline preceding functional decline (25). The present analyses are also consistent with the expectation that higher dependence levels should be associated with other indicators of functional impairment, as evidenced by the finding that progressors were more likely to have multiple caregivers or live with another person. However, the categorical nature of dependence levels means that those at higher levels of dependence at baseline may not be as likely to progress due to ceiling effects.
A change of one dependence level can have practical consequences. For example, we estimate that a patient who declines one level in dependence requires 40 more hours of caregiving time per month, and incurs €35.85 per month more in medical costs and €299.57 per month more in total societal costs. Recently, Darba and Kaskens (9) used the DS to evaluate 343 patients with AD dementia in Spain. They found that with each additional 1-point increase in DS, there was a 13.5% increase in direct medical care costs, a 25.3% increase in social care costs, and a 214.7% increase in informal care costs over 6 months. These data highlight the social and economic ramifications of deteriorating independence.
Selecting which outcome(s) to evaluate can be a challenge for an illness with as global an impact as AD dementia. Outcomes of interest reflect the needs and goals of patients, families, caregivers, clinicians, researchers, payers, and healthcare agencies. While cognition is traditionally thought to be the chief variable of interest in AD dementia, and is the hallmark impairment of the disease (26), assessment of functional abilities, behavior, caregiver burden, QoL, resource utilization, and costs also provide worthwhile information, which can be used to stage AD dementia and assess disease impact (27). Once a relevant outcome(s) of interest is chosen, it is then important to select an instrument that will provide valid and meaningful data (27, 28).
Our results support the suggestions of Zhu et al. (6, 29) that dependence is a distinct component of disability in AD dementia, and that assessment of patient function and dependence provides information not available from the MMSE or other cognitive measures. Further, increases in dependence are associated with other indicators of increasing disease severity such as more caregiver time, higher caregiver burden, and higher care costs. The additional association between cognition and dependence suggests that dependence is a useful summary measure of AD severity.
We believe that dependence levels may be considered as interim clinical milestones that reflect AD dementia-related functional deficits and can be a useful metric to monitor disease course. Measuring dependence levels may offer valuable information for tracking, managing, and treating patients with AD dementia. These assessments may be particularly informative when characterizing such patients in the mildest stage of disease, a time when discrete clinical milestones are difficult to ascertain. Whether measurable shifts in dependency can occur within 18 months with enough sensitivity to make it an appropriate 18-month endpoint for phase 3 clinical trials remains to be determined, but this metric may prove to be useful for longer-term trials, trials enriched with a population more likely to progress, or observational studies.

Strengths of the Study

The GERAS study included a large population of patients representative of the AD dementia continuum. The aim of the GERAS study was to address some of the limitations of previous cost studies by using well-established, standardized methods for assessing resource use and caregiver time over a longer follow-up period. The study also provides unique information on the societal costs of AD dementia in community-dwelling patients, both across different severity levels and between countries (14).

Study Limitations

One limitation of GERAS is that the study sample did not include the entire spectrum of functional impairment. Specifically, enrolled patients were required to be community dwelling, thus excluding those who were severely impaired. In addition, patients with cognitive impairment who did not yet show signs of dementia (prodromal AD) were also excluded. This limits the external validity of the results. The frequency of users of AD medications may be overestimated because study centers were mostly specialist clinics (14).
Another limitation is that although about 40% of the patients in the moderately severe/severe group had an MMSE score of <10 (14), the cut-offs used in the GERAS study may bias the moderately severe/severe AD dementia group toward those with less-severe symptoms. Additionally, because observed discontinuation rates tend, in general, to be higher in those with greater cognitive and functional disabilities, dependence progression may have been underestimated in our study, which focused only on completers. Also, differences in patient characteristics between AD dementia severity groups in each country may have confounding effects on resource utilization and associated costs, but these have not been taken into account. Lastly, cultural differences between countries regarding the care of patients with AD dementia, as well as differences in healthcare systems, mean that the resource utilization and cost data from the three countries in the GERAS study (France, Germany, and the UK) may limit the generalizability of the findings (14). It should also be noted that patients were enrolled in GERAS based on NINCDS-ADRDA clinical criteria; however, no biomarker information or genotype status was captured. Further research is warranted to evaluate the power of these factors to predict progression status.



Understanding the types of changes that patients with AD dementia undergo can help determine what sorts of milestones may be suitable for clinical use and/or use in clinical trials and observational studies. Our study demonstrated that dependence levels may provide unique information on clinical progression beyond what is captured with a cognitive measure. We have also shown the pragmatic consequences of increasing dependence on patients, caregivers, and costs. Identification of several baseline factors associated with functional deterioration opens up possibilities for early intervention.


Funding: This study was supported by Eli Lilly and Company.

Acknowledgments: The authors would like to acknowledge Amy Rothman Schonfeld, Gill Gummer and Caroline Spencer (Rx Communications, Mold, UK) for assistance with the preparation of this article, funded by Eli Lilly and Company. The authors would also like to thank Dr. Yaakov Stern for sharing his insights and expertise during the course of this research.

Disclosures: KK-W, JSA, MB and WY are all employees and minor shareholders of Eli Lilly and Company. SG has received clinical trial support from Eli Lilly and Company, Roche, TauRx and Lundbeck Pharmaceuticals, was a DSMB member for ADCS, ATRI, API and Eisai, was a scientific advisor for AbbVie, Advantage, Alzheon, Axovant, Boehringer-Ingelheim, Firalis, Heptares, IntelGen, Kalgene, Eli Lilly and Company, Lundbeck Pharmaceuticals, Novartis, Otsuka, Servier, Sanofi, Schwabe, Takeda, TauRx, TVM Capital and Roche. DMR has served as a consultant for Eli Lilly and Company, Biogen Idec, Lundbeck Pharmaceuticals, and serves as a member of the Scientific Advisory Board for Neurotrack. DG has received personal fees from Biomed Central, vTv Therapeutics, Janssen Immunotherapy, Prothena Corporation, the Michael J Fox Foundation, and has received grants from California Institute for Regenerative Medicine and the Michael J Fox Foundation.



1.    McKhann G, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7:263-269.
2.    Mitchell AJ, Meader N, Pentzek M. Clinical recognition of dementia and cognitive impairment in primary care: a meta-analysis of physician accuracy. Acta Psychiatr Scand 2011;124:165-183.
3.    Stella F, Forlenza OV, Laks J, et al. Caregiver report versus clinician impression: disagreements in rating neuropsychiatric symptoms in Alzheimer’s disease patients. Int J Geriatr Psychiatry 2015 9;30:1230-1237.
4.    Kahle-Wrobleski K, Coley N, Lepage B, et al., for the PLASA/DSA Group. Understanding the complexities of functional ability in Alzheimer’s disease: more than just basic and instrumental factors. Curr Alzheimers Res 2014;11:357-366.
5.    Molnar FJ, Man-Son-Hing M, Fergusson D. Systematic review of measures of clinical significance employed in randomized controlled trials of drugs for dementia. J Am Geriatr Soc 2009;57:536-546.
6.    Zhu CW, Leibman C, Townsend R, et al. Bridging from clinical endpoints to estimates of treatment value for external decision makers. J Nutr Health Aging 2009;13:256-259.
7.    Scherer RK, Scarmeas N, Brandt J, et al. The relation of patient dependence to home health aide use in Alzheimer’s disease. J Gerontol A Biol Sci Med Sci 2008;63:1005-1009.
8.    Jones RW, Romeo R, Trigg R, et al., for the DADE Investigator Group. Dependence in Alzheimer’s disease and service use costs, quality of life, and caregiver burden: The DADE study. Alzheimers Dement 2015 11:280-290. DOI: 10.1016/j.jalz.2014.03.001.
9.    Darba J, Kaskens L. Relationship between patient dependence and direct medical- social-, indirect – and informal-care costs in Spain. ClinicoEcon Outcomes Res 2015;7:387-395.
10.    Gauthier S, Leuzy A, Racine E, et al. Diagnosis and management of Alzheimer’s disease: past, present and future ethical issues. Prog Neurobiol 2013;110:102-113.
11.    Lenderking MR, Wyrwich KW, Stolar M, et al. Reliability, validity, and interpretation of the dependence scale in mild to moderately severe Alzheimer’s disease. Am J Azheimers Dis Other Dement 2013; 28:738-749.
12.    Brickman AM, Riba A, Bell K, et al. Longitudinal assessment of patient dependence in Alzheimer disease. Arch Neurol 2002;59:1304-1308.
13.    Stern Y, Albert SM, Sano M, et al. Assessing patient dependence in Alzheimer’s disease. J Gerontol 1994;49:M216-222.
14.    Wimo A, Reed CC, Dodel R, et al. The GERAS study: a prospective observational study of costs and resource use in community dwellers with Alzheimer’s disease in three European countries – study design and baseline findings. J Alzheimers Dis 2013;36:385-399.
15.    Kahle-Wrobleski K, Andrews JS, Belger M, et al. Clinical and economic characteristics of milestones along the continuum of Alzheimer’s disease: Transforming functional scores into levels of dependence. J Prev Alzheimers Dis Feb 2015;2:115-120.
16.    McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984;34:939-944.
17.    Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198.
18.    National Institute of Health and Clinical Excellence. NICE technology appraisal guideline 217. Donepezil, galantamine, rivastigmine and memantine for the treatment of Alzheimer’s disease. March 2011. Accessed March 19, 2015.
19.    Galasko D, Bennett D, Sano M, et al. An inventory to assess activities of daily living for clinical trials in Alzheimer’s disease. The Alzheimer’s Disease Cooperative Study. Alzheimer Dis Assoc Disord 1997;11(Suppl 2):S33-S39.
20.    The EuroQol Group. Euro-Qol – a new facility for the measurement of health-related quality of life. Health Policy 1990;16:199-208.
21.    Cummings JL, Mega M, Gray K, et al. The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology 1997;48(Suppl 6):S10-S16.
22.    Wimo A, Wetterholm AL, Mastey V, et al. Evaluation of resource utilization and caregiver time in anti-dementia drug trials: a quantitative battery. In Wimo A, Jonsson B, Karlsson G, (eds). The Health Economics of Dementia, 1998. John Wiley and Sons, London: pp. 465-499.
23.    Zarit SH, Reever KE, Bach-Peterson J. Relatives of the impaired elderly: correlates of feelings of burden. Gerontologist 1980;20:649-655.
24.    Dodel R, Belger M, Reed C, et al. Determinants of societal costs in Alzheimer’s disease: GERAS study baseline results. Alzheimer’s Dement 2015:11:933-945.
25.    Liu-Seifert H, Siemers E, Sundell K, et al. Cognitive and functional decline and their relationship in patients with mild Alzheimer’s dementia. J Alzheimers Dis 2015;43:949-955. doi: 10.3233/JAD-140792.
26.    Vellas B, Carrillo MC, Sampaio C, et al, EU/US/CTAD Task Force Members. Designing drug trials for Alzheimer’s disease: what we have learned from the release of the phase III antibody trials: a report from the EU/US/CTAD Task Force. Alzheimer’s Dement 2013;9:438-444.
27.    Green Park Collaborative. Evidence Guidance Document. Design of Clinical Studies of Pharmacologic Therapies for Alzheimer’s Disease. 11/04/13.
28.    Doody RS, Thomas RG, Farlow M, et al., for the Alzheimer’s Disease Cooperative Study Steering Committee; and Siemers E, Liu-Seifert H, and Mohs for the Solanezumab Study Group. Phase 3 trials of solanezumab for mild-to-moderate Alzheimer’s disease. N Engl J Med 2014;370:311-321.
29.    Zhu CW, Leibman C, McLaughlin T, et al. The effects of patient function and dependence on costs of care in Alzheimer’s Disease. J Am Geriatr Soc 2008;56:1147-1503.



L. Ruffini1, F. Lauretani2, M. Scarlattei1, A. Ticinesi2, T. Meschi2, C. Ghetti4, G. Serreli1, M. Maggio3, P. Caffarra5


1. Nuclear Medicine Unit, Academic Hospital of Parma, Parma, Italy; 2. Geriatric Rehabilitation Department, Academic Hospital of Parma, Parma, Italy; 3. Department of Clinical and Experimental Medicine, Section of Geriatrics, University of Parma; 4. Medical physics, Academic Hospital of Parma, Parma, Italy; 5. Department of Neuroscience, University of Parma, Parma, Italy

Corresponding Author: Livia Ruffini, MD, Nuclear Medicine Unit, Academic Hospital of Parma, Parma, Italy, E-mail:

J Prev Alz Dis 2016;3(3):127-132
Published online May 31, 2016,



A significant progress has been made in the understanding of the neurobiology of Alzheimer’s disease. The post-mortem studies are the gold standard for a correct histopathological diagnosis, contributing to clarify the correlation with cognitive, behavioral and extra-cognitive domains. However, the relationship between pathological staging and clinical involvement remains challenging.
Neuroimaging, including positron emission tomography (PET) and magnetic resonance, could help to bridge the gap by providing in vivo information about disease staging. In the last decade, advances in the sensitivity of neuroimaging techniques have been described, in order to accurately distinguish AD from other causes of dementia.
Fluorodeoxyglucose-traced PET (FDG-PET) is able to measure cerebral metabolic rates of glucose, a proxy for neuronal activity, theoretically allowing detection of AD. Many studies have shown that this technique could be used in early AD, where reduced metabolic activity correlates with disease progression and predicts histopathological diagnosis. More recently, molecular imaging has made possible to detect brain deposition of histopathology-confirmed neuritic β-amyloid plaques (Aβ) using PET. Although Aβ plaques are one of the defining pathological features of AD, elevated levels of Aβ can be detected with this technique also in older individuals without dementia. This raises doubts on the utility of Aβ PET to identify persons at high risk of developing AD.
In the present case-series, we sought to combine metabolic information (from FDG-PET) and amyloid plaque load (from Aβ PET) in order to correctly distinguish AD from other forms of dementia. By selecting patients with Aβ PET + / FDG-PET + and Aβ PET – / FDG-PET +, we propose an integrated algorithm of clinical and molecular imaging information to better define type of dementia in older persons.


Key words: Molecular imaging, amyloid PET, 18F-FDG PET, Alzheimer disease.



Beta amyloid deposition in cerebral tissue is one of the major histopathological hallmarks of Alzheimer’s disease (AD) and the development of radiotracers to visualize β-amyloid plaques in vivo has been proposed as a biomarker of AD.
The Cochrane Library recent reports examined the utilization of the (11)C-PIB-PET for the early diagnosis of AD and other types of dementia in adult persons (1). The conversion from mild cognitive impairment (MCI) to AD was evaluated in nine of the examined studies. Of the 274 participants included in the meta-analysis, 112 developed Alzheimer’s dementia. Sensitivity ranged between 83% and 100% while specificity between 46% and 88%. They concluded that given the heterogeneity in the conduct and interpretation of the test and the lack of defined thresholds for determination of test positivity, its routine use in clinical practice cannot be recommended. However, a recent meta-analysis (2), showed that, among participants with dementia, the prevalence of amyloid positivity was associated with clinical diagnosis, age, and APOE genotype. These findings indicate the potential clinical utility of amyloid imaging for differential diagnosis in early-onset dementia. This technique can also support the clinical diagnosis in subjects who are negative for APOE ε4 status and older than 70 years (2). Moreover, among persons without dementia, the prevalence of cerebral amyloid pathology as determined by positron emission tomography (PET) or cerebrospinal fluid findings was associated with age, APOE genotype, and presence of cognitive impairment. These findings suggest a 20- to 30-year interval between first development of amyloid positivity and the onset of dementia (3). A robust difference was also found in the amyloid PET tracer PIB retention in cerebral tissue between AD patients and healthy controls (HC).
However, the utilization of FDG-PET for the diagnosis of AD is still debated. The variability of specificity values and the absence of defined thresholds for determination of test positivity among different studies is still a major issue (1). Thus, current evidence does not support the routine use of FDG-PET scans in clinical practice in persons with MCI (4). FDG PET scan is also an expensive test, therefore it is important that its accuracy is clearly demonstrated and its protocol adequately standardized before it can be widely used. More uniform approaches to thresholds sufficient sample sizes are needed in future studies to make definitive recommendations (5).
In the present case-series, we combined metabolic (from FDG-PET) and amyloid plaque load (from Aβ PET) information in older individuals with dementia, in order to: a) identify AD from other forms of dementia, b) select patients with Aβ PET + / FDG-PET + and Aβ PET – / FDG-PET + and c) propose an algorithm to integrate clinical and imaging information to better define different types of dementia in older persons.



Patients were evaluated at the Center for Cognitive Disorders, AUSL of Parma, while the PET scans were performed at the Nuclear Medicine Department of University-Hospital of Parma (Parma, Italy). Patients were first evaluated by neurologists with a standard clinical evaluation (6) and then referred to a neuropsychologist with long-term experience in clinical and experimental neuropsychology of degenerative diseases (7). The diagnosis of mild cognitive impairment (MCI) (8) and dementia syndrome, independent of etiology, were established using a standard evaluation protocol based on the 1984 NINCDS–ADRDA criteria (9). The neuropsychological battery included test assessing abstract reasoning, memory, attention, language, praxis and visuo-perceptive functions. All tests included in the neuropsychology battery have normal ranges and cut-offs available for the Italian population (10-12). Depressive  symptoms were assessed by the 15-item Geriatric Depression Scale (GDS-15) which is a widely used screening instrument for depressive symptoms in the elderly. The 15-item Geriatric Depression Scale (GDS-15) detects changes in depressive symptoms after a major negative life event (13). All patients underwent a brain MRI or CT scan in the previous 3 months. Chronic drug treatment were recorded. Missing data were integrated by checking in original clinical sheet. The data were treated in agreement with Italian law for the privacy guaranty. The study received approval by the institutional review board, and all patients signed informed consent.

PET/CT imaging

Both 18F-FDG and amyloid PET scans were performed using a whole-body hybrid system Discovery IQ (GE Healthcare) operating in three-dimensional detection mode, in two different days. Head holder was used to restrict patient and head movement was checked on a regular basis.


After an overnight fast, 200 MBq of 18F-FDG  were administered intravenously in a quiet, dimly lit examination room. The brain PET/CT recording was started 30 min after tracer injection. During the 30-minute uptake period, participants were left undisturbed in a darkened room and instructed to rest quietly without activity with their eyes close, as commonly recommended.
The brain CT was first recorded to provide the attenuation correction map (140 kV, 25 mA, 512×512 matrix, 3.75-mm slice thickness, scan Type Helical full 0.8 s, No of images 79, Rec Fov 30 cm, recon type standard). CT was immediately followed by a 3D-PET recording during a 10-min period (FOV 30 cm, recon type QCHD and VPHD, matrix size 256×256).
Quantitative analysis was performed using the SPM5 software implemented in Matlab R2014a (12). The patient PET dataset was spatially normalized using the SPM5 PET template and smoothed with a Gaussian filter of 8 mm FWHM. Differences in CMRglc (patient  vs. normal) were assessed on a voxel-by-voxel basis, using a paired t-test. The results were displayed on the Tailarach atlas.

