jpad journal

AND option

OR option


P. de Souto Barreto1,2, K. Pothier3, G. Soriano1,2, M. Lussier4,5, L. Bherer4,5,6, S. Guyonnet1,2, A. Piau1,2, P.-J. Ousset1, B. Vellas1,2

1. Gerontopole of Toulouse, Institute of Ageing, Toulouse University Hospital (CHU Toulouse), Toulouse, France; 2. UPS/Inserm UMR1027, University of Toulouse III, Toulouse, France; 3. University of Tours, EA 2114, PAVEA Laboratory, Tours, France; 4. Département de Médecine, Université de Montréal, Montréal, Canada; 5. Centre de recherche de l’Institut universitaire de gériatrie de Montréal, Montréal, Canada; 6. Centre de recherche de l’Institut de cardiologie de Montréal. Montréal, Canada

Corresponding Author: Professor Philipe de Souto Barreto, Gérontopôle de Toulouse, Institut du Vieillissement, 37 Allées Jules Guesde, F-31000 Toulouse, France, Phone: (+33) 561 145 668, Fax: (+33) 561 145 640, e-mail:

J Prev Alz Dis 2020;
Published online December 4, 2020,



Importance/Objective: To describe the feasibility and acceptability of a 6-month web-based multidomain lifestyle training intervention for community-dwelling older people and to test the effects of the intervention on both function- and lifestyle-related outcomes.
Design: 6-month, parallel-group, randomized controlled trial (RCT).
Setting: Toulouse area, South-West, France.
Participants: Community-dwelling men and women, ≥ 65 years-old, presenting subjective memory complaint, without dementia.
Intervention: The web-based multidomain intervention group (MIG) received a tablet to access the multidomain platform and a wrist-worn accelerometer measuring step counts; the control group (CG) received only the wrist-worn accelerometer. The multidomain platform was composed of nutritional advices, personalized exercise training, and cognitive training.
Main outcomes and measures: Feasibility, defined as the proportion of people connecting to ≥75% of the prescribed sessions, and acceptability, investigated through content analysis from recorded semi-structured interviews. Secondary outcomes included clinical (eg, cognitive function, mobility, health-related quality of life (HRQOL)) and lifestyle (eg, step count, food intake) measurements.
Results: Among the 120 subjects (74.2 ±5.6 years-old; 57.5% women), 109 completed the study (n=54, MIG; n=55, CG). 58 MIG subjects connected to the multidomain platform at least once; among them, adherers of ≥75% of sessions varied across multidomain components: 37 people (63.8% of 58 participants) for cognitive training, 35 (60.3%) for nutrition, and three (5.2%) for exercise; these three persons adhered to all multidomain components. Participants considered study procedures and multidomain content in a positive way; the most cited weaknesses were related to exercise: too easy, repetitive, and slow progression. Compared to controls, the intervention had a positive effect on HRQOL; no significant effects were observed across the other clinical and lifestyle outcomes.
Conclusions and Relevance: Providing multidomain lifestyle training through a web-platform is feasible and well-accepted, but the training should be challenging enough and adequately progress according to participants’ capabilities to increase adherence. Recommendations for a larger on-line multidomain lifestyle training RCT are provided.

Key words: Multimodal lifestyle intervention, exercise, cognitive stimulation, nutritional advice, web-based intervention.



In recent years, the multidomain strategy, an approach characterized by the combination of several lifestyle interventions, have received increasing attention, with the development of large randomized controlled trial (RCT) by our team (1) and others (2, 3). The rationale behind the multidomain lifestyle strategy is simple: if interventions, such as physical exercise, healthy nutrition, and cognitive stimulation, that have already proven their effectiveness for improving/maintaining individual’s health during aging are combined, the benefits will then be potentiated.
Although several studies have investigated the effects of multidomain lifestyle strategies on functional-related outcomes during aging (4), mixed findings were obtained. The interventions tested so far are difficult to transpose into actual clinical practice, mainly because they are burdensome (eg, participants must visit research facilities several times a week or month) and expensive (eg, they require specialized professionals, such as exercise instructors). Several multidomain lifestyle platforms exist (5), however, to the best of our knowledge, none of them tested a training (not just health information, counselling or motivation) intervention, dedicated to older adults, through a randomized controlled trial (RCT) design. The advantages of having an online multidomain lifestyle training platform for which the content and procedures have proven their feasibility, acceptability, and efficacy, are multiple, since such a platform may be accessed at any time by the participants, being adapted to individual’s time constraints, and does not need the physical presence of both participants and specialized professionals. These advantages are still more important in the current context of social isolation and population containment caused by the COVID-19 pandemics.
The main purposes of the present article were to describe the feasibility and acceptability of a 6-month randomized controlled trial (RCT) of a web-based multidomain lifestyle training intervention for community-dwelling older people with spontaneous memory complaint. Secondary objectives were to test the effects of the intervention on both function- (eg, cognition, mobility) and lifestyle-related (eg, PA, food intake) outcomes.



A detailed description of the eMIND trial has been published elsewhere (6). eMIND is a 6-month pilot, parallel-arm, RCT of a multidomain lifestyle intervention composed of cognitive training, exercise training, and nutritional advices, among community-dwelling older adults from Toulouse area, South-West, France. The protocol was approved by the ethic committee of Tours (CPP Tours; 2017T2-10); the first recruitment occurred in December 2017 and the last study visit in September 2019. The trial was registered in a publicly accessible registry (; NCT03336320). All participants signed an informed consent before undertaking study procedures.