18F-fluorbetaben PET

All cerebral emission scans began 90 minutes after a mean injection of 4 MBq/kg weight (240-360 MBq) of 18F-fluorbetaben. For each subject, 10-minute frames were acquired to ensure movement-free image acquisition. All PET sinograms were reconstructed with a 3-D iterative algorithm, with corrections for randomness, scatter, photon attenuation and decay, which produced images with an isotropic voxel of 2×2×2 mm and a spatial resolution of approximately 5-mm full-width at a half-maximum at the field of view center.
PET images were assessed visually by two trained, independent readers blinded each other with a  previously described technique (14-15).
Visual assessment of florbetaben PET images was performed by a three-grade scoring system (RCTU – regional cortical tracer uptake) comparing the activity in cortical gray matter (frontal cortex, posterior cingulate cortex/precuneus; lateral temporal cortex, and parietal cortex) with activity in adjacent cortical white matter. The RCTU scores (1 = no binding, 2 = minor binding, and 3 = pronounced binding) were then condensed into a single three-grade scoring system for each PET scan, the BAPL score: 1 = no b-amyloid load, 2 = minor b-amyloid load, 3 = significant b-amyloid load, with the resulting scores condensed into a binary interpretation (score 1 = negative; score 2 or 3 = positive).


The complexity of the diagnosis of dementia in older persons: three suggestive case reports

First Case (probable AD)

A 74-year-old woman with memory complaints was evaluated. Her medical history was characterized only by hypertensive cardiomyopathy.  At the initial cognitive assessment, she showed a nearly normal Mini Mental Examination State (MMSE 23.4/30 adjusted for age and education). Cerebral MRI showed mild cerebral atrophy and slight hippocampal atrophy. The neurological examination was free of neurological signs. At the Geriatric Depression Scale (GDS) she scored 1/15 and her Neuropsychiatric Inventory score was 0/144. The neuropsychological evaluation showed deficits in immediate and cue total verbal recall and constructional apraxia. The FDG-PET scans showed a significantly hypo-metabolism involving frontal, temporal and parietal lobe (Figure 1a). The amyloid PET was then performed, indicating a significant presence of amyloid plaques (BAPL 3) (Figure 1b). The diagnosis of probable Alzheimer’s disease was thus established.

Figure 1a. Metabolic PET. Tracer: 18F-FDG

a) Quantitative assessment of FDG uptake: SPM results on a voxel-by-voxel basis; b) Results of the analysis displayed on the Tailarach atlas


Figure 1b. Amyloid PET (axial sections). Tracer: 18F-Fluorbetaben


Second Case (probable FTD)

A 75-year-old woman with “hesitant and agrammatical” language was evaluated. Her medical history was null. Her cerebral CT scan was normal. At the initial cognitive assessment, she showed a pathological Mini Mental Examination State (MMSE 21.4/30 adjusted for age and education). The neurological examination was normal. At the cognitive evaluation she was found having both memory and extra-amnesic disorders relating to consistent difficulties on speech production. The FDG-PET scans showed a significantly reduced metabolic activity in many cerebral areas, including prevalently the frontal lobe (Figure 2a). The amyloid PET was negative for brain amyloid deposition  (Figure 2b). The diagnosis of primary progressive aphasia due to probable Fronto-Temporal Dementia was established.

Figure 2a. Metabolic PET. Tracer: 18F-FDG

Quantitative assessment of FDG uptake: SPM results on a voxel-by-voxel basis.


Figure 2b. Amyloid PET/CT. Tracer: 18F-Fluorbetaben


Third Case (probable vascular dementia, VaD)

A 79-year-old men with history of hypertension and type 2 diabetes was evaluated for progressive loss of language. At the initial cognitive assessment, he showed a slight deficit on Mini Mental Examination State (MMSE 22.3/30 adjusted for age and education). His cerebral MRI was positive for the presence of small vascular lesions (not of lacunar type), in peri-ventricular and peri-trigonal areas and cortical atrophy. At the neurological examination bilateral Babinski sign and palmo-mental reflex were present. The cognitive evaluation showed multiple deficits on abstract reasoning, verbal fluency, executive and constructional functions, delayed memory recall difficulties. The initial diagnosis was of mild cognitive impairment due to cerebral vascular pathology. The FDG-PET scans however showed a significantly hypo-metabolism in different cerebral areas involving the frontal, temporal and parietal lobes (Figure 3a). The amyloid PET showed a significant presence of amyloid plaques (BAPL 3) (Figure 3b). The conclusive diagnosis was therefore of probable Alzheimer’s disease with associated cerebral vascular lesions.

Figure 3a. Metabolic PET. Tracer: 18F-FDG

a) Quantitative assessment of FDG uptake: SPM results on a voxel-by-voxel basis; b) Results of the analysis displayed on the Tailarach atlas


Figure 3b. Amyloid PET (axial sections). Tracer: 18F-Fluorbetaben


Figure 4. PET Algorithm Proposed



Discussion and proposition of a possible algorithm for the diagnosis of dementia, integrating neuropsychological and neuroimaging information

We briefly described three case reports underlying the utility of PET imaging in the differential diagnosis of dementia types. FDG-PET and amyloid-PET scans seem to be  useful  in the diagnostic algorithms of dementia, even if some limitations are present.
Our case-series introduce some considerations for a probably more appropriate prescription of amyloid and FDG-PET (Figure 4). When well defined neuropsychological deficits, regarding memory plus other cognitive domains, are present and the available international criteria are satisfied, an amyloid PET scan positive for a significant plaque deposition is suggestive for the diagnosis of AD. Conversely, when the amyloid PET is negative, the FDG-PET results should be considered. If the FDG-PET is indicative of hypomethabolism in the frontal areas, a diagnosis of FTD is highly probable (case report two). Moreover, the execution of the amyloid PET might also help in selected cases when the differential diagnosis between VaD and AD is not clear. In the frequent cases where MRI scans are positive for diffuse small vascular lesions, not clearly of lacunar origin, the amyloid PET might contribute to confirm a degenerative disease, especially when the clinical history is not suggestive of VaD. On the contrary, even in the presence of vascular lesions, a positive amyloid PET allows the clinical diagnosis of AD, supporting the hypothesis of a primary neurodegenerative diseases overlapping to cerebral vascular lesions (case report three).
Our case-series highlights the need of establishing a clear diagnostic algorithm emphasizing the role of the neuropsychological and clinical assessment combined with the information derived from nuclear medicine in order to achieve an early diagnosis allowing the most accurate and effective therapeutic options.
Since the burden of dementia in older persons is predicted to increase, it is imperative to develop accurate diagnostic techniques, particularly in older individuals where cerebrovascular lesions and neurodegenerative diseases may coexist.
The application of FDG-PET study to evaluate brain metabolic features in degenerative forms of dementia and amyloid PET to assess β-amyloid plaque load, in the absence of lacunar vascular lesions with significant memory loss, makes the diagnosis of Alzheimer’s disease more likely, thus reducing errors in differential diagnosis.
On the other hand, when cognitive deficits tend to spare memory and mainly involve language, a positive FDG-PET with a negative amyloid PET allow to suggest the presence of other forms of neurodegenerative dementia, such as FTD.
Our hypothesis to use molecular imaging techniques as complementary rather than redundant information is in accordance with recent data published by Besson et al. (17). They suggested  that MRI and FDG-PET biomarkers should be used in combination with amyloid PET, offering an additive contribution instead of reflecting the same process of neurodegeneration (17). The amyloid PET seems to have a high grade of specificity to discriminate AD from FTD, as confirmed by recent data, showing that the (18)F-florbetapir uptake is significantly different between AD and FTD patients (18). Then, another study found 18F-florbetaben to be useful in distinguishing patients with AD from healthy controls, as well as from patients with other neurodegenerative disorders. In fact, 96% of patient with AD and 60% of patients with MCI displayed broadly distributed cortical 18F-florbetaben retention, compared to 9% of patients with FTD, 25% of patients with VaD, 29% of patients with dementia with Lewy bodies (DLB), and 16% of healthy controls (19).
In conclusion, earlier (20) and more accurate diagnosis of AD (21), by using amyloid PET may help to accurately identify patients with AD, contributing to improve patient’s health status, despite the high costs of this technique. The proposed hypotheses should be confirmed in future studies with adequate sample size and prospective studies to confirm, even post-mortem, diagnosis suggested by integrating neuroimaging and neuropsychological findings.


Acknowledgments: The authors do not have any conflict of interest in the publication of this case series, have all contributed to the conception of the description and in the writing of this case series, and have approved the manuscript in its present form.

Conflict of interest: None.

Ethical standard: We allowed the good clinical practice for detecting different type of dementia in older persons.



1.    Zhang S, Smailagic N, Hyde C, et al. (11)C-PIB-PET for the early diagnosis of Alzheimer’s disease dementia and other dementias in people with mild cognitive impairment (MCI). Cochrane Database Syst Rev 2014;7:CD010386.
2.    Ossenkoppele R, Jansen WJ, Rabinovici GD, et al. Prevalence of amyloid PET positivity in dementia syndromes: a meta-analysis. JAMA 2015;313:1939-49.
3.    Jansen WJ, Ossenkoppele R, Knol DL, et al. Prevalence of cerebral amyloid pathology in persons without dementia: a meta-analysis.  JAMA 2015;313:1924-38.
4.    Smailagic N, Vacante M, Hyde C, et al. ¹⁸F-FDG PET for the early diagnosis of Alzheimer’s disease dementia and other dementias in people with mild cognitive impairment (MCI). Cochrane Database Syst Rev. 2015;1:CD010632.
5.    Sabri O, Sabbagh MN, Seibyl J, et al. Florbetaben PET imaging to detect amyloid beta plaques in Alzheimer’s disease: phase 3 study. Alzheimers Dement 2015;11:964-74.
6.    Ferrucci L, Bandinelli S, Cavazzini C, et al. Neurological examination findings to predict limitations in mobility and falls in older persons without a history of neurological disease. Am J Med. 2004;116:807-15.
7.    Caffarra P, Ghetti C, Concari L, et al. Differential patterns of hypoperfusion in subtypes of mild cognitive impairment. Open Neuroimag J 2008;2:20-8.
8.    Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7:270-9.
9.    McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7:263-9.
10.    Spinnler H, Tognoni G. Standardizzazione e taratura italiana di test neuropsicologici. Ital J Neurol Sci 1987;8:1–120.
11.    Caffarra P, Vezzadini G, Dieci F, et al. Modified Card Sorting Test: normative data. J Clin Exp Neuropsychol 2004;26:246-50.
12.    Caffarra P, Vezzadini G, Zonato F, et al. A normative study of a shorter version of Raven’s progressive matrices 1938. Neurol Sci 2003;24:336-9.
13.    Vinkers DJ, Gussekloo J, Stek ML, et al. The 15-item Geriatric Depression Scale (GDS-15) detects changes in depressive symptoms after a major negative life event. The Leiden 85-plus Study. Int J Geriatr Psychiatry 2004;19:80-84.
14.    Good CD, Johnsrude IS, Ashburner J, et al. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 2001;14:21–36.
15.    Barthel H, Luthardt J, Becker G, et al. Individualized quantification of brain beta-amyloid burden: results of a proof of mechanism phase 0 florbetaben PET trial in patients with Alzheimer’s disease and healthy controls. Eur J Nucl Med Mol Imaging 2011;38:1702–14.
16.    Barthel H, Gertz HJ, Dresel S, et al. Cerebral amyloid-beta PET with florbetaben (18F) in patients with Alzheimer’s disease and healthy controls: a multicentre phase 2 diagnostic study. Lancet Neurol 2011;10:424–35.
17.    Besson FL, La Joie R, Doeuvre L, et al. Cognitive and Brain Profiles Associated with Current Neuroimaging Biomarkers of Preclinical Alzheimer’s Disease. J Neurosci 2015;35:10402-11.
18.    Kobylecki C, Langheinrich T, Hinz R, et al. 18F-florbetapir PET in patients with frontotemporal dementia and Alzheimer disease. J Nucl Med 2015;56:386-91.
19.    Barthel H, Luthardt J, Becker G, et al. Individualized quantification of brain β-amyloid burden: results of a proof of mechanism phase 0 florbetaben PET trial in patients with Alzheimer’s disease and healthy controls. Eur J Nucl Med Mol Imaging 2011;38:1702–1714.
20.    Mondadori CR, Buchmann A, Mustovic H, et al. Enhanced brain activity may precede the diagnosis of Alzheimer’s disease by 30 years. Brain 2006;129:2908-22.
21.    Mosconi L, Sorbi S, de Leon MJ, et al. Hypometabolism exceeds atrophy in presymptomatic early-onset familial Alzheimer’s disease. J Nucl Med 2006;47:1778-86.



J. Delrieu1,2, S. Andrieu2, M. Pahor3, C. Cantet2, M. Cesari1,2, P.J. Ousset1,2, T. Voisin1,2, B. Fougère1, S. Gillette1,2, I. Carrie1, B. Vellas1,2


1. Gérontopôle, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France; 2. INSERM U 1027, Toulouse, France; 3. Department of Aging and Geriatric Research, University of Florida-Institute on Aging, Gainesville, FL, USA 

Corresponding Author: Julien Delrieu, Gérontopôle, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France, Phone: +33 (0) 561776426, Email : 

J Prev Alz Dis 2016;3(3):151-159
Published online March 8, 2016,



Objectives: An international group proposed the existence of “cognitive frailty”, a condition defined by simultaneous presence of physical frailty and cognitive impairment in the absence of dementia. The objective was to compare the neuropsychological profiles in subgroups of elders differentiated across their physical frailty (Fried phenotype) and cognitive status (Clinical Dementia Rating score) to characterize the “cognitive frailty” entity. 

Method: We studied baseline characteristics of 1,617 subjects enrolled in Multidomain Alzheimer Disease Preventive Trial (MAPT). Included subjects were aged 70 years or older and presented at least 1 of the 3 following clinical criteria: (1) Memory complaint spontaneously reported to a general practitioner, (2) limitation in one instrumental activity of daily living, (3) slow gait speed. Subjects with dementia were not included in the trial. 

Results: “Cognitive frailty individuals” significantly differed from “individuals with cognitive impairment and without physical frailty”, scoring worse at executive, and attention tests. They presented subcortico-frontal cognitive pattern different of Alzheimer Disease. Cognitive performance of subjects with 3 criteria or more of the frailty phenotype are cognitively more impaired than subjects with only one.

Discusion: The characterization of “cognitive frailty” must be done in frail subjects to set up specific preventive clinical trials for this population. 


Key words: Alzheimer Disease, elderly, MAPT trial, cognitive frailty, physical frailty.  



The frailty syndrome has recently attracted attention of the scientific community and public health authorities in numerous countries as risk factor for several age-related negative outcomes in older persons (1). In parallel, dementia and cognitive disorders also represent major healthcare and social priorities. The most commonly used definition of frailty was developed by Fried et al. in the Cardiovascular Health Study and in the Women’s Health and Aging Studies (2). Frailty was operationally defined as a clinical condition meeting 3 out of 5 criteria closely related to the physical domain: weak muscle strength, slow gait speed, unintentional weight loss, exhaustion, and sedentary behavior (3). Up to date, frailty and cognitive impairment have mostly been studied in parallel with very few attempts of simultaneously considering them. However, some recent work has started considering cognition as part of the definition of frailty, especially from an epidemiological viewpoint. Several biological and clinical conditions may underlie the age-related physical and cognitive declines: 1) depression (4), 2) cardiovascular risk factors (5), 3) genetic mutations (e.g. APO-E4) (6), 4) behavioral factors (e.g. low education, unhealthy dietary patterns, low physical and mental activity, smoking, high alcohol consumption), 5) oxidative damage and functional changes in the hippocampus and prefrontal cortex (7), 6) accumulation of common brain pathological findings (e.g. Alzheimer’s disease pathology, microinfarcts, nigral neuronal loss) (8–11), and 7) Low grade chronic inflammation. The absence of consideration of cognitive impairment in frailty syndrome could contribute to important heterogeneity of this entity (12)..      

An International Consensus Group organized by the International Academy on Nutrition and Aging (IANA) and the International Association of Gerontology and Geriatrics (IAGG), proposed the identification of the “cognitive frailty” condition (13). “Cognitive frailty” was hypothetically described as a clinical condition characterized by the simultaneous presence of both physical frailty and cognitive impairment. In particular, the key factors defining such a condition included: 1) presence of physical frailty and cognitive impairment, and 2) exclusion of concurrent Alzheimer’s disease (AD) dementia or other dementias. To identify “cognitive frailty”, the panel of experts suggested that all frail subjects should perform a comprehensive cognitive assessment exploring memory performance as well as other cognitive functions, in particular executive functions (with Montreal Cognitive assessment test (MoCa) (14), Mini Mental state Examination (MMSE) (15), Alzheimer’s Disease Assessment Scale-cognitive subscale (ADAS-Cog) (16), and speed processing tests). However, currently, cognitive pattern of “cognitive frailty” is not clearly characterized and the panel of experts of IANA and IAGG described a hypothetical condition without data available to support it. The cognitive profile identification of “cognitive frailty individuals” could be interesting because a potential for reversibility could also characterize this entity. Our hypothesis is that “cognitive frailty” would be a specific cognitive entity different of that met in AD, and which would be the witness of a more general impairment of the individual.

The main objective of our study was to determine the specific neuropsychological profile of “cognitive frailty individuals”, based on a sample of older adults of the Multidomain Alzheimer Preventive Trial (MAPT), aged 70 years and over, living in the community without any clinical sign of dementia (17,18). In addition, we aimed to assess the association of physical frailty severity with cognitive performance of “cognitive frailty individuals” and “frail older individuals”. 



MAPT study

The MAPT study was a 4-arm randomized controlled trial aimed at assessing the effects of isolated supplementation with omega-3 fatty acid, an isolated multidomain intervention (consisting of nutritional counseling, physical exercise, cognitive stimulation), or a combination of the 2 interventions, versus placebo, on cognitive functions modifications in older persons aged 70 years and older. A total of 1,680 subjects were enrolled in 13 memory clinics and followed up for 3 years. After the baseline assessment, participants also underwent cognitive, functional, and biological assessments after 6, 12, 24 and 36 months. The protocol is registered on a public-access clinical trial database (, Number: NCT01513252). Written, informed consent was obtained from all participants. 


Included subjects were aged 70 years or older and presented at least 1 of the 3 following clinical criteria: (1) Memory complaint spontaneously reported to a general practitioner, (2) limitation in one instrumental activity of daily living (IADL, i.e., inability in the use the telephone, shopping, preparation of meals, housekeeping, laundry, transportation, medication use, or management of money, (3) slow gait speed (i.e., ≤0.8 m/s). Subjects with dementia, limitation in basic activities of daily living (bathing, dressing, toileting, transferring, continence, eating) and suffering from severe depression were not included in the trial. . 

Clinical data  

Clinical Visits were scheduled every 6 months to assess physical condition, diseases and corresponding treatments, adherence to multi-domain intervention. Cognitive and functional assessments were conducted at baseline, six months, and annually at 1, 2 and 3 years by independent research staff blinded to intervention. 

Cognitive assessment

The battery of neuropsychological tests included the free and cued selective reminding test (FCRST, focused on verbal episodic memory/recall) (19), the Controlled Oral Word Association Test and Category Naming Test (COWAT and CNT, for verbal fluency) (20), the Digit Symbol Substitution Subtest of the Wechsler Adult Intelligence Scale-Revised (for attention and executive function) (21), the Trail-Making Test (TMT, measuring switching) (22), the Mini Mental State Examination (MMSE) (15), and the Clinical Dementia Rating Scale (CDR) (23). Two visual-analogue scales were administered, to assess memory function and the consequences of memory impairment in everyday life. Depressive symptoms was assessed with the Geriatric Depression Scale-15 items (GDS) (24). 

Physical and frailty assessment

Frailty was evaluated using the classification system proposed by Fried et al., based on assessments of grip strength, timed walking, unintentional weight loss, fatigue, and physical activity (3). In addition, functional assessment included the Alzheimer Disease Cooperative Study-Activities of Daily Living Prevention Instrument (ADCS-ADL) (25) and the Short Physical Performance Battery (SPPB) (26).

Classification of groups

Participants were classified into four groups according to the presence of cognitive impairment and/or frailty syndrome (Box 1).  The four groups were mainly defined as follows:

– Group 1: “Robust older persons” with no evidence of physical frailty (i.e., no frailty criteria) and absence of cognitive impairment (i.e., CDR=0), 

– Group 2: “Frail older individuals” with at least one Fried criteria and without cognitive impairment (i.e., CDR=0),

– Group 3: “Individuals with cognitive impairment and without physical frailty” with no Fried criteria and with cognitive impairment (i.e., CDR=0.5).   

– Group 4: “Cognitive frailty individuals” with at least one Fried criterion and with cognitive impairment (i.e., CDR=0.5),


Box 1. Definitions used to establish the 4 sub-groups of this study



We compared clinical characteristics, in particular neuropsychological profile, of subjects according to their frailty and cognitive status. We compared in a first time 4 groups of subjects: group 1, group 2, group 3, and group 4 (group 4 is the reference group for this analysis). In a second time, to evaluate the impact of physical frailty severity, we compared cognitive performance according to the number of physical frailty criteria (1, 2, and 3 or more) among subjects with CDR score of 0 and 0.5 (group with only 1 Fried criteria is the reference group for this analysis).

We used χ2 or Fisher’s exact (for expected values <5) tests for categorical variables, one way analyses of variance for quantitative variables with normal distributions (Student tests), and non-parametric tests (Kruskal-Wallis test) for quantitative variables without normal distributions. We compared characteristics of “frailty cognitive individuals” (group 4) with group 1, 2, and 3; using univariate polytomic regressions for categorical variables and univariate linear regressions or Kruskal-Wallis tests for continuous variables. In the absence of a normal distribution, variables were transformed and tested on square root or logarithmic value in order to obtain normal distributions. A multivariate analysis was also conducted to test the effect of potential confounding factors: 1) age, gender, socio-cultural level, BMI, and GDS for the first analysis (cognitive profile of “cognitive frailty” group), and 2) age, gender, socio-cultural level, and GDS for the second analysis (impact of number of physical frailty criteria on cognitive performance).

P values were based on two-sided tests. To account for the multiplicity of tests with an overall risk of 5%, each comparison compared to the reference group is considered significant if the «p» is <0.05/number of comparison, either 0.017 for 3 comparisons (first analysis), and 0.025 for 2 comparisons (second analysis). Analyses were performed using SAS software version 9.4 (SAS institute, Cary, NC, USA).




Figure 1 shows the flow chart of this study. Table 1 shows baseline characteristics for the 1,617 MAPT participants studied in this work. “Cognitive frailty individuals” with at least 1 Fried criterion and with cognitive impairment (i.e., CDR=0.5), represented 356 subjects, 22% of our study population. 


Figure 1. Flow chart of the study


Table 1. Baseline characteristics of MAPT study subjects (n=1617)

MMSE, Mini Mental Scale Examination; CDR, Clinical Dementia Rating Score; SPPB, Short Physical Performance Battery; ADCS-ADL, Alzheimer’s Disease Cooperative Study-Activities of Daily Living; TMT, Trail Making Test; COWAT, Controlled Oral Word Association Test; CNT, Categorial naming testing ; FCRST, Free and Cued Selective Reminding Test; GDS, Geriatric Depression rating; Visual Analogue Scale 1, Visual Analogue Scale, memory functioning; Visual Analogue Scale 2, Visual Analogue Scale, consequences in everyday life. Index of cuing (%) = (free recall-total recall) / (free recall-48).


“Cognitive frailty individuals” significantly differed from “individuals with cognitive impairment and without physical frailty” for age, GDS and Body Mass Index (Body Mass Index), from “frail older individuals” for gender, GDS and age, and from “robust older persons” for age, gender, education years, GDS and BMI (table 2).


Table 2. Cognitive profile of “cognitive frailty individuals” in MAPT study

MMSE, Mini Mental Scale Examination; CDR, Clinical Dementia Rating Score; SPPB, Short Physical Performance Battery; ADCS-ADL, Alzheimer’s Disease Cooperative Study-Activities of Daily Living; TMT, Trail Making Test; COWAT, Controlled Oral Word Association Test; CNT, Categorial naming testing ; FCRST, Free and Cued Selective Reminding Test; GDS, Geriatric Depression rating; Visual Analogue Scale 1, Visual Analogue Scale, memory functioning; Visual Analogue Scale 2, Visual Analogue Scale, consequences in everyday life; Index of cuing (%) = (free recall-total recall) / (free recall-48); * indicates that subjects from the “cognitive frailty individuals” significantly differ from either the “normal older individuals” or “frail older individuals” or “individuals with cognitive impairment and without physical frailty” (p<0.017) in multivariate analysis (ajustement for age, gender, socio-cultural level, BMI, and GDS)


Cognitive profile of “cognitive frailty” group 

“Cognitive frailty individuals” significantly differed with lower performance from “frail older individuals” and “robust older persons” for all cognitive tests (MMSE, CDR-SB, FCRST, TMT-A and –B, WAIS-R coding, CNT, and COWAT), visual analogue scales and some physical frailty tests (handgrip strength and slow gait speed) (table 2) in bivariate and multivariate analysis.  

“Cognitive frailty individuals” significantly differed with lower performance from “individuals with cognitive impairment without physical frailty” for CDR-SB, free recall and delayed free recall of FCRST, TMT-A and TMT-B, WAIS-R coding, CNT, and visual analogue scales (table 2). Multivariate analysis indicated that “cognitive frailty individuals” and “individuals with cognitive impairment without physical frailty” had similar profiles on FCRST, TMT-B, visual analogue scale 1, and CNT although “cognitive frailty individuals” demonstrated significantly more impairment in visual analogue scale 1, CDR-SB, TMT-A, and WAIS-R coding.


Impact of physical frailty severity on cognitive performance 

Subjects without cognitive impairment (i.e., CDR=0)

In multivariate analysis, subjects with only 1 Fried criterion significantly differed with better performance from subjects with 3 Fried criteria and more for delayed free recall of FCRST, CDR-SB, and visual analogue scale 1 (table 3).


Table 3. Impact of physical frailty severity on cognitive performance in “frail older individuals”

MMSE, Mini Mental Scale Examination; CDR-SB, Clinical Dementia Rating Score-Sum of Boxes; TMT, Trail Making Test; COWAT, Controlled Oral Word Association Test; CNT, Categorial naming testing ; FCRST, Free and Cued Selective Reminding Test; Visual Analogue Scale 1, Visual Analogue Scale, memory functioning; Visual Analogue Scale 2, Visual Analogue Scale, consequences in everyday life; Index of cuing (%) = (free recall-total recall) / (free recall-48); *indicates that subjects from the “Fried=1” group significantly differ from either the ‘Fried=2” or “Fried ≥ 3” groups (p<0.025) in multivariate analysis (ajustement for age, gender, socio-cultural level, and GDS).


Subjects with cognitive impairment (i.e., CDR=0.5)

In multivariate analysis, subjects with only 1 Fried criteria significantly differed with better performances from subjects with 3 Fried criteria and more for WAIS-R coding, and CDR-SB (table 4).


Table 4. Impact of physical frailty severity on cognitive performance in “cognitive frailty individuals”

MMSE, Mini Mental Scale Examination; CDR-SB, Clinical Dementia Rating Score-Sum of Boxes; TMT, Trail Making Test; COWAT, Controlled Oral Word Association Test; CNT, Categorial naming testing ; FCRST, Free and Cued Selective Reminding Test; Visual Analogue Scale 1, Visual Analogue Scale, memory functioning; Visual Analogue Scale 2, Visual Analogue Scale, consequences in everyday life; Index of cuing (%) = (free recall-total recall) / (free recall-48), *indicates that subjects from the “Fried=1” group significantly differ from either the ‘Fried=2” or “Fried ≥ 3” groups (p<0.025) in multivariate analysis (ajustement for age, gender, socio-cultural level, and GDS).



In the bivariate analysis, “cognitive frailty individuals” significantly differed with lower performance from “individuals with cognitive impairment and without physical frailty” for CDR-SB, free recall and delayed free recall of FCRST, TMT-A and TMT-B, WAIS-R coding, CNT, and visual analogue scales. Multivariate analysis demonstrated significantly more impairment in visual analogue scale 1, CDR-SB, and WAIS-R coding.

“Cognitive frailty” has been conceived as a clinical condition characterized by the simultaneous presence of both physical frailty and cognitive impairment (after exclusion of dementia). To our knowledge, no previous study has attempted to determinate the cognitive profile of “cognitive frailty individuals” in comparison of subjects with cognitive impairment and without physical frailty. MAPT study provided an opportunity to describe the cognitive functions of a large sample of subjects with cognitive and physical performances well characterized. To achieve this objective, “cognitive frailty individuals” were included from MAPT study on the basis of the following: 1) CDR of 0.5 to objective the cognitive impairment. In MAPT study, all included participants at baseline had basic activities of daily living preserved (inclusion criteria) and no dementia. So, we have considered subjects with CDR score of 0.5 as MCI subjects. 2) At least one Fried Criterion of physical frailty and not 3 or more, because we wanted to cover the entire spectrum of physical frailty and pre-frailty population seen in memory clinic and geriatric centers in this analysis.

Cognitive profile of “cognitive frailty individuals” was an amnesic MCI multidomain. In fact, in bivariate analysis, memory, attention, and executive performances of “cognitive frailty individuals” were lower than in “individuals with cognitive impairment and without physical frailty” (as we may consider as MCI without physical frailty individuals). The multivariate analysis indicated that “cognitive frailty individuals” demonstrated significantly only more impairment for executive functions than in “individuals with cognitive impairment and without physical frailty”. Altered executive functions were mainly processing speed (TMT-A and WAIS-R coding), selective attention (WAIS-R coding) and mental flexibility (semantic fluency). The dissociation in semantic and phonenic fluency could be support the degradation in semantic knowledge in the “cognitive frailty individuals”. WAIS-R coding assessed the scanning and tracking aspect of attention. This test has also been found to measure aspects of visual selective attention and processing speed. Research using previous versions of the WAIS in non-clinical samples has suggested that the age-related decline in WAIS-R coding scores is related to motor ability (27). Performances in our study sample with physical frailty, probably due to lower executive and attention performances but also due to lower motor abilities. The main characteristic of the FCSRT was to assess verbal episodic memory with semantic cueing that permitted one to control for encoding and to facilitate retrieval in order to isolate the storage capacities. The cued recall technique, used in the FCSRT, was aimed at enhancing the recall performance by presentation of semantic cues that help for encoding and for retrieval processes. In this study, free recall and delayed free recall performances were lower in “cognitive frailty individuals” than in “individuals with cognitive impairment and without physical frailty”. Total recall, delayed total recall and index of cuing were not significantly different between these 2 groups. This memory pattern differed from amnesic syndrome of the medial temporal (or hippocampal) which is characterized by a low free recall performance with a decreased total recall because of insufficient effect of cueing (28). The ability to benefit from cues mainly reflected impairment in strategies to retrieve stored information, as “subcortico-frontal dementia”. The motor features contributing to physical frailty derive from motor control systems which reside in the brain including basal ganglia, brainstem, frontal and subcortico-frontal areas. Thus, it is likely that physical frailty and cognition may show some degree of inter-relationship due to the effect on both from processes occurring in the brain (29). For example, the presence of cerebrovascular disease (8) and nigral neuronal loss (9) in older adults is associated with higher levels of frailty and lower levels of physical and cognitive functions, and could be responsible of “subcortico-frontal dementia”.  Depressive symptoms are also related to cognitive outcomes (30,31), in particular to executive functions. Kelaiditi et al maintained that “cognitive frailty” is characterized by reduced cognitive reserve. “Cognitive frailty” could be viewed as simply the inverse of cognitive reserve (32). 

We also estimated the association of physical frailty severity (number of frailty criteria) on cognitive performance. Subjects with 2 criteria, and 3 criteria or more, had more impaired cognitive scores (in particular for executive functions) than subjects with only one. Thus, more physical frailty would be severe and more cognitive performances would be impaired. The association between cognition performance and physical frailty severity seemed to be more important in normal cognitive functioning group (“frail older individuals” with at least 1 Fried criteria and CDR=0) than in “cognitive frailty individuals”. This cross-sectional study was not be able to assess the causal direction, whether physical frailty impacts cognitive performance or whether low cognitive performance impacts physical frailty.  However, physical frailty probably could impact directly administration of cognitive testing, and cognitive scores more impaired in severe physical frailty could be in relation with both motor and cognitive performance. One other hypothesis is that we are more likely to see effects of physical frailty on cognition in normal cognitive functioning group because “frail older individuals” have not yet cognitive impairment, and in ”cognitive frailty individuals”, probably some subjects have already prodromal AD (or MCI due to AD) which could decrease cognitive effect of physical frailty severity. 

The main key points of this study are: 1) it’s the first study which estimated the cognitive profile of subjects with “cognitive frailty”, 2) the large sample study, and 3) the neuropsychological battery realized in MAPT study permitting the well characterization of executive and memory functions. The absence of specific cognitive functions assessment (language, perception, praxia) did not allow to rule on the integrity supposed by these functions in “cognitive frailty individuals”.

“Cognitive frailty individuals” had executive and attention performance worse than “individuals with cognitive impairment and without physical frailty”. They presented a subcortico-frontal cognitive pattern, different of AD which a cortical neurodegenerative dementia. So, after exclusion of dementia and cognitive impairment diagnosis, it seems to be really important to use adequate cognitive screening tools to diagnosis “cognitive frailty individuals” in parallel of usual physical frailty, or cognitive markers because they would be an interesting target for specific prevention intervention. To identify “cognitive frailty”, we could suggest that frail subjects should perform as screening tests Frontal Assessment Battery (33), the 5 words test (34); and FCRST, TMT-A, TMT-B, WAIS-R coding and verbal fluencies as diagnosis tests. We could also propose Mattis Dementia Rating Scale (35).

This large population (“cognitive frailty” sample represented 22% of the population of MAPT study) could be targeted for non-specific multi-domain trials. The advantages of targeted “cognitive frailty individuals” for multi-domain prevention trials include the importance of intervening and potentially slowing or reversing the frailty syndrome, the large numbers of persons affected, and the ability to target these individuals through primary care physicians.  Disadvantages to target this population include the broad heterogeneity and presence of multiple morbidities within this population and the likelihood of poor compliance. The selection of the study sample may not be fully representative of the general population. In addition, the neurobiology of frailty has yet to be defined. Endpoints of a study in this population could include both physical and cognitive functions, in particular attention and executive tests.  

“Cognitive frailty” could represent a cognitive entity with specific neuropsychological patterns (executive and selective attention). The results of this cross-sectional study could justify a clinical follow-up to assess the cognitive evolution of “cognitive frailty individuals”. In ”cognitive frailty individuals”, probably some subjects have already prodromal AD. A longitudinal study could permit to determinate cognitive decline of “cognitive frailty individuals” and extract subjects who convert to AD in the longitudinal follow-up to better characterize cognitive pattern and trajectory of this entity. Pathophysiological mechanisms of “cognitive frailty” are currently unknown. Ancillary neuroimaging studies of MAPT could provide an opportunity to better understand the relation between “cognitive frailty” and cerebral atrophy, white matter hyperintensities, and amyloid deposits.


Authors contribution: J.D. was involved in the writing of the manuscript. J.D., M.P., S.A., B.F., T.V., PJ.O., S.G., I.C., M.C., and B.V. were involved in the review of the subsequent drafts. J.D., S.A., I.C., S.G., and B.V. were involved in the conception and organization of the research project. C.C. was involved in the execution of the statistical analysis.

Conflicts of interest: MAPT study was partially funded by the Institut de Recherche Pierre Fabre, Exhonit Therapeutics and Avid Radiopharmaceuticals Inc.

Funding: This study was supported by grants from the French Ministry of Health (PHRC 2008), and the Institut de Recherche Pierre Fabre (manufacturer of the omega-3 supplement). The promotion of this study was supported by the University Hospital Center of Toulouse. Biological sample collection was supported by Exhonit Therapeutics. AV45-MAPT study was supported by Avid Radiopharmaceuticals Inc and LABEX IRON Innovative Radiopharmaceuticals in Oncology and Neurology. Marco Pahor was supported by the University of Florida’s Claude D. Pepper Older Americans Independence Center (NIH/NIA P30AG028740). 

Principal investigator: Bruno Vellas (Toulouse); Coordination: Sophie Gillette-Guyonnet ; Project leader: Isabelle Carrié ; CRA: Lauréane Brigitte ;  Investigators: Catherine Faisant, Françoise Lala, Julien Delrieu; Psychologists: Emeline Combrouze, Carole Badufle, Audrey Zueras ; Methodology, statistical analysis and data management: Sandrine Andrieu, Christelle Cantet, Virginie Gardette, Christophe Morin; Multidomain group: Gabor Abellan Van Kan, Charlotte Dupuy, Yves Rolland (physical and nutritional components), Céline Caillaud, Pierre-Jean Ousset (cognitive component), Françoise Lala (preventive consultation) (Toulouse).  The cognitive component was designed in collaboration with Sherry Willis from the University of Seattle, and Sylvie Belleville, Brigitte Gilbert and Francine Fontaine from the University of Montreal. Co-Investigators in associated centre: Jean-François Dartigues, Isabelle Marcet, Fleur Delva, Alexandra Foubert, Sandrine Cerda (Bordeaux); Marie-Noëlle-Cuffi, Corinne Costes (Castres); Olivier Rouaud, Patrick Manckoundia, Valérie Quipourt, Sophie Marilier, Evelyne Franon (Dijon); Lawrence Bories, Marie-Laure Pader, Marie-France Basset, Bruno Lapoujade, Valérie Faure, Michael Li Yung Tong, Christine Malick-Loiseau, Evelyne Cazaban-Campistron (Foix); Françoise Desclaux, Colette Blatge (Lavaur); Thierry Dantoine, Cécile Laubarie-Mouret, Isabelle Saulnier, Jean-Pierre Clément, Marie-Agnès Picat, Laurence Bernard-Bourzeix, Stéphanie Willebois, Iléana Désormais, Noëlle Cardinaud (Limoges); Marc Bonnefoy, Pierre Livet, Pascale Rebaudet, Claire Gédéon, Catherine Burdet, Flavien Terracol (Lyon), Alain Pesce, Stéphanie Roth, Sylvie Chaillou, Sandrine Louchart (Monaco); Kristelle Sudres, Nicolas Lebrun, Nadège Barro-Belaygues (Montauban); Jacques Touchon, Karim Bennys, Audrey Gabelle, Aurélia Romano, Lynda Touati, Cécilia Marelli, Cécile Pays (Montpellier); Philippe Robert, Franck Le Duff, Claire Gervais, Sébastien Gonfrier (Nice); Yves Gasnier and Serge Bordes, Danièle Begorre, Christian Carpuat, Khaled Khales, Jean-François Lefebvre, Samira Misbah El Idrissi, Pierre Skolil, Jean-Pierre Salles (Tarbes). MRI group: Carole Dufouil (Bordeaux), Stéphane Lehéricy, Marie Chupin, Jean-François Mangin, Ali Bouhayia (Paris); Michèle Allard (Bordeaux); Frédéric Ricolfi (Dijon); Dominique Dubois (Foix); Marie Paule Bonceour Martel (Limoges); François Cotton (Lyon); Alain Bonafé (Montpellier); Stéphane Chanalet (Nice); Françoise Hugon (Tarbes); Fabrice Bonneville, Christophe Cognard, François Chollet (Toulouse). PET scans group: Pierre Payoux, Thierry Voisin, Julien Delrieu, Sophie Peiffer, Anne Hitzel, (Toulouse); Michèle Allard (Bordeaux); Michel Zanca (Montpellier); Jacques Monteil (Limoges); Jacques Darcourt (Nice). Medico-economics group: Laurent Molinier, Hélène Derumeaux, Nadège Costa (Toulouse). Biological sample collection: Christian Vincent, Bertrand Perret, Claire Vinel (Toulouse).



1. Morley JE. Frailty: diagnosis and management. J Nutr Health Aging. 2011 Aug;15(8):667–70. 

2. Ensrud KE, Blackwell TL, Redline S, Ancoli-Israel S, Paudel ML, Cawthon PM, et al. Sleep disturbances and frailty status in older community-dwelling men. J Am Geriatr Soc. 2009 Nov;57(11):2085–93. 

3. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001 Mar;56(3):M146–56. 

4. Panza F, Frisardi V, Capurso C, D’Introno A, Colacicco AM, Imbimbo BP, et al. Late-life depression, mild cognitive impairment, and dementia: possible continuum? Am J Geriatr Psychiatry Off J Am Assoc Geriatr Psychiatry. 2010 Feb;18(2):98–116. 

5. Phan HM, Alpert JS, Fain M. Frailty, inflammation, and cardiovascular disease: evidence of a connection. Am J Geriatr Cardiol. 2008 Apr;17(2):101–7. 

6. Snejdrlova M, Kalvach Z, Topinkova E, Vrablik M, Prochazkova R, Kvasilova M, et al. APOE polymorphism as a potential determinant of functional fitness in the elderly regardless of nutritional status. Neuro Endocrinol Lett. 2011;32 Suppl 2:51–4. 

7. Ho Y-Y, Matteini AM, Beamer B, Fried L, Xue Q, Arking DE, et al. Exploring biologically relevant pathways in frailty. J Gerontol A Biol Sci Med Sci. 2011 Sep;66(9):975–9. 

8. Buchman AS, Yu L, Boyle PA, Levine SR, Nag S, Schneider JA, et al. Microvascular brain pathology and late-life motor impairment. Neurology. 2013 Feb 19;80(8):712–8. 

9. Buchman AS, Yu L, Wilson RS, Schneider JA, Bennett DA. Association of brain pathology with the progression of frailty in older adults. Neurology. 2013 May 28;80(22):2055–61. 

10. Buchman AS, Schneider JA, Leurgans S, Bennett DA. Physical frailty in older persons is associated with Alzheimer disease pathology. Neurology. 2008 Aug 12;71(7):499–504. 

11. Buchman AS, Yu L, Wilson RS, Boyle PA, Schneider JA, Bennett DA. Brain Pathology Contributes to Simultaneous Change in Physical Frailty and Cognition in Old Age. J Gerontol A Biol Sci Med Sci. 2014 Aug 18; 

12. Bergman H, Ferrucci L, Guralnik J, Hogan DB, Hummel S, Karunananthan S, et al. Frailty: an emerging research and clinical paradigm–issues and controversies. J Gerontol A Biol Sci Med Sci. 2007 Jul;62(7):731–7. 

13. Kelaiditi E, Cesari M, Canevelli M, van Kan GA, Ousset P-J, Gillette-Guyonnet S, et al. Cognitive frailty: rational and definition from an (I.A.N.A./I.A.G.G.) international consensus group. J Nutr Health Aging. 2013 Sep;17(9):726–34. 

14. Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005 Apr;53(4):695–9. 

15. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975 Nov;12(3):189–98. 

16. Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer’s disease. Am J Psychiatry. 1984 Nov;141(11):1356–64. 

17. Gillette-Guyonnet S, Andrieu S, Dantoine T, Dartigues J-F, Touchon J, Vellas B, et al. Commentary on “A roadmap for the prevention of dementia II. Leon Thal Symposium 2008.” The Multidomain Alzheimer Preventive Trial (MAPT): a new approach to the prevention of Alzheimer’s disease. Alzheimers Dement J Alzheimers Assoc. 2009 Mar;5(2):114–21. 

18. Vellas B, Carrie I, Gillette-Guyonnet S, Touchon J, Dantoine T, Dartigues J françois, et al. MAPT Study: A Multidomain Approach for Preventing Alzheimer’s Disease: Design and Baseline Data. The Journal of Prevention of Alzheimer’s Disease. Volume 1, Number 1; 2014;13–22. 

19. Grober E, Buschke H, Crystal H, Bang S, Dresner R. Screening for dementia by memory testing. Neurology. 1988 Jun;38(6):900–3. 

20. Cardebat D, Doyon B, Puel M, Goulet P, Joanette Y. [Formal and semantic lexical evocation in normal subjects. Performance and dynamics of production as a function of sex, age and educational level]. Acta Neurol Belg. 1990;90(4):207–17. 

21. Wechsler D. Wechsler adult intelligence scale-revised. 1981; 

22. Reitan R. Validity of the Trail Making Test as an indicator of brain damage. Percept Mot Skills. 1958;271–6. 

23. Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for the staging of dementia. Br J Psychiatry J Ment Sci. 1982 Jun;140:566–72. 

24. Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res. 1982 1983;17(1):37–49. 

25. Galasko D, Bennett DA, Sano M, Marson D, Kaye J, Edland SD, et al. ADCS Prevention Instrument Project: assessment of instrumental activities of daily living for community-dwelling elderly individuals in dementia prevention clinical trials. Alzheimer Dis Assoc Disord. 2006 Dec;20(4 Suppl 3):S152–69. 

26. Guralnik JM, Ferrucci L, Pieper CF, Leveille SG, Markides KS, Ostir GV, et al. Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance battery. J Gerontol A Biol Sci Med Sci. 2000 Apr;55(4):M221–31. 

27. Kreiner DS, Ryan JJ. Memory and motor skill components of the WAIS-III Digit Symbol-Coding subtest. Clin Neuropsychol. 2001 Feb;15(1):109–13. 

28. Sarazin M, Berr C, De Rotrou J, Fabrigoule C, Pasquier F, Legrain S, et al. Amnestic syndrome of the medial temporal type identifies prodromal AD: a longitudinal study. Neurology. 2007 Nov 6;69(19):1859–67. 

29. Buchman AS, Bennett DA. Cognitive frailty. J Nutr Health Aging. 2013 Sep;17(9):738–9. 

30. Barnes DE, Alexopoulos GS, Lopez OL, Williamson JD, Yaffe K. Depressive symptoms, vascular disease, and mild cognitive impairment: findings from the Cardiovascular Health Study. Arch Gen Psychiatry. 2006 Mar;63(3):273–9. 

31. Saczynski JS, Beiser A, Seshadri S, Auerbach S, Wolf PA, Au R. Depressive symptoms and risk of dementia: the Framingham Heart Study. Neurology. 2010 Jul 6;75(1):35–41. 

32. Woods AJ, Cohen RA, Pahor M. Cognitive frailty: frontiers and challenges. J Nutr Health Aging. 2013 Sep;17(9):741–3. 

33. Dubois B, Slachevsky A, Litvan I, Pillon B. The FAB: a Frontal Assessment Battery at bedside. Neurology. 2000 Dec 12;55(11):1621–6. 

34. Dubois B, Touchon J, Portet F, Ousset PJ, Vellas B, Michel B. [“The 5 words”: a simple and sensitive test for the diagnosis of Alzheimer’s disease]. Presse Médicale Paris Fr 1983. 2002 Nov 9;31(36):1696–9. 

35. Paulsen JS, Butters N, Sadek JR, Johnson SA, Salmon DP, Swerdlow NR, et al. Distinct cognitive profiles of cortical and subcortical dementia in advanced illness. Neurology. 1995 May;45(5):951–6. 


K.J. Anstey1, R. Eramudugolla1, D.E. Hosking1, N.T. Lautenschlager2,3, R.A. Dixon4

1. Centre for Research on Ageing, Health and Wellbeing, Research School of Population Health, Australian National University; 2. Academic Unit for Psychiatry of Old Age, St. Vincent’s Health, Department of Psychiatry, University of Melbourne; 3. School of Psychiatry and Clinical Neurosciences & WA Centre for Health and Ageing, University of Western Australia; 4. Department of Psychology, University of Alberta, Canada

Corresponding Author: Kaarin J. Anstey, Centre for Research on Ageing, Health and Wellbeing, Florey Building 54, Mills Road, Australian National University, Canberra, ACT 0200, Australia  Tel: (612) 61258410; Fax: (612) 61251558; Email:

J Prev Alz Dis 2015;2(3):189-198
Published online June 16, 2015,


Dementia risk reduction is a global health and fiscal priority given the current lack of effective treatments and the projected increased number of dementia cases due to population ageing. There are often gaps among academic research, clinical practice, and public policy. We present information on the evidence for dementia risk reduction and evaluate the progress required to formulate this evidence into clinical practice guidelines. This narrative review provides capsule summaries of current evidence for 25 risk and protective factors associated with AD and dementia according to domains including biomarkers, demographic, lifestyle, medical, and environment. We identify the factors for which evidence is strong and thereby especially useful for risk assessment with the goal of personalising recommendations for risk reduction. We also note gaps in knowledge, and discuss how the field may progress towards clinical practice guidelines for dementia risk reduction.

Key words: Risk factor, Alzheimer disease, cognitive decline, prevention, risk assessment.  


Until effective therapeutics for Alzheimer’s disease (AD) are available, secondary prevention is an important focus for health professionals in relation to dementia. Secondary prevention encompasses risk reduction in those with multiple established risk factors for dementia, as well as interventions to slow progression of cognitive decline in adults with cognitive impairment or minor neurocognitive disorders (1). The development and integration of risk assessment and clinical risk management for AD and dementia is emerging rapidly in the research field, with the development of online risk assessment tools, education programs and lifestyle interventions (2, 3). However, for this research to have substantial public policy improvements, it must be translated, tested, and evaluated in clinical and community settings.  At present, risk management, including risk reduction and protection elevation, is the only available approach with potential for a large impact on the projected rates of dementia given population aging. The emerging focus on the development and implementation of efficacious dementia risk reduction protocols is consistent with best practice as applied to other chronic conditions or diseases and it has the advantages of being relatively inexpensive and translatable.   

The development of clinical practice guidelines and policies relating to dementia prevention requires a sound evidence base on risk and protective factors, as well as a framework for applying advice to relevant population-subgroups. This review provides an overview of 25 risk and protective factors for late-life AD and dementia and then describes how this knowledge may inform the development of risk assessment procedures and risk management targets in the context of clinical practice guidelines.

Multidomain assessment of risk and protective factors for AD

Many studies report research on risk and protective factors in relation to AD, while some only report results in relation to a general outcome of dementia. In this article we consult published meta-analyses and large-scale cohort studies to note which risk and protective factors have been linked to AD specifically or to AD and dementia more generally. There are fewer reports relating risk and protective factors specifically to vascular cognitive impairment because the prevalence of distinct vascular dementia (VaD) is lower than AD: hence, this review does not address that outcome.  There is often both vascular and Alzheimer pathology contributing to neurocognitive disorders. Notably, a lack of precision in diagnosis is a feature of the observational research on which much of the risk and protective factors have been based.

Established risk and protective factors for AD and dementia come from several domains, exacerbating the complexity of conducting a thorough risk assessment. That some factors may operate interactively or synergistically increases the need for careful interpretation of risk profiles. Assessing multiple domains of risk simultaneously permits an evaluation of overall risk profile, including the development of panels of risk factors, risk factor composite scores, and interactions among two or more synergistic risk factors.  Systematic review and meta-analysis of the current state of knowledge of risk factors for dementia is beyond the scope of a single article. We present a summary of the current knowledge in this area, drawing on key review articles, meta-analyses (1, 4-8) and individual papers, with more detail provided on recent findings on diet and dementia risk, which is an emerging area of interest.

Our framework for linking risk and protective factors to individual patient outcomes is depicted in Figure 1 (adapted from (9)). This figure shows the relationships between risk and protective factors for dementia, and clinical assessment goals. As depicted in the first column, there are five clusters of risk and protective factors, including biomarkers (not reviewed in this article), demographic factors, lifestyle, medical and environmental risk.  The second column illustrates that these factors flow together in individual cases and likely interact (synergistically, interactively, complementary) in unique ways to lead to individualized outcomes.  The third column identifies three general clusters of cognitive/clinical status outcomes, including relatively healthy brain and cognitive aging, typical non-demented cognitive trajectories, and cognitive impairment and dementia.  Subsequently, the fourth column presents an associated direction of personalized consultation regarding risk-related recommendations.  For healthy brain aging, advice will focus on sustained protective support and risk reduction.  For typical or normative aging, advice should focus on risk reduction and increase in protective behaviors.  For preclinical dementia or MCI, advice may focus on immediate risk control and perhaps reduction, especially targeting modifiable factors in the medical and lifestyle domains. Table 1 summarises the information on risk and protective factors as organized into the same demographic, lifestyle, medical and environmental domains.

Table 1. Summary of findings on risk and protective factors for AD and dementia


The role of biomarkers: Indicators of mechanisms

Risk and protective factors exert detectable and potentially manageable influence on the course of neurodegenerative disease through relatively diffuse or as yet undetermined biological mechanisms.  In some cases, biological markers (biomarkers) can be linked to somewhat more specific biological pathways associated with cognitive impairment, AD, and dementia.  It is likely that biomarkers will play an increasingly important role in risk assessment in the future (eg (10)).  Given the heterogeneity of aetiologies, mechanisms and phenotypes of dementia, biomarkers and their pathways may be ultimately used in compiling risk profiles for groups or individuals. Notably, modifiable risk factors that interact with specific biomarkers offer specific opportunities for early alterations in the course of the disease. Although a review of biomarkers for AD is beyond the scope of this article, the information provided by some well-known and typically accessible biomarkers can provide crucial supplemental information for the overall dementia risk profile (11). However, it is known that at autopsy, a large minority of individuals who died without dementia have AD neuropathology (12, 13); hence, detection of disease processes does not necessarily mean that an individual will develop AD.  This uncertainty does not change the need to target preventive strategies among at-risk individuals but indicates that individuals with biomarkers for AD may not necessarily express the clinical symptoms. (Autosomal mutations are an exception.)  As with all diseases that have multiple risk factors, where prediction (prior to diagnosis) is never entirely accurate, and which develop after decades of gradual accumulation of pathology, a comprehensive risk appraisal is required.

Figure 1. Model of Multidomain influences on major cognitive phenotypes in ageing including AD and dementia.Risk assessment and consultation goals are indicated for each phenotype

Adapted from Dolcos et al (2012)

Eventually, biomarkers of common pathogenic processes leading to brain ageing and concomitant Alzheimer pathology may provide a framework by which to organise risk and protective factors. For example, markers of inflammation are evident in a number of conditions that increase risk of AD including abdominal obesity, Type II diabetes and exposure to air pollution. However, most useful for present purposes would be information pertaining to genetic variants with known and elevated risk for AD. A recent large meta-analysis of genome-wide association studies in those of European ancestry found 11 new susceptibility loci for AD (14). The current state of knowledge indicates important supplemental information for constructing risk profiles could include (at least) APOE ε4 status.  For this article, we turn attention to risk and protective functions from the four domains of factors that operate through indirect (and often modifiable) pathways.

The demographic domain

Demographic risk factors include both modifiable and determinable non-modifiable characteristics, and enable profiling of population sub-groups at increased risk of dementia using population-level characteristics. Risk of AD and dementia strongly increases with chronological age (15), and in most countries is higher for women than men (15, 16). Low levels of formal school education increases the risk of AD and dementia (17). At present it is unknown whether increasing levels of education later in life confers the same protection as equivalent years of education obtained earlier in life. Higher levels of education appear to be associated with high level of cognitive function into late life, but not with reduced rate of decline (18).  In a related area, results have been inconsistent regarding bilingualism as a possible protective factor against late onset dementia. Although some evidence has suggested bilinguals have a delayed onset of dementia due to increased cognitive reserve (19) others have studied samples including monolinguals and bilinguals and found no difference in rate of cognitive decline or onset of dementia (20). A demographic characteristic that is rarely discussed in detail is race. It appears that specific racial and ethnic groups have higher rates of AD risk factors.  Some groups may have a higher or lower risk in relation to specific biological risk factors such as APOE (21), with evidence the APOE ε4 allele does not influence dementia progression in sub-Saharan Africans. Among developing countries, prevalence estimates of dementia for adults aged 65 and older are higher in certain Asian and Latin American countries, but are low (1-3%) in India and sub-Saharan Africa (22).  A recent study has shown that adults of Hispanic origin have earlier onset of dementia than non-Hispanics, adjusting for APOE genotype (23).  However, not enough data are available to produce quantitative pooled estimates of these effects.  Far more research is required to evaluate how risk profiles vary by race and ethnicity, which may potentially explain significant variation in the strength of specific genetic, medical or lifestyle factors as risk or protective in relation to AD. However, there is now sufficient evidence to incorporate age and sex into risk scores for incident dementia.

Lifestyle domain

Lifestyle-related risk factors for AD and dementia have been the focus of much recent research due to their modifiability. The prime lifestyle factors for which there is a body of evidence in relation to dementia risk include physical activity, diet, smoking, cognitive engagement and social engagement. 

Physical activity

There is consistent evidence that physical activity is associated with reduced AD and dementia risk, with higher levels of activity associated with the lowest risk (24). The benefits of physical activity for cognitive health appear to accumulate over the life course. For example, higher fitness levels in young adulthood has been linked with better cognitive outcomes in mid-adulthood (25), and better midlife fitness has been linked to reduced risk of late-life dementia (26). However, there is also evidence that taking up physical activity in old age can still impact positively on cognitive and functional performance (27). The effect of physical activity on brain ageing and neurodegeneration is also corroborated by neuroimaging studies and intervention studies (28, 29) and intervention durations of 6 months and longer are reported as being  more effective than shorter durations (30). To date, the majority of positive findings of trials are from samples of cognitively healthy older adults. A much smaller number of trials to date focused on trials with at-risk populations, especially those with subjective memory complaints or mild cognitive impairment. Some studies have reported significant benefits in the cognitive domains of attention, executive functions and memory (31, 32); however, other reports did not demonstrate such benefits (33). The inconsistency in results highlights the need for more high-quality later randomized controlled trials and a number of those are currently under way.

Dietary components and dietary patterns

The dietary component with the strongest link to AD and dementia risk reduction is oily fish, with three or more servings a week being associated with lower risk (34, 35).  Studies have consistently shown a relationship between low levels of alcohol intake (rather than abstinence) and reduced risk of AD, dementia and cognitive decline (36, 37). However, it is possible that this association partly reflects selection bias. Specifically, abstainers may include former heavy drinkers, with resultant poor cognitive and general health, and heavy drinkers are less likely to persist in longitudinal studies (38) . There is some evidence that n-3 fatty acids (39) and Vitamin B  may be beneficial for those in the early stages of decline although a recent meta-analysis of 11 trials found no cognitive benefits associated with Vitamin B supplementation (40).

The Mediterranean diet (MeDi) (Figure 2) was shown in a meta-analysis of five studies conducted with over 2-8 years follow-up to be associated with 33% reduced risk of cognitive impairment (MCI or AD) (41) and adherence to the MeDi has also been associated with reduced cognitive decline (42, 43). The cognitive benefits of the MeDi were confirmed by a 5-year randomised controlled trial (RCT). Those who consumed a MeDi supplemented with extra-virgin olive oil or mixed nuts had higher mean Mini-Mental State Examination (MMSE) scores and Clock Drawing Test scores than those who consumed a low fat control diet (44). The DASH diet (Dietary Approaches to Stop Hypertension) (45) includes whole grains, poultry, fish, and nuts and is reduced in saturated fats, red meats, sweets, and sugar-containing beverages. The two studies that have investigated associations between DASH dietary patterns and cognitive decline both found the DASH diet to be protective against cognitive decline (43, 46).

Figure 2. The Traditional Mediterranean Diet Pyramid

Adapted from Willett 1995 et al. (Copyright 1994 Oldways Preservation & Exchange Trust)


Smoking in late-life has been shown to increase the risk of AD, VaD and dementia (47, 48) and it is inferred from this that smoking earlier in adulthood is also associated with increased risk, although specific data on this are presently lacking. Smoking cessation is associated with less late-life (over 70) cognitive decline and brain atrophy than continued smoking (49) providing strong support for advising patients to cease smoking even at older ages.

Cognitive engagement 

Engaging in cognitively stimulating activities in late life (e.g., reading, playing puzzles and attending museums and concerts) is associated with a lower risk of AD and dementia (50, 51).  However effective dosage and type of cognitive activity are not yet known, and to date, there is no reliable evidence for an effect of cognitive training programs on delaying dementia. Research on the benefits of cognitive engagement is often confounded, as individuals with higher initial cognitive ability also engage in more cognitively stimulating lifestyles.

Social engagement

There is consistent evidence that higher levels of social engagement are associated with reduced risk of AD and dementia (52, 53), even in adults with the APOE e4 genotype (54), and there is RCT evidence that it increases brain volume (55). Social engagement measures include different types of relationships, living arrangements, size and quantity of social networks and amount of social activities.

Medical domain

Cardiometabolic risk factors in midlife have been linked to late-life cognitive decline, AD and late-life dementia (56, 57) but late-life cardiometabolic risk factors have less clear associations with dementia. The link between abnormally high or low blood pressure in late life and dementia risk is inconsistent (58), with some evidence that low blood pressure may increase cognitive decline through reduced perfusion (59).  While high blood pressure increases risk of stroke, and stroke is a strong risk factor for dementia (60), high blood pressure in late life has not been consistently linked with cognitive decline or AD. Findings relating blood pressure to risk of AD and dementia are complex, and influenced by methodological issues such as the length of follow up, whether or not treatment is evaluated, and whether trajectories are modelled that examine both increasing and decreasing hypertension at different stages of the adult life-course (58). Support has not been found in systematic reviews for a link between hypertension and AD (61). It is possible that the inconsistency associated with blood pressure and AD is due to the measure of blood pressure used in cohort studies. Peripheral hypertension is usually measured, and yet in old age, peripheral hypertension has a low correlation with central hypertension, which is the true risk factor for cerebrovascular changes and AD (62). In general, it appears that high blood pressure in midlife may represent a risk for dementia in later life (63). If untreated, hypertension in middle age that increases into old age may increase dementia risk, although a decline in blood pressure is seen in the period prior to the development of AD.

A recent systematic review of atrial fibrillation (AF) as a risk factor for cognitive impairment (defined as MMSE<24) or any type of dementia (DSM-IV criteria) identified an increased risk associated with AF both with and without history of stroke, despite study heterogeneity (64).

Stroke has also been considered as a risk factor for AD even after controlling for the presence of other cardiometabolic risk factors such as hypertension, diabetes and heart disease (65, 66). Individuals with a history of stroke had an earlier onset as well as a higher incidence of AD relative to those without a stroke history, although the risk was highest in those with recognized vascular risk factors. It is possible that a stroke may accelerate or bring above threshold the level of neuropathology and cognitive impairment required for progression to AD in those with mild or sub-clinical pathology (65).  In terms of other forms of dementia and vascular dementia in particular, a meta-analysis of 30 studies (66) reported that even after adjusting for other vascular risk factors, recurrent stroke increases the prevalence of dementia, with a rise in incidence after each additional stroke, suggesting stroke contributes significantly to the pathology leading to dementia over and above existing cardiovascular risk.

There is consistent evidence that elevated blood glucose (including Type II diabetes) increases the risk for cognitive decline, AD and dementia (67), and that this risk is independent of other cardiometabolic risk factors associated with diabetes (68, 69). There is also emerging evidence that pre-diabetic and sub-clinical levels of high blood glucose also predict cognitive decline and dementia (69, 70) although there have been no meta-analyses of this association.

Obesity and being overweight during midlife has been associated with increased late-life AD and dementia risk (71) with midlife obesity conferring double the risk of late-life AD.  The relationship between late-life obesity and dementia risk is unclear (71, 72) with the balance of evidence presently suggesting that it is not associated with increased risk. One study has shown that weight loss predicted AD similarly in overweight and normal weight adults, indicating that trajectory of weight loss rather than BMI was predictive of dementia in older adults (73).

High serum cholesterol during midlife is associated with elevated AD and dementia risk (74); however, this relationship is not consistently evident for high cholesterol in late life (75).

Systematic reviews have shown that clinically diagnosed depression and depressive symptoms, in both midlife and late-life, are each consistently associated with elevated risk of AD, cognitive decline and dementia (76-78). In late-life, evidence suggests that late-onset depression may also represent a prodrome of AD (79). Altogether these findings suggest that screening for depression is an essential component of risk profiling for AD.

Head injury during adulthood is associated with increased AD and dementia risk, specifically where the injuries were moderate or severe and occurred frequently (80-82). Repetitive concussion and head injuries as experienced by boxers is also associated with risk of developing a distinct neurodegenerative syndrome (83-85). Although not modifiable, information on history of head injury may contribute to an overall picture of a patient’s accumulated lifetime exposure to risks for AD.

Elevated plasma homocysteine increases the risk of vascular disease and stroke (86) and has been associated, albeit inconsistently, with poorer cognitive performance and dementia risk (87). Homocysteine levels increase with age and are dependent on Vitamin B metabolism (86). A number of trials in older adults have tested whether lowering homocysteine by vitamin B supplementation slows cognitive decline (40). Those with MCI and higher initial baseline homocysteine levels demonstrated better cognitive and clinical outcomes (88) and reduced brain atrophy (89). Recent preliminary evidence also suggests that the impact of homocysteine on older-age cognition may be dependent on its interactive effects with cholesterol (90).

There is some evidence from observational studies that certain classes of drugs such as anti-hypertensives, statins (91, 92), non-steroidal anti-inflammatories (NSAIDS) (93)  (but see (94)) and hormone replacement therapy (HRT) (95), are associated with reduced AD and dementia risk. However, RCTs are either lacking or, apart from some isolated findings, do not generally support dementia risk reduction through statin therapy (96) , the use of anti-hypertensives (97), or anti-inflammatories (98). A recent follow-up of RCTs indicate that HRTs have a complex pattern of risks and benefits for women’s health (99) and are thus not recommended for dementia prevention. One large study showed increased risk of dementia associated with NSAIDS (94). In one review of hypertensives, a significant effect (-18% incidence) was found for diuretic or dihydropyridine calcium channel blockers as part of active treatment for hypertension, although the overall pooled effect of hypertensives was not significant (100). Therefore, at this point in time, while medication is recommended to treat medical conditions associated with risk of AD, we lack high quality evidence of their role in prevention. On the other hand, the reduction in rates of dementia observed recently in several countries has been speculatively attributed in part to better management of cardiovascular risk factors, better health care and increased levels of education (101) .

Drugs with anticholinergic properties are used for treating common medical conditions such as asthma, urinary incontinence, seasonal allergies, insomnia, depression, and other psychiatric conditions (102, 103). Age-related decrease in cholinergic receptors and in the increase in blood-brain barrier permeability (102) increase the risk of anticholinergic medication causing cognitive impairment. The Adult Changes in Thought study found that a 10-year cumulative dose-response relationship was demonstrated between anticholinergic drug use and increased risk for all-cause dementia and AD. Associations remained robust across sub-group analyses and subclasses of anticholinergic medication use (103). These findings add weight to earlier clinical recommendations that elderly patients using anticholinergic medication should be monitored for cognitive dysfunction and if adverse effects are suspected, medications could be withdrawn (102).

Environmental domain

There is evidence for increased risk of dementia (Parkinson’s Disease with Dementia, and Alzheimer’s Disease and other dementias) in individuals exposed to very high levels of pesticides (104, 105). While there is insufficient data to date on the link between air pollution and dementia risk, there is some evidence high levels of air pollution is associated with greater cognitive impairment in older adults (106-108) and air pollution has been associated with AD neuropathology (108).

Conclusion: From risk assessment to risk reduction

This review has evaluated evidence pertaining to 25 factors that have been associated in epidemiological literature with increased risk of AD and dementia in some studies. Some of these factors are now considered uncontroversial risk or protective factors for AD (109) despite the lack of RCT evidence. The way in which knowledge about risk factors for AD and dementia is obtained does not map well onto the hierarchical model of the widely used GRADE system for ranking the quality of evidence. This is because most of the information on dementia risk is epidemiological and not experimental. However, clinical practice guidelines typically use systems such as GRADE in their development using consensus among experts. We argue that in the field of dementia prevention, it will be necessary to carefully consider the optimal methods of grading evidence so that routinely prioritising RCTs may not be the best approach. It is more appropriate that bodies of evidence are considered holistically or integratively (animal models, short-term RCTs, long term epidemiological studies, neuropathological evidence) in relation to putative risk factors and their mechanisms. Adopting rigorous yet realistic criteria is likely to be the most pragmatic approach to developing guidelines while evidence is still being collated and evaluated in this field.

Our review demonstrates that the multiple domains of risk and protection are populated by a variety of specific factors that independently or interactively may contribute to incident dementia in older adults. The body of evidence continues to grow and understanding of risk factors is becoming increasingly nuanced with (a) synergistic and modifier effects being increasingly addressed in observational research and (b) more specific (rather than global) questions being addressed in clinical trials. Broader questions that can now start to be addressed include consideration of the requirements to develop specific clinical practice guidelines for dementia prevention, and methods for evaluating risk assessment for AD and dementia.

To move this field towards evidence-based clinical practice guidelines, research needs to provide more specific information on the quantity or dose of factors that are required for protection. For example, specific prescriptions of physical activity, or specific dietary advice, should be further evaluated in trials. Recent findings from multidomain trials will be able to guide research and personal advice (110, 111). In some cases the qualitative types or range of activities that are protective is not known. For example, the value of brain training compared with a lifestyle of active reading has never been evaluated, and water-based physical activity has rarely been evaluated in relation to level and trajectory of cognitive function in aging. This lack of specific information reduces the clarity with which practical advice may be developed and distributed to individuals.

For risk factors such as blood glucose, more specific guidelines on levels of risk are required, as recent research shows that even within the normal range, variation in blood glucose is associated with future risk of dementia (70). Clinical practice guidelines also need to take account of the patient’s age and life course, considering the cumulative or synergistic effects of risk factors on other risk factors and outcomes that may ultimately increase late-life dementia risk or protection. For example, obesity in young adulthood may increase the risk of Type II Diabetes in mid-adulthood which in turn increases the risk of dementia in late adulthood.

Risk assessment for AD is now possible using well researched questionnaires, medical tests, checklists and online tools. There are now tools available to address risk in a range of ages and circumstances, from the population level in middle age, through to risk among older adults with brain atrophy and impaired IADLs. Risk assessment tools provide clinicians with validated means of assessing risk or identifying areas where protection may be increased, when they are administered to the appropriate group.

There is not yet a standard practice in this field but the evidence is now strong enough to support personalized recommendations for risk reduction by increasing levels of education in young adulthood, increasing physical, cognitive and social activity throughout adulthood, reducing cardiovascular risk factors including diabetes in middle-age, through lifestyle and medication, treating depression, adopting a healthy diet and physical activity, avoiding pesticides and heavy air pollution and teaching avoidance of all potential dangers to brain health while enhancing potential protective factors.

Now that risk assessment for dementia is possible, researchers and policy makers can also start to identify markers of success in dementia prevention interventions. We have previously argued that risk reduction, as opposed to dementia prevention, is a more realistic and useful immediate goal when focussing on the majority of the population who are middle-aged and have no cognitive symptoms (112). At the individual level, risk reduction is the most meaningful outcome, as the time frames for dementia prevention are so long and it is not possible to estimate the true contribution of genetic risk factors at the individual level. At the population level, estimates of incidence and prevalence over decades remain the most objective measures of disease burden; however, there are other factors that may indicate success in prevention strategies. These include delaying the age of onset of dementia, slowing the progression of disease and reducing the overall health burden associated with dementia.

Acknowledgements: KJA is funded by NHMRC Research Fellowship # 1002560. RAD is supported in part by a Canada Research Chair (Tier 1) and a grant from the National Institutes of Health (NIA; R01 AG008235). The research is supported by the Dementia Collaborative Research Centres. DH is funded by Australian Research Council Centre of Excellence in Population Ageing Research (#CE110001029).

Conflict of interests: Authors have no conflicts of interest to declare.


1. Fratiglioni L, Qiu C. Prevention of cognitive decline in ageing: dementia as the target, delayed onset as the goal. Lancet Neurol. 2011;10:778-9.

2. Barnes DE, Covinsky KE, Whitmer RA, Kuller LH, Lopez OL, Yaffe K. Commentary on «Developing a national strategy to prevent dementia: Leon Thal Symposium 2009.» Dementia risk indices: A framework for identifying individuals with a high dementia risk. Alzheimer’s & Dementia 2010;6:138-41.

3. Solomon A, Mangialasche F, Richard E, Andrieu S, Bennett DA, Breteler M, et al. Advances in the prevention of Alzheimer’s disease and dementia. J Intern Med. 2014;275:229-50.

4. Qiu C, Kivipelto M, von Strauss E. Epidemiology of Alzheimer’s disease: occurrence, determinants, and strategies toward intervention. Dialogues Clin Neurosci. 2009;11:111-28.

5. Ritchie K, Carriere I, Ritchie CW, Berr C, Artero S, Ancelin ML. Designing prevention programmes to reduce incidence of dementia: prospective cohort study of modifiable risk factors. BMJ. 2010;341:c3885.

6. Barnes DE, Covinsky KE, Whitmer RA, Kuller LH, Lopez OL, Yaffe K. Predicting risk of dementia in older adults: The late-life dementia risk index. Neurol. 2009;73:173-9.

7. Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol. 2011;10:819-28.

8. Williams JW, Plassman BL, Burke J, Benjamin S. Preventing Alzheimer’s disease and cognitive decline. Evid Rep Technol Assess (Full Rep). 2010:1-727.

9. Dolcos S, MacDonald SW, Braslavsky A, Camicioli R, Dixon RA. Mild cognitive impairment is associated with selected functional markers: integrating concurrent, longitudinal, and stability effects. Neuropsychology. 2012;26:209-23.

10. Waragai M, Hata S, Suzuki T, Ishii R, Fujii C, Tokuda T, et al. Utility of SPM8 plus DARTEL (VSRAD) ombined with magnetic resonance spectroscopy as djunct techniques for screening and predicting dementia due to Alzheimer’s disease in clinical practice. J Alzheimers Dis. 2014.

11. Hall JR, Wiechmann AR, Johnson LA, Edwards M, Barber RC, Cunningham R, et al. The impact of APOE status on relationship of biomarkers of vascular risk and systemic inflammation to neuropsychiatric symptoms in Alzheimer’s disease. J Alzheimers Dis. 2014.

12. Katzman R, Terry R, DeTeresa R, Brown T, Davies P, Fuld P, et al. Clinical, pathological, and neurochemical changes in dementia: a subgroup with preserved mental status and numerous neocortical plaques. Ann Neurol. 1988;23:138-44.

13. Stern Y. Cognitive reserve and Alzheimer disease. Alzheimer Dis Assoc Disord. 2006;20:S69-74.

14. Lambert J-C, Ibrahim-Verbaas CA, Harold D, Naj AC, Sims R, Bellenguez C, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet. 2013;45:1452-8.

15. Jorm AF, Dear KB, Burgess NM. Projections of future numbers of dementia cases in Australia with and without prevention. Aust N Z J Psychiatry. 2005;39:959-63.

16. Alzheimer Disease International. World Alzheimer’s Report. In: Prince M, Jackson J, editors. London: King’s College London; 2009.

17. Caamaño-Isorna F, Corral M, Montes-Martínez A, Takkouche B. Education and dementia: a meta-analytic study. Neuroepidemiology. 2006;26:226-32.

18. Zahodne LB, Glymour MM, Sparks C, Bontempo D, Dixon RA, MacDonald SW, et al. Education does not slow cognitive decline with aging: 12-year evidence from the Victoria longitudinal study. J Int Neuropsychol Soc. 2011;17:1039-46.

19. Bialystok E, Craik FI, Luk G. Bilingualism: consequences for mind and brain. Trends Cogn Sci. 2012;16:240-50.

20. Zahodne LB, Schofield PW, Farrell MT, Stern Y, Manly JJ. Bilingualism does not alter cognitive decline or dementia risk among Spanish-speaking immigrants. Neuropsychology. 2014;28:238-46.

21. Murrell JR, Price B, Lane KA, Baiyewu O, Gureje O, Ogunniyi A, et al. Association of apolipoprotein E genotype and Alzheimer disease in African Americans. Arch Neurol. 2006;63:431-4.

22. Kalaria RN, Maestre GE, Arizaga R, Friedland RP, Galasko D, Hall K, et al. Alzheimer’s disease and vascular dementia in developing countries: prevalence, management, and risk factors. Lancet Neurol. 2008;7:812-26.

23. Fitten LJ, Ortiz F, Fairbanks L, Bartzokis G, Lu P, Klein E, et al. Younger age of dementia diagnosis in a Hispanic population in southern California. Int J Geriatr Psychiatry. 2014.

24. Hamer M, Chida Y. Physical activity and risk of neurodegenerative disease: a systematic review of prospective evidence. Psychol Med. 2009;39:3-11.

25. Zhu N, Jacobs DR, Jr., Schreiner PJ, Yaffe K, Bryan N, Launer LJ, et al. Cardiorespiratory fitness and cognitive function in middle age: The CARDIA Study. Neurol. 2014;82:1339-46.

26. Defina LF, Willis BL, Radford NB, Gao A, Leonard D, Haskell WL, et al. The association between midlife cardiorespiratory fitness levels and later-life dementia: a cohort study. Ann Intern Med. 2013;158:162-8.

27. Almeida OP, Khan KM, Hankey GJ, Yeap BB, Golledge J, Flicker L. 150 minutes of vigorous physical activity per week predicts survival and successful ageing: a population-based 11-year longitudinal study of 12 201 older Australian men. Br J Sports Med. 2014;48:220-5.

28. Erickson KI, Raji CA, Lopez O, Becker JT, Rosano C, Newman AB, et al. Physical activity predicts gray matter volume in late adulthood. Neurology. 2010;75:1415-22.

29. Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci USA. 2011;108:3017-22.

30. Lautenschlager NT. Physical activity and cognition in older adults with mild cognitive impairment and dementia. Neurodegen Dis Management. 2013;3:211-8.

31. Lautenschlager NT, Cox KL, Flicker L, Foster JK, van Bockxmeer FM, Xiao J, et al. Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: a randomized trial. JAMA. 2008;300:1027-37.

32. Baker LD, Frank LL, Foster-Schubert K, Green PS, Wilkinson CW, McTiernan A, et al. Effects of aerobic exercise on mild cognitive impairment: a controlled trial. Arch Neurol. 2010;67:71-9.

33. Scherder EJ, Van Paasschen J, Deijen JB, Van Der Knokke S, Orlebeke JF, Burgers I, et al. Physical activity and executive functions in the elderly with mild cognitive impairment. Aging Ment Health. 2005;9:272-80.

34. Barberger-Gateau P, Raffaitin C, Letenneur L, Berr C, Tzourio C, Dartigues JF, et al. Dietary patterns and risk of dementia: the Three-City cohort study. Neurol. 2007;69:1921-30.

35. Huang TL, Zandi PP, Tucker KL, Fitzpatrick AL, Kuller LH, Fried LP, et al. Benefits of fatty fish on dementia risk are stronger for those without APOE ε4. Neurology. 2005;65:1409-14.

36. Anstey KJ, Mack HA, Cherbuin N. Alcohol consumption as a risk factor for dementia and cognitive decline: meta-analysis of prospective studies. Am J Geriatr Psychiatry. 2009;17:542-55.

37. Peters R, Peters J, Warner J, Beckett N, Bulpitt C. Alcohol, dementia and cognitive decline in the elderly: a systematic review. Age Ageing. 2008;37:505-12.

38. Roizen R, Fillmore K, Chikritzhs T, Stockwell T. Light-to-moderate drinking and dementia risk: the former drinkers problem re-visited. Addict Research Theory. 2013;21:181-93.

39. Freund-Levi Y, Eriksdotter-Jonhagen M, Cederholm T, Basun H, Faxen-Irving G, Garlind A, et al. Omega-3 fatty acid treatment in 174 patients with mild to moderate Alzheimer disease: OmegAD study: A randomized double-blind trial. Arch Neurol. 2006;63:1402-8.

40. Clarke R, Bennett D, Parish S, Lewington S, Skeaff M, Eussen SJ, et al. Effects of homocysteine lowering with B vitamins on cognitive aging: meta-analysis of 11 trials with cognitive data on 22,000 individuals. Am J Clin Nutr. 2014;100:657-66.

41. Singh B, Parsaik AK, Mielke MM, Erwin PJ, Knopman DS, Petersen RC, et al. Association of Mediterranean diet with mild cognitive impairment and Alzheimer’s disease: a systematic review and meta-analysis. J Alzheimers Dis. 2014;39:271-82.

42. Lourida I, Soni M, Thompson-Coon J, Purandare N, Lang IA, Ukoumunne OC, et al. Mediterranean diet, cognitive function, and dementia: a systematic review. Epidemiology. 2013;24:479-89.

43. Tangney CC, Hong L, Wang Y, Barnes L, Schneider JA, Bennett DA, et al. Relation of DASH- and Mediterranean-like dietary patterns to cognitive decline in older persons. Neurology. 2014.

44. Martínez-Lapiscina EH, Clavero P, Toledo E, Estruch R, Salas-Salvadó J, San Julián B, et al. Mediterranean diet improves cognition: the PREDIMED-NAVARRA randomised trial. J Neurol, Neurosurg & Psychiatry. 2013.

45. Vogt TM, Appel LJ, Obarzanek EVA, Moore TJ, Vollmer WM, Svetkey LP, et al. Dietary Approaches to Stop Hypertension: rationale, design, and methods. J Am Diet Assoc. 1999;99:S12-S8.

46. Wengreen H, Munger RG, Cutler A, Quach A, Bowles A, Corcoran  C, et al. Prospective study of Dietary Approaches to Stop Hypertension- and Mediterranean-style dietary patterns and age-related cognitive change: the Cache County study on memory, health and aging. Am J Clin Nutr. 2013;98:1263-71.

47. Anstey KJ, von Sanden C, Salim A, O’Kearney R. Smoking as a risk factor for dementia and cognitive decline: a meta-analysis of prospective studies. Am J Epidemiol. 2007;166:367-78.

48. Peters R, Poulter R, Warner J, Beckett N, Burch L, Bulpitt C. Smoking, dementia and cognitive decline in the elderly: a systematic review. BMC Geriatr. 2008;8:36.

49. Almeida OP, Garrido GJ, Alfonso H, Hulse G, Lautenschlager NT, Hankey GJ, et al. 24-month effect of smoking cessation on cognitive function and brain structure in later life. Neuroimage. 2011;55:1480-9.

50. Wilson RS, Scherr PA, Schneider JA, Tang Y, Bennett DA. Relation of cognitive activity to risk of developing Alzheimer disease. Neurology. 2007;69:1911-20.

51. Valenzuela M, Brayne C, Sachdev P, Wilcock G, Matthews F. Cognitive lifestyle and long-term risk of dementia and survival after diagnosis in a multicenter population-based cohort. Am J Epidemiol. 2011;173:1004-12.

52. Bassuk SS, Glass TA, Berkman LF. Social disengagement and incident cognitive decline in community-dwelling elderly persons. Ann Intern Med. 1999;131:165-73.

53. Fratiglioni L, Wang HX, Ericsson K, Maytan M, Winblad B. Influence of social network on occurrence of dementia: a community-based longitudinal study. Lancet. 2000;355:1315-9.

54. Ferrari C, Xu WL, Wang HX, Winblad B, Sorbi S, Qiu C, et al. How can elderly apolipoprotein E ε4 carriers remain free from dementia? Neurobiol Aging. 2013;34:13-21.

55. Mortimer JA, Ding D, Borenstein AR, DeCarli C, Guo Q, Wu Y, et al. Changes in brain volume and cognition in a randomized trial of exercise and social interaction in a community-based sample of non-demented Chinese elders. J Alzheimers Dis. 2012;30:757-66.

56. Kivipelto M, Helkala EL, Laakso MP, Hanninen T, Hallikainen M, Alhainen K, et al. Midlife vascular risk factors and Alzheimer’s disease in later life: longitudinal, population based study. BMJ. 2001;322:1447-51.

57. Fratiglioni L, Qiu C. Prevention of common neurodegenerative disorders in the elderly. Exp Gerontol. 2009;44:46-50.

58. Joas E, Backman K, Gustafson D, Ostling S, Waern M, Guo X, et al. Blood pressure trajectories from midlife to late life in relation to dementia in women followed for 37 years. Hypertension. 2012;59:796-801.

59. Glodzik L, Rusinek H, Pirraglia E, McHugh P, Tsui W, Williams S, et al. Blood pressure decrease correlates with tau pathology and memory decline in hypertensive elderly. Neurobiol Aging. 2014;35:64-71.

60. Dregan A, Wolfe CD, Gulliford MC. Does the influence of stroke on dementia vary by different levels of prestroke cognitive functioning?: A cohort study. Stroke. 2013;44:3445-51.

61. Power MC, Weuve J, Gagne JJ, McQueen MB, Viswanathan A, Blacker D. The association between blood pressure and incident Alzheimer disease: A systematic review and meta-analysis. Epidemiology. 2011;22:646-59.

62. Palatini P, Casiglia E, Gasowski J, Gluszek J, Jankowski P, Narkiewicz K, et al. Arterial stiffness, central hemodynamics, and cardiovascular risk in hypertension. Vascular health and risk management. 2011;7:725-39.

63. Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol. 2011.

64. Kwok CS, Loke YK, Hale R, Potter JF, Myint PK. Atrial fibrillation and incidence of dementia: a systematic review and meta-analysis. Neurology. 2011;76:914-22.

65. Honig LS, Ming-Xin T, Albert S, Costa R, Luchsinger JA, Manly JJ, et al. Stroke and the risk of Alzheimer disease. Arch Neurol. 2003;60:1707-12.

66. Pendlebury ST, Rothwell PM. Prevalence, incidence, and factors associated with pre-stroke and post-stroke dementia: a systematic review and meta-analysis. Lancet Neurol. 2009;8:1006-18.

67. Cheng G, Huang CT, Deng H, Wang H. Diabetes as a risk factor for dementia and mild cognitive impairment: a meta-analysis of longitudinal studies. Intern Med J. 2012;42:484-91.

68. Biessels GJ, Staekenborg S, Brunner E, Brayne C, Scheltens P. Risk of dementia in diabetes mellitus: a systematic review. Lancet Neurol. 2006;5:64-74.

69. Lu FP, Lin KP, Kuo HK. Diabetes and the risk of multi-system aging phenotypes: a systematic review and meta-analysis. PLOS-One. 2009;4.

70. Crane PK, Walker R, Hubbard RA, Li G, Nathan DM, Zheng H, et al. Glucose levels and risk of dementia. N Engl J Med. 2013;369:540-8.

71. Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev. 2011.

72. Sabia S, Kivimaki M, Shipley M, Marmot M, Singh-Manoux A. Body mass index over the adult life course and cognition in late midlife: the Whitehall II cohort study. Am J Clin Nutr. 2009;89:601-7.

73. Power BD, Alfonso H, Flicker L, Hankey GJ, Yeap BB, Almeida OP. Changes in body mass in later life and incident dementia. Int Psychogeriatr. 2013;25:467-78.

74. Anstey KJ, Lipnicki DM, Low LF. Cholesterol as a risk factor for dementia and cognitive decline: a systematic review of prospective studies with meta-analysis. Am J Geriatr Psychiatry. 2008;16:343-54.

75. Di Paolo G, Kim T-W. Linking lipids to Alzheimer’s disease: cholesterol and beyond. Nat Rev Neurosci. 2011;12:285-96.

76. Ownby RL, Crocco E, Acevedo A, John V, Loewenstein D. Depression and risk for Alzheimer disease: systematic review, meta-analysis, and metaregression analysis. Arch Gen Psychiatry. 2006;63:530-8.

77. Barnes DE, Yaffe K, Byers AL, McCormick M, Schaefer C, Whitmer RA. Midlife vs late-life depressive symptoms and risk of dementia: differential effects for Alzheimer disease and vascular dementia. Arch Gen Psychiatry. 2012;69:493-8.

78. Diniz BS, Butters MA, Albert SM, Dew MA, Reynolds CF. Late-life depression and risk of vascular dementia and Alzheimer’s disease: systematic review and meta-analysis of community-based cohort studies. Br J Psychiatry. 2013;202:329-35.

79. Lee GJ, Lu PH, Hua X, Lee S, Wu S, Nguyen K, et al. Depressive symptoms in mild cognitive impairment predict greater atrophy in Alzheimer’s disease-telated regions. Biol Psychiatry. 2012;71:814-21.

80. Fleminger S, Oliver DL, Lovestone S, Rabe-Hesketh S, Giora A. Head injury as a risk factor for Alzheimer’s disease: the evidence 10 years on; a partial replication. J Neurol Neurosurg Psychiatry. 2003;74:857-62.

81. Mayeux R, Ottman R, Maestre GE, Ngai BS, Tang MX, Ginsberg H, et al. Synergistic effects of traumatic head injury and apolipoprotein-e4 in patients with Alzheimer’s disease. Neurology. 1995;45:555-7.

82. Plassman BL, Havlik RJ, Steffens DC, Helms MJ, Newman TN, Drosdick D, et al. Documented head injury in early adulthood and risk of Alzheimer’s disease and other dementias. Neurology. 2000;56:1158-66.

83. Jordan BD. The clinical spectrum of sport-related traumatic brain injury. Nat Rev Neurol. 2013;9:222-30.

84. Smith DH, Johnson VE, Steward W. Chronic neuropathologies of single and repetitive TBI: substrates of dementia? Nat Rev Neurol. 2013:211-21.

85. Stern RA, Danshvar DH, Baugh CM, Seichepine DR, Montenigro PH, Riley DO, et al. Clinical presentation of chronic traumatic encephalopathy. Neurology. 2013;81:1122-9.

86. The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke. JAMA. 2002;288:2015-22.

87. Ho RCM, Cheung MWL, Fu E, Win H, Zaw MH, Ng A, et al. Is high homocysteine level a risk factor for cognitive decline in elderly? A systematic review, meta-analysis, and meta-regression. Am J Geriatr Psychiatry. 2011;19:607-17.

88. de Jager CA, Oulhaj A, Jacoby R, Refsum H, Smith AD. Cognitive and clinical outcomes of homocysteine-lowering B-vitamin treatment in mild cognitive impairment: a randomized controlled trial. Int J Geriatr Psychiatry. 2012;27:592-600.

89. Smith AD, Smith SM, de Jager CA, Whitbread P, Johnston C, Agacinski G, et al. Homocysteine-lowering by B vitamins slows the rate of accelerated brain atrophy in mild cognitive impairment: A randomized controlled trial. PLOS-One. 2010.

90. Cheng Y, Jin Y, Unverzagt FW, Su L, Yang L, Ma F, et al. The relationship between cholesterol and cognitive function is homocysteine-dependent. Clin Interv Aging. 2014;9:1823-9.

91. Song Y, Nie H, Xu Y, Zhang L, Wu Y. Association of statin use with risk of dementia: a meta-analysis of prospective cohort studies. Geriatr Gerontol Int. 2013;13:817-24.

92. Wong WB, Lin VW, Boudreau D, Devine EB. Statins in the prevention of dementia and Alzheimer’s disease: A meta-analysis of observational studies and an assessment of confounds. Pharmacoepidemiol Drug Saf. 2013;22:345-58.

93. Etminan M, Gill S, Samii A. Effect of non-steroidal anti-inflammatory drugs on risk of Alzheimer’s disease: systematic review and meta-analysis of observational studies. BMJ. 2003;327:128.

94. Breitner JC, Haneuse SJ, Walker R, Dublin S, Crane PK, Gray SL, et al. Risk of dementia and AD with prior exposure to NSAIDs in an elderly community-based cohort. Neurology. 2009;72:1899-905.

95. Zandi PP, Carlson MC, Plassman BL. Hormone replacement therapy and incidence of Alzheimer disease in older woment: the Cache County study. JAMA. 2002;288:2123-9.

96. Scott HD, Laake K. Statins for the prevention of Alzheimer’s disease and dementia. Cochrane database of systematic reviews. 2009;CD003160:1-11.

97. McGuiness B, O’Hare J, Craig D, Bullock R, Maloug R, Passmore P. Cochrane review on ‘Statins for the treatment of dementia’. Int J Geriatr Psychiatry. 2013;28:119-26.

98. Breitner JC, Baker LD, Montine TJ, Meinert CL, Lyketsos C, Ashe KH, et al. Extended results of the Alzheimer’s disease anti-inflammatory prevention trial. Alzheimer’s & Dementia. 2011;7:402-11.

99. Manson JE, Chlebowski RT, Stefanick ML, et al. Menopausal hormone therapy and health outcomes during the intervention and extended post-stopping phases of the women’s health initiative randomized trials. JAMA. 2013;310:1353-68.

100. Staessen JA, Thijs L, Richart T, Odili AN, Birkenhager WH. Placebo-controlled trials of blood pressure-lowering therapies for primary prevention of dementia. Hypertension. 2011;57:e6-7.

101. Matthews FE, Arthur A, Barnes LE, Bond J, Jagger C, Robinson L, et al. A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. Lancet. 2013.

102. Campbell N, Boustani M, Limbil T, Ott C, Fox C, Maidment I, et al. The cognitive impact of anticholinergics: A clinical review. Clin Interv Aging. 2009;4:225-33.

103. Gray SL, Anderson ML, Dublin S, et al. Cumulative use of strong anticholinergics and incident dementia: A prospective cohort study. JAMA Int Med. 2015;175:401-7.

104. Hayden KM, Norton MC, Darcey D, Ostbye T, Zandi PP, Breitner JC, et al. Occupational exposure to pesticides increases the risk of incident AD: the Cache County study. Neurology. 2010;74:1524-30.

105. Priyadarshi A, Khuder SA, Schaub EA, Shrivastava S. A meta-analysis of Parkinson’s disease and exposure to pesticides. Neurotoxicology. 2000;21:435-40.

106. Chang K-H, Chang M-Y, Muo C-H, Wu T-N, Chen C-Y, Kao C-H. Increased risks of dementia in patients exposed to nitrogen dioxide and carbon monoxide: A population-based retrospective cohort study. PLOS-One. 2014.

107. Brown TP, Rumsby PC, Capleton AC, Rushton L, Levy LS. Pesticides and Parkinson’s Disease: is there a link? Environ Health Perspect. 2006;114:156-64.

108. Calderon-Garciduenas L, Kavanaugh M, Block M, D’Angiulli A, Delgado-Chavez R, Torres-Jardon R, et al. Neuroinflammation, hyperphosphorylated tau, diffuse amyloid plaques, and down-regulation of the cellular prion protein in air pollution exposed children and young adults. J Alzheimers Dis. 2012;28:93-107.

109. 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;13:788-94.

110. Ngandu T, Lehtisalo J, Solomon A, Levalahti E, Ahtiluoto S, Antikainen R, 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. Lancet. 2015.

111. Vellas B, Carrie I, Gillette-Guyonnet S, Touchon J, Dantoine T, Dartigues JF, et al. MAPT study: a multidomain approach for preventing Alzheimer’s disease: design and baseline data. Journal for the Prevention of Alzheimer’s Disease. 2014;1:13-22.

112. Anstey KJ, Bahar-Fuchs A, Herath P, Rebok GW, Cherbuin N. A 12-week multidomain intervention versus active control to reduce risk of Alzheimer’s disease: study protocol for a randomized controlled trial. Trials. 2013;14:60-.

113. Kalantarian S, Stern TA, Mansour M, Ruskin JN. Cognitive impairment associated with atrial fibrillation a meta-analysis. Ann Intern Med. 2013;158:338-46.

114. Gudala K, Bansal D, Schifano F, Bhansali A. Diabetes mellitus and risk of dementia: A meta-analysis of prospective observational studies. Journal of Diabetes Investigation. 2013;4:640-50.

115. Loef M, Walach H. Midlife obesity and dementia: meta-analysis and adjusted forecast of dementia prevalence in the United States and China. Obesity. 2013;21:E51-E5.

116. Gao Y1 HC, Zhao K, Ma L, Qiu X, Zhang L, Xiu Y, Chen L, Lu W, Huang C, Tang Y, Xiao Q. Depression as a risk factor for dementia and mild cogniive impairment: a meta-analysis of longitudinal studies. Int J Geriatr Psychiatry. 2013;28:441-9.

117. Ho RCM, Cheung MWL, Fu E, Win H, Zaw MH, A. N, et al. Is high homocysteine level a risk factor for cognitive decline in elderly? A systematic review, meta-analysis, and meta-regression. Am J Geriatr Psychiatry. 2011;19:607-17.

118. Zhou B, Teramukai S, Fukushima M. Prevention and treatment of dementia or Alzheimer’s disease by statins: a meta-analysis. Dement Geriatr Cogn Disord. 2007;23:194-201.

119. Swiger KJ, Manalac RJ, Blumenthal RS, Blaha MJ, Martin SS. Statins and cognition: A systematic review and meta-analysis of short- and long-term cognitive effects. Mayo Clinic Proceedings. 2013;88:1213-21.

120. Levi Marpillat N, Macquin-Mavier I, Tropeano AI, Bachoud-Levi AC, Maison P. Antihypertensive classes, cognitive decline and incidence of dementia: a network meta-analysis. J Hypertens. 2013;31:1073-82.

121. Chang-Quan H, Hui W, Chao-Min W, Zheng-Rong W, Jun-Wen G, Yong-Hong L, et al. The association of antihypertensive medication use with risk of cognitive decline and dementia: a meta-analysis of longitudinal studies. Int J Clin Pract. 2013;65:1295-305.

122. Wang J, Tan L, Wang HF, Tan CC, Meng XF, Wang C, et al. Anti-inflammatory drugs and risk of Alzheimer’s disease: an updated systematic review and meta-analysis. J Alzheimers Dis. 2014.

123. LeBlanc ES, Janowsky J, Chan BKS, Nelson HD. Hormone replacement therapy and cognition: systematic review and meta-analysis. JAMA. 2001;285:1489-99.




S. Villeneuve, W.J. Jagust


Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley CA, USA. 

Corresponding Author: Sylvia Villeneuve, Helen Wills Neuroscience Institute, 132 Barker Hall, MC# 3190, University of California, Berkeley, CA 94720 USA,
E-mail:, Telephone: 510-643-6616, Fax: 510-642-3192

J Prev Alz Dis 2015;2(1):64-70
Published online Januay 29, 2015,


Vascular risk factors (e.g. hypertension, dyslipidemia and diabetes) are well known risk factors for Alzheimer’ disease. These vascular risk factors lead to vascular brain injuries, which also increase the likelihood of dementia. The advent of amyloid PET imaging has helped establish that vascular risk factors also lead to Alzheimer’s disease via pathways that are independent from vascular brain injuries, at least, when vascular brain injuries are measured as white matter lesions and infarcts. While vascular brain injuries (white matter lesions and infarcts) do not seem to influence amyloid pathology, some evidence from amyloid imaging suggests that increased vascular risk is related to increased amyloid burden. Furthermore, while vascular brain injuries and amyloid have an additive and independent impact on brain integrity, vascular risk factors might potentiate the impact of amyloid on cortical thickness on brain regions vulnerable to Alzheimer’s disease. New research should further explore and confirm, or refute, possible interactions between amyloid and vascular risk factors on brain integrity and cognition. Neuroimaging tools used to assess vascular brain integrity should also be expanded. Measuring cortical blood flow or damage to the capillary system might, for instance, give insight about how vascular risk factors can be associated to amyloid burden and impact. These findings also stress the need for monitoring vascular risk factors in midlife as a strategy for Alzheimer’s disease prevention. .

Key words: Alzheimer’ disease, amyloid, vascular brain injuries, vascular risk factors, treatment.


How do vascular factors, such as vascular diseases or vascular brain injuries (VBI, also often called cerebrovascular disease), increase the risk of Alzheimer disease (AD)? Are Alzheimer and vascular pathologies independent diseases, or does the presence of one pathology influence the presence and the impact of the other? Can vascular risk factors be good preventive targets for AD, and if so, why and when should they be targeted? Although the answers to these questions remain unclear, they represent some of the oldest issues in understanding relationships between Alzheimer and vascular diseases. Neuroimaging has long helped us to detect and quantify brain vascular diseases. The more recent advent of amyloid imaging now permits the detection and quantitation of amyloid-beta (Aβ), permitting new types of studies to explore complex relationships between Alzheimer and vascular pathologies. The current review first presents a brief overview of knowledge about the association between AD pathology and vascular factors (both VBI and vascular risk factors) from epidemiology and autopsy studies. We subsequently address what has been learned since the advent of in-vivo Aβ imaging. Possible avenues for prevention and treatments are also explored along with future research directions. This review does not intend to be an exhaustive review of the literature, but more an overview of where we are and where we should go next.    


The cause(s) of Alzheimer’s disease

The major obstacle to AD prevention and treatment is that the cause(s) of the disease is still unknown. In 1991, it was proposed that cerebral amyloid deposition represents the key pathogenic mechanism of AD development (1). The amyloid hypothesis suggested that amyloid initiates a cascade of pathological events, including the overexpression of neurofibrillary tangles, that lead to neurodegeneration and cognitive decline (2). The amyloid hypothesis finds its strongest support in the several varieties of familial AD that invariably result from genetic mutations which influence amyloid accumulation. In late onset AD, however, the causes likely include a combination of genetic, environmental, and lifestyle factors that act in concert to influence individual risk for development of disease and its associated symptoms. Specifically, while amyloid deposition seems still to be a key feature of late disease, other factors moderate its impact on brain integrity and cognition. Also, because late onset AD patients do not have a genetic mutation that causes early Aβ production, other genetic and environmental factors must influence Aβ accumulation. Identifying these factors and understanding the mechanism by which they influence the risk of AD is important from a prevention point of view, but also to guide new drug development.

Vascular and cerebrovascular diseases as risk factors for Alzheimer’ disease: knowledge from epidemiology and autopsy research studies

Vascular risk factors such as hypertension, dyslipidemia and diabetes are well known risk factors for AD (3). When looking at the prevalence of vascular factors compared to other risk factors (Table 1), it is evident that vascular factors should be a particular target for AD prevention. Furthermore, individuals with multiple vascular risks have more than twice the risk of developing dementia associated with AD compared to elderly without vascular risk factors (4). These vascular risk factors lead to VBI (e.g. white matter lesions and infarcts), which also increase the likelihood of dementia (3, 5). In fact, autopsy studies suggest that the most prevalent cause of dementia is mixed dementia, often defined by the presence of Alzheimer plus vascular pathologies (6). Autopsy studies further suggest that, while about a quarter of people can be free from dementia when presenting with Alzheimer pathology with no other comorbidity, very few persons (less than 7%) can stay free from dementia when both Alzheimer and vascular pathologies are present (6). Interestingly, autopsy studies also showed that less severe Alzheimer’s pathology is needed to develop Alzheimer’s dementia in the presence of infarcts or white matter lesions (7). Given the strong co-occurrence between both diseases, Alzheimer’s and vascular dementia are often presented as a continuum: with pure Alzheimer’s or vascular dementia representing the two extremes, and ‘mixed’ dementia in between and representing most older people with dementia.

Table 1. Risk factors for Alzheimer’ disease

Presented are common risk factors for AD; 1. Factors that have additionally been associated with increased brain Aβ. For hypercholesterolemia, both low HDL and high LDL cholesterol, but not total cholesterol, have been associated with increase Aβ (34). Aggregate vascular risk has also been associated with increased Aβ (33).

Because both pathologies frequently occur together, it is a major challenge to assign the degree of importance to either of them with regard to their effects on brain and cognitive integrity. Before Aβ-imaging, assessing the respective impact of both pathologies was only possible in autopsy-defined groups. However, even with the availability of autopsy data, or now with the availability of quantitative measures of Aβ deposition, assigning a role to each pathology when they are mixed is problematic. This is because such effects likely depend on the amount of each pathology, the length of time the pathology has been present, the location of pathology (particularly true for cerebrovascular disease which can be more focal than Aβ), and many aspects of the individual subject’s genetic, medical, and environmental background that could increase or limit susceptibility to each pathological process. Another challenging question, based on the strong associations between Alzheimer’s and vascular pathologies, is whether the impact of both pathologies are independent and additive, or if the presence of one pathology influences the presence and the impact of the other. It is possible that 1) both pathologies share common drivers (i.e. age, Apolipoprotein E (ApoE)) but act via independent pathways, 2) that one pathology drives the other pathology and/or 3) that both pathologies interact and that the join effect of both pathologies on brain and cognition is greater than their sum. Autopsy studies reported insufficient data supporting a direct link (options 2 and 3 above) between Alzheimer and vascular pathologies (8). It has therefore been assumed that both diseases occur and act independently, and additively increase the risk of dementia. Figure 1 schematises these independent pathways.

Figure 1. Independent Alzheimer and vascular pathways: an autopsy based model

Autopsy studies suggest that Alzheimer and vascular pathologies increase the risk of AD via independent and additive pathways. Because both pathologies frequently co-occur and because vascular risk factors such as hypertension and diabetes are well known risk factors for AD, mixed dementia is often considered the most frequent type of dementia. VBI : vascular brain injuries. Aβ: amyloid-beta


Aβ imaging

In 2004, the first in-vivo radiotracer to specifically track brain Aβ was reported (9). The Pittsburgh Compound B (PIB)-PET tracer is a 11C radiotracer that binds to fibrillar deposits of Aβ protein in plaques and cerebrovascular amyloid (CAA). Since then several 18F-labeled (half-life of 110 min) compounds have been created. Using these radiotracers it is now possible to track brain and cognitive changes associated with “pure” Alzheimer or vascular dementia, as well as subtle cognitive changes that are independent from both pathologies, which might include what is often termed normal aging. It is also possible to assess the relationship between Aβ, VBI and vascular risk factors in-vivo and test if Aβ and vascular factors act via independent or common pathways.

Figure 2. Alzheimer and vascular independent and shared pathways : an in-vivo based model

Proposed conception of the relationship between AD and vascular factors. While Aβ burden and vascular brain injuries (VBI, white matter lesions and infarcts) still have distinct pathways, vascular risk factors are associated with both Aβ burden and VBI. Vascular risk factors are also associated with brain integrity via a pathway that is independent from Aβ and VBI. Accordingly, vascular risk factors should be a particular target for prevention. Aβ: amyloid-beta. BBB: blood brain carrier. CBF: cerebral blood flow


Aβ and vascular brain injuries: independent or dependant pathways?

Supporting autopsy findings, many in-vivo studies assessing a relationship between Aβ and VBI (white matter lesions or infarcts) found no or slight correlation between the two factors in cognitively normal older adults, or older adults in preclinical or clinical phases of AD (10-18), even though increased white matter lesions have sometimes been reported in AD patients (14, 15).  Increased PIB-PET signal has in turn been associated with increased white matter lesions in persons presenting cerebral amyloid angiopathy (CAA) (19). Therefore CAA might have a stronger relationship with VBI than parenchymal Aβ. Whether transient Aβ increase follows an acute vascular event in humans, as has been suggested in rodents (20), still needs to be tested.

Concerning the impact of Aβ and VBI on brain and cognitive integrity, it seems that both factors mainly act via independent pathways, which is also in line with autopsy studies. Lower cerebrospinal fluid Aβ (which is inversely associated with brain Aβ) has been associated with decreased temporoparietal metabolism while greater white matter lesions have been associated with decreased frontal metabolism in individuals with mild cognitive impairment that subsequently progressed to dementia (21). Hippocampal volume and precuneus thickness have further been found to mediate (account for) the relationship between Aβ and memory (22-24), while frontal thickness has been reported to mediate the relationship between VBI and executive function (22) in cognitively impaired patients. These results do not imply that VBI cannot target brain regions typically affected by AD pathology (18, 25), but they suggest that VBI has a predominant impact on frontal functions. Similarly, while VBI is primarily associated with executive dysfunctions, it is not restricted to them, or to the impact of frontal-executive dysfunctions on other cognitive domains (10, 11, 26).  

The association between white matter structural integrity, measured with diffusion tensor imaging (DTI), and Aβ needs to be further explored given the inconsistent results reported in the literature (13, 27). Furthermore, even an association between these DTI changes and white matter lesions (13, 27) does not exclude the possibility that they are not all from vascular origin. The question of whether VBI potentiates the association between Aβ and functional connectivity, or if VBI and Aβ affect different brain networks, also needs further exploration. Indeed, while evidence suggests a link between Aβ and brain network functions measured with functional MRI (26), the independent or shared impact of white matter lesions on brain connectivity is unknown.

Figure 3. Impact of Aβ, VBI and vascular risk factors on cortical thickness in older adults with a spectrum of vascular diseases

Legend: Statistical cortical maps showing the association among Aβ, VBI (white matter hyperintensity), vascular risk (FCRP score) and cortical thickness in a sample of 66 older (64 for VBI) adults enriched for vascular diseases. Results suggest that increased vascular risk, increased Aβ burden and increased VBI are associated with thinner cortex. Statistical surface maps were created using a vertex-wise statistical thresholds of p < 0.05. The analyses are corrected for age, cognitive status, and multiple comparisons. This figure is based on a previously publication (38) 

With the constant improvement of neuroimaging tools, it is now possible to go beyond the assessment of WML/infarcts (or DTI) and explore other cerebrovascular mechanisms that might be related to Aβ. Indeed the multiple pathways by which amyloid and vascular factors could be linked do not necessary involve white matter lesions or infarcts (3, 28, 29), a fact that should be keep in mind and further explored. A recent study suggested for instance an association between Aβ and lower cerebral blood flow assessed using MRI-based arterial spin labelling (30). One explanation might be that lower blood flow diminishes Aβ clearance, which in turn reduces cerebral blood flow via a harmful vicious cycle (28). Assessing the integrity of the blood brain barrier using an MR contrast (28), brain vasoreactivity using carbon dioxide inhalation (31), or cerebral blood volume (as a proxy of capillary density) using a contrast agent and functional MRI (32), in relationship to Aβ would also be of interest. Even if still difficult to examine using existing neuroimaging tools, assessing the link between Aβ and microinfarcts might lead to new insight about the relationship between vascular factors and Aβ.

Aβ and vascular risk factors: independent or dependant pathways?

Although VBI and vascular risk factors, such as hypertension, cholesterol and diabetes, are linked, vascular risk factors can occur in the absence of VBI and vice versa. Vascular risk factors and VBI should therefore be considered and treated as two separate entities. While no clear association has been found between Aβ and VBI, there is strong evidence suggesting that vascular risk factors (aggregate or independent risk) are associated with increased brain Aβ (33-36). Importantly, some of these observations were found in late middle age subjects (36), suggesting that intervention targeting vascular risk factors should probably be started in midlife. Supporting that idea, the impact of vascular risk factors on brain integrity can already be detected in young adults (37). While the process by which vascular risk factors might lead to Aβ are mainly unknown, assessing these “other cerebrovascular mechanisms” are of interest since changes in cerebral blood flow, diminution of blood brain barrier permeability and vascular oxidative metabolism are all possible mechanisms by which vascular risk factors might increase Aβ burden (28).

In one of our previous studies we suggested that vascular risk factors interact with Aβ to reduce cortical thickness in brain regions known to be vulnerable to AD (38). This observation was independent of VBI and found when looking at aggregate vascular burden (assessed using the total Framingham cardiovascular risk profile, FCRP, score) or levels of circulating high-density lipoprotein (HDL) cholesterol. These data suggest that the impact of Aβ on cortical thickness might be potentiated by the presence of vascular burden and/or vice versa. In this same study, we also presented results suggesting that vascular risk factors can be associated with cortical thinning independently of Aβ and VBI. Therefore, vascular risk factors could influence AD risk via at least three pathways: 1) by increasing VBI, 2) by facilitating Aβ burden (and having a synergistic effect with it on brain integrity), and 3) by direct effects on the brain independently of Aβ and VBI (Figure 2). This last pathway should not be neglected as vascular risk factors can start early in life and therefore probably have a wide spread impact by the time a person reaches 65 years old, as suggested in Figure 3. Even if vascular risk factors do not lead to dementia by themselves, they probably diminish the “brain reserve”, conceptualised as a buffer that allows individuals to stay free from cognitive impairment in the presence of brain pathology. These “frail” brains might also be more vulnerable to other brain pathologies, as the interaction with Aβ suggests (38).

In this same study, it was also suggested that cholesterol-lowering medications might be protective against the negative impact of vascular risk factors and Aβ on cortical thickness. Both higher FCRP and higher Aβ burden were associated with less cortical thinning in subjects that were taking cholesterol-lowering drugs when compared with subjects who were not taking cholesterol drugs. This finding, which needs replication, is in line with other studies suggesting that statins confer some level of neuro-protection against late-life development of AD (39, 40, see also 41). Given that statin treatment has shown no reliable effect on clinical symptoms in subjects with dementia, it is more than plausible that statins only have an impact when started in midlife. Also, not all classes of statins necessarily confer the same protective benefit (40), an effect that needs to be better understood. 

Table 2. Preventive and treatment targets for Alzheimer’ disease

Presented are prevention and treatment targets for AD.  This table is intended to present avenues that should be explored and is not restricted to available treatments or treatments that have been found to be beneficial. 


Apolipoprotein E, Aβ and vascular factors

ApoE is a well-known genetic risk factor for AD (3, 42), with ~ 60% of AD patients presenting at least one ε4 allele (43). Interestingly, ApoE seems to be a common upstream driver to both Aβ and vascular burden, reinforcing the association between these two factors. ApoE, for example has been suggested to play a key role in Aβ accumulation and clearance, with ApoE4 being associated with increased Aβ burden (44) and ApoE2 being associated with lower Aβ burden (45). Because of its role in lipid metabolism regulation, ApoE4 also influences vascular risk factors and cardiovascular diseases (46), which in turn affects the risk of AD, as presented previously. Other mechanisms by which ApoE4 might influence the clinical expression of AD include neuronal inflammation, less efficient neuronal repair, diminished blood barrier integrity, increased tau phosphorylation, neurofibrillary tangle formation, neuronal mitochondrial dysfunction, and decreased GABAergic interneuron selectivity (47, 48). Given its wide range of functions, ApoE is probably a key factor to target for AD prevention and treatment.

Alzheimer’s disease prevention and treatment

Because the disease starts up to 30 years before the onset of dementia (49), and because vascular risk factors already impair the brain in middle age (37), preventive strategies for AD (table 2) should be implemented as early as possible. Most of these strategies should also be adopted in late life as they may still confer a benefit. For instance, while monitoring vascular risk may be a good prevention target in midlife, treating vascular risk in late life as been found to improve cognition in individuals with mild cognitive impairment (50).

In addition to targeting vascular risk factors, one obvious treatment target for AD is anti-amyloid therapies. Even if these therapies failed in dementia patients, they might have beneficial impact at preclinical or presymptomatic stages of the disease (51). Clinical trials enrolling subjects with autosomal-dominant familial AD and cognitively normal amyloid-positive older adults from the general population are currently ongoing. Given the possible vascular side effect of anti-amyloid therapies (52), vascular brain health will be monitored closely in these trials as they may predict adverse side effects.

Several other avenues should be tested for AD prevention and treatment in addition to Aβ and vascular therapies as the disease is multifactorial. More importantly research should be done to assess the mechanism by which these other factors can diminish the risk of AD since it might lead to new treatments. For instance, depression is a well-known risk factor for dementia. More recently it was suggested that individuals with a lifetime history of depression present increased brain Aβ (53). While depression might be secondary to Aβ accumulation, it is also important to assess if anti-depresant medication can slow Aβ accumulation, and hopefully AD progression, as was recently suggested (54). Similarly, enhanced lifetime cognitive activity has been shown to buffer the effect of ApoE4 on Aβ burden (55). While this information is valuable by itself for preventative strategies, understanding the mechanism by which cognitive activity might influence Aβ burden could point to new treatment strategies. Such strategies could include approaches such as the antiepileptic levetiracetam (56) (assuming that cognitive activity attenuate Aβ secretion via modulation of neural activity) or cognitive training protocols (57).

As mentioned previously, ApoE is a major risk factors for both Alzheimer and vascular pathologies. Increasing effort should therefore focus on developing and testing drugs that modify ApoE expression and function. Promoting ApoE levels, increasing ApoE receptor 2, blocking domain interaction in ApoE4, or restoring brain vascular integrity are all potential interesting targets (48, 58-60).


The absence of a relationship between Aβ and VBI (here defined as white matter lesions and infarcts), as well as their independent impact on cognition and brain integrity, suggests that both factors mainly act via independent pathways. Vascular risk factors however, seem to have a more direct impact on AD since increased vascular risk factors have been associated with increased Aβ burden. Furthermore, increased vascular risk factors might potentiate the impact of Aβ burden on cortical thickness. Given these findings, and the fact that vascular risk factors are often treatable, they should represent key factors for prevention.

Finally, while everyone is impatiently awaiting the results of anti-amyloid therapies in asymptomatic individuals, other treatments strategies should also be targeted. Among them, drugs that modify ApoE metabolism and function might be promising. Effort should also be made to understand how protective and risk factors such as lifestyle, psychiatric symptoms and sleep, influence AD risk.

Acknowledgements: The authors would like to thank J. Vogel for his review of the manuscript prior to publication.

Funding: SV is supported by a Canadian Institutes of Health Research post-doctoral fellowship and WJJ is supported by a NIH grants AG034570.

Conflict of interest: The authors report no conflict of interest.


1. Selkoe DJ. The Molecular Pathology of Alzheimer’s Disease. Neuron 1991;6:487-498.

2. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 2002;297:353-356.

3. Kalaria RN, Akinyemi R, Ihara M. Does vascular pathology contribute to Alzheimer changes? Journal of the neurological sciences 2012;322:141-147.

4. Luchsinger JA, Reitz C, Honig LS, Tang MX, Shea S, Mayeux R. Aggregation of vascular risk factors and risk of incident Alzheimer disease. Neurology 2005;65:545-551.

5. Schneider JA, Wilson RS, Bienias JL, Evans DA, Bennett DA. Cerebral infarctions and the likelihood of dementia from Alzheimer disease pathology. Neurology 2004;62:1148-1155.

6. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology 2007;69:2197-2204.

7. Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Markesbery WR. Brain infarction and the clinical expression of Alzheimer disease. The Nun Study. JAMA 1997;277:813-817.

8. Chui HC, Zheng L, Reed BR, Vinters HV, Mack WJ. Vascular risk factors and Alzheimer’s disease: are these risk factors for plaques and tangles or for concomitant vascular pathology that increases the likelihood of dementia? An evidence-based review. Alzheimers Res Ther 2012;3:36.

9. Klunk WE, Engler H, Nordberg A, et al. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh Compound-B. Annals of neurology 2004;55:306-319.

10. Hedden T, Mormino EC, Amariglio RE, et al. Cognitive profile of amyloid burden and white matter hyperintensities in cognitively normal older adults. The Journal of neuroscience : the official journal of the Society for Neuroscience 2012;32:16233-16242.

11. Marchant NL, Reed BR, DeCarli CS, et al. Cerebrovascular disease, beta-amyloid, and cognition in aging. Neurobiology of aging 2012;33:1006 e1025-1036.

12. Marchant NL, Reed BR, Sanossian N, et al. The Aging Brain and Cognition: Contribution of Vascular Injury and Abeta to Mild Cognitive Dysfunction. JAMA Neurol 2013:1-8.

13. Chao LL, Decarli C, Kriger S, et al. Associations between white matter hyperintensities and beta amyloid on integrity of projection, association, and limbic fiber tracts measured with diffusion tensor MRI. PloS one 2013;8:e65175.

14. Provenzano FA, Muraskin J, Tosto G, et al. White matter hyperintensities and cerebral amyloidosis: necessary and sufficient for clinical expression of Alzheimer disease? JAMA Neurol 2013;70:455-461.

15. Zhou Y, Yu F, Duong TQ, for the Alzheimer’s Disease Neuroimaging I. White matter lesion load is associated with resting state functional MRI activity and amyloid pet but not FDG in mild cognitive impairment and early alzheimer’s disease patients. Journal of magnetic resonance imaging : JMRI 2013;9:102-109..

16. Park JH, Seo SW, Kim C, et al. Effects of cerebrovascular disease and amyloid beta burden on cognition in subjects with subcortical vascular cognitive impairment. Neurobiology of aging 2014;35:254-260.

17. Grimmer T, Faust M, Auer F, et al. White matter hyperintensities predict amyloid increase in Alzheimer’s disease. Neurobiology of aging 2012;33:2766-2773.

18. Hedden T, Schultz AP, Rieckmann A, et al. Multiple Brain Markers are Linked to Age-Related Variation in Cognition. Cerebral cortex 2014; pii:bhu238. [Epub ahead of print]

19. Gurol ME, Viswanathan A, Gidicsin C, et al. Cerebral amyloid angiopathy burden associated with leukoaraiosis: a positron emission tomography/magnetic resonance imaging study. Annals of neurology 2013;73:529-536.

20. Garcia-Alloza M, Gregory J, Kuchibhotla KV, et al. Cerebrovascular lesions induce transient beta-amyloid deposition. Brain : a journal of neurology 2011;134:3697-3707.

21. Haight TJ, Landau SM, Carmichael O, et al. Dissociable effects of Alzheimer disease and white matter hyperintensities on brain metabolism. JAMA Neurol 2013;70:1039-1045.

22. Ye BS, Seo SW, Kim GH, et al. Amyloid burden, cerebrovascular disease, brain atrophy, and cognition in cognitively impaired patients. Alzheimer’s & dementia : the journal of the Alzheimer’s Association 2014; ;pii: S1552-5260(14)02419-4. [Epub ahead of print]

23. Mormino EC, Kluth JT, Madison CM, et al. Episodic memory loss is related to hippocampal-mediated beta-amyloid deposition in elderly subjects. Brain : a journal of neurology 2009;132:1310-1323.

24. Villeneuve S, Reed BR, Wirth M, et al. Cortical thickness mediates the effect of beta-amyloid on episodic memory. Neurology 2014;82:761-767.

25. Wirth M, Villeneuve S, Haase CM, et al. Associations between Alzheimer disease biomarkers, neurodegeneration, and cognition in cognitively normal older people. JAMA Neurol 2013;70:1512-1519.

26. Jagust W. Vulnerable neural systems and the borderland of brain aging and neurodegeneration. Neuron 2013;77:219-234.

27. Glodzik L, Kuceyeski A, Rusinek H, et al. Reduced glucose uptake and Abeta in brain regions with hyperintensities in connected white matter. NeuroImage 2014;100:684-691.

28. Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nature reviews Neuroscience 2011;12:723-738.

29. de la Torre JC. Is Alzheimer’s disease a neurodegenerative or a vascular disorder? Data, dogma, and dialectics. Lancet neurology 2004;3:184-190.

30. Mattsson N, Tosun D, Insel PS, et al. Association of brain amyloid-beta with cerebral perfusion and structure in Alzheimer’s disease and mild cognitive impairment. Brain : a journal of neurology 2014;137:1550-1561.

31. Glodzik L, Randall C, Rusinek H, de Leon MJ. Cerebrovascular reactivity to carbon dioxide in Alzheimer’s disease. Journal of Alzheimer’s disease : JAD 2013;35:427-440.

32. Brickman AM, Khan UA, Provenzano FA, et al. Enhancing dentate gyrus function with dietary flavanols improves cognition in older adults. Nature neuroscience 2014;17:1798-1803.

33. Reed BR, Marchant NL, Jagust WJ, DeCarli CC, Mack W, Chui HC. Coronary risk correlates with cerebral amyloid deposition. Neurobiology of aging 2012;33:1979-1987.

34. Reed B, Villeneuve S, Mack W, Decarli C, Chui HC, Jagust W. Associations between serum cholesterol levels and cerebral amyloidosis. JAMA Neurol 2014;71:195-200.

35. Rodrigue KM, Rieck JR, Kennedy KM, Devous MD, Diaz-Arrastia R, Park DC. Risk Factors for beta-Amyloid Deposition in Healthy Aging: Vascular and Genetic Effects. JAMA Neurol 2013:1-7.

36. Langbaum JB, Chen K, Launer LJ, et al. Blood pressure is associated with higher brain amyloid burden and lower glucose metabolism in healthy late middle-age persons. Neurobiology of aging 2012;33:827 e811-829.

37. Maillard P, Seshadri S, Beiser A, et al. Effects of systolic blood pressure on white-matter integrity in young adults in the Framingham Heart Study: a cross-sectional study. Lancet neurology 2012;11:1039-1047.

38. Villeneuve S, Reed BR, Madison CM, et al. Vascular risk and Abeta interact to reduce cortical thickness in AD vulnerable brain regions. Neurology 2014;83:40-47.

39. Jick H, Zornberg GL, Jick SS, Seshadri S, Drachman DA. Statins and the risk of dementia. Lancet 2000;356:1627-1631.

40. Wolozin B, Wang SW, Li NC, Lee A, Lee TA, Kazis LE. Simvastatin is associated with a reduced incidence of dementia and Parkinson’s disease. BMC medicine 2007;5:20.

41. Leduc V, De Beaumont L, Theroux L, et al. HMGCR is a genetic modifier for risk, age of onset and MCI conversion to Alzheimer’s disease in a three cohorts study. Molecular psychiatry 2014doi: 10.1038/mp.2014.81. [Epub ahead of print]

42. Mahley RW, Huang Y. Alzheimer disease: multiple causes, multiple effects of apolipoprotein E4, and multiple therapeutic approaches. Annals of neurology 2009;65:623-625.

43. Ward A, Crean S, Mercaldi CJ, et al. Prevalence of apolipoprotein E4 genotype and homozygotes (APOE e4/4) among patients diagnosed with Alzheimer’s disease: a systematic review and meta-analysis. Neuroepidemiology 2012;38:1-17.

44. Strittmatter WJ, Saunders AM, Schmechel D, et al. Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A 1993;90:1977-1981.

45. Jiang Q, Lee CY, Mandrekar S, et al. ApoE promotes the proteolytic degradation of Abeta. Neuron 2008;58:681-693.

46. Villeneuve S, Brisson D, Marchant NL, Gaudet D. The potential applications of Apolipoprotein E in personalized medicine. Frontiers in aging neuroscience 2014;6:154.

47. Mahley RW, Huang Y. Apolipoprotein e sets the stage: response to injury triggers neuropathology. Neuron 2012;76:871-885.

48. Liu CC, Kanekiyo T, Xu H, Bu G. Apolipoprotein E and Alzheimer disease: risk, mechanisms and therapy. Nature reviews Neurology 2013;9:106-118.

49. Villemagne VL, Burnham S, Bourgeat P, et al. Amyloid beta deposition, neurodegeneration, and cognitive decline in sporadic Alzheimer’s disease: a prospective cohort study. Lancet neurology 2013;12:357-367.

50. Chertkow H, Massoud F, Nasreddine Z, et al. Diagnosis and treatment of dementia: 3. Mild cognitive impairment and cognitive impairment without dementia. CMAJ : Canadian Medical Association journal = journal de l’Association medicale canadienne 2008;178:1273-1285.

51. Sperling RA, Jack CR, Jr., Aisen PS. Testing the right target and right drug at the right stage. Science translational medicine 2011;3:111cm133.

52. Yamada M. Predicting cerebral amyloid angiopathy-related intracerebral hemorrhages and other cerebrovascular disorders in Alzheimer’s disease. Frontiers in neurology 2012;3:64.

53. Wu KY, Hsiao IT, Chen CS, et al. Increased brain amyloid deposition in patients with a lifetime history of major depression: evidenced on 18F-florbetapir (AV-45/Amyvid) positron emission tomography. European journal of nuclear medicine and molecular imaging 2014;41:714-722.

54. Sheline YI, West T, Yarasheski K, et al. An antidepressant decreases CSF Abeta production in healthy individuals and in transgenic AD mice. Science translational medicine 2014;6:236re234.

55. Wirth M, Villeneuve S, La Joie R, Marks SM, Jagust WJ. Gene-environment interactions: lifetime cognitive activity, APOE genotype, and beta-amyloid burden. The Journal of neuroscience : the official journal of the Society for Neuroscience 2014;34:8612-8617.

56. Bakker A, Krauss GL, Albert MS, et al. Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron 2012;74:467-474.

57. Belleville S, Clement F, Mellah S, Gilbert B, Fontaine F, Gauthier S. Training-related brain plasticity in subjects at risk of developing Alzheimer’s disease. Brain : a journal of neurology 2011;134:1623-1634.

58. Bu G. Apolipoprotein E and its receptors in Alzheimer’s disease: pathways, pathogenesis and therapy. Nature reviews Neuroscience 2009;10:333-344.

59. Champagne D, Pearson D, Dea D, Rochford J, Poirier J. The cholesterol-lowering drug probucol increases apolipoprotein E production in the hippocampus of aged rats: implications for Alzheimer’s disease. Neuroscience 2003;121:99-110.

60. Mahley RW, Huang Y. Small-molecule structure correctors target abnormal protein structure and function: structure corrector rescue of apolipoprotein E4-associated neuropathology. Journal of medicinal chemistry 2012;55:8997-9008.