Inclusion criteria were: ≥ 65 years-old; Mini-Mental State Examination (MMSE(7)) ≥ 24; presenting subjective memory complaints; easy access to internet (Internet access at least twice a week). Exclusion criteria were: illness with life expectancy less than six months; diagnosis of dementia according with DSM-V; diagnosis of neurodegenerative diseases, particularly Parkinson; major depression; any health condition potentially deteriorated by exercise; dependency in ≥ 1 activity of daily living (basic ADL); already participating in exercise or cognitive training ≥ 2 times/week in the last 2 months.
One hundred-twenty participants were enrolled in eMIND (74.2 years-old ± 5.6; 57.5% women).

Randomization and masking

Participants were randomized after baseline assessments in a 1:1 ratio to either multidomain intervention group (MIG) or control group (CG) using a dedicated at-distance website. Concealment of group allocation was warranted by using opaque envelops, stored in a safe and locked place. Outcome assessors were blinded to group assignment.


Figure 1 displays the main procedures of eMIND. Participants randomized to MIG received a tablet (model: HP x2 210 G2 – 10.1) providing access to a secured, password-encrypted platform that respected all the laws and regulations in France. Data was stored in an approved database for health data. MIG participants also received a commercial wrist-worn accelerometer (model: FitBit Flex 2) that provides objectively measured step counts. The accelerometer was synchronized with the tablet. Participants could access the platform for as many times as they wanted.

Figure 1. Flow of the eMIND study procedures

After baseline assessments and randomization, MIG received a tablet with access to the multidomain platform for six months and a wrist-worn accelerometer; CG received the wrist-worn accelerometer and information on multidomain activities available in the website of the Toulouse University Hospital. The lowest part of the figure (MIG side) displays examples of how participants received the multidomain lifestyle training in their tablets, with cognitive training (eg, different Stroop tasks and N-back), and videos for both exercise (eg, chair rise – for illustration purposes, we put sit and up-right positions together, but participants had videos) and nutritional advice (eg, proteins).


The web multidomain platform focused on three lifestyles: nutritional advice, and exercise and cognitive training. The platform was equipped with a chat, to facilitate communication of participants with the research team, a personalized agenda showing the day-by-day activities (eg, exercise and cognitive training to be done, nutritional advices), a library area where the content of the interventions and educational material on lifestyles were available. Participants were requested to follow both exercise and cognitive training twice a week, and nutritional advices every fifteen days; for that, they should only click on the activities displayed in their personal agenda. The content of each lifestyle is briefly described below.

Exercise training

Three different 6-month exercise programs, with increasingly challenging exercises, were proposed according to individual’s baseline physical function (according to the short physical performance battery (SPPB)). Several videos with different exercises were elaborated: eg, chair rise, walking in the line, flexions and extension in the knee, walking backwards, tiptoe standing, etc. Each exercise video had subtitles with big letters to facilitate the understanding. The exercise program was developed by an experienced exercise scientist on the basis of exercise principles (eg, frequency, intensity, load progression), focusing on the lower-body. A typical exercise session was composed of three different types of strength exercises, three balance exercises (the number of both sets and repetitions per set varied according to individual’s physical function), a specific advice on aerobic exercise (type, session duration, mode (continuous or in bouts), intensity and how to subjectively reach the self-perceived exertion), and a set of flexibility exercises for warm up and cooling down.

Cognitive training

A computerized cognitive training was provided using Neuropeak (, a password encrypted platform developed by researchers from the University of Montreal, Canada. Participants trained on cognitive tasks mainly related with executive functions, with a 6-month cognitive training pre-established personalized program with progressive difficulty; participants received active feedback encouraging them to perform beyond their baseline performances. Three types of tasks were performed: the dual-task, Stroop task, and N-back task. For the dual task (divided attention training), participants had to identify vehicles and fruits, both separately and concurrently (8). For the Stroop task (inhibition and switching training), participants were presented four different conditions: concordant, counting, discordant, and switching. For the N-back (updating training), digits were sequentially presented and the participant had to indicate if the digit was the same than the digit presented “n” steps earlier.

Healthy nutrition advice

Twelve videos (about five-to-eight minutes per video, two videos per month for the 6-month intervention) on nutritional advices were produced for this study by an experienced hospital dietitian from the department of geriatrics (Toulouse University Hospital), being organized under important nutritional topics for older adults, such as, proteins, fat, hydration, fruits and vegetables, calcium and osteoporosis, etc. Advices were based on the French recommendations from the “Programme National Nutrition Santé” (PNNS) ( A quiz with three questions was used to facilitate the retention of the key messages for each nutritional video. A personalized approach for nutritional advices had been planned for people at-risk of nutritional deficiency (mini-nutritional assessment(9) (MNA) ≤ 23.5).

Control group

All controls received the same accelerometer than MIG, but did not have access to the multidomain web-platform. They received a link to information on multidomain activities produced by the research team ( CG participants were asked to bring their own smartphone/tablet at the baseline visit and the research team helped them to synchronize the accelerometer with the smartphone/tablet.

Main outcome measures

For this pilot RCT, feasibility and acceptability of study procedures and tools were the main outcomes. Feasibility was assessed through the adherence to the multidomain protocol. Participants accessing all the three interventions (clicking on the multidomain contents in their personal agenda in the web-platform) for at least 75% of the requirements were considered adherers, confirming the feasibility of the study procedures. Acceptability was assessed through the content analysis of recorded semi-structured interviews performed at the post-intervention assessments. The main element of acceptability was defined by the question “In your opinion, can this web-platform be used in its current state?”, which was anchored by the following responses: No, it requires major changes for improving its easiness-of-use; No, but it requires only minor adaptations; Yes, but a few modifications could render it easier to use; Yes, the platform is easy-to-use in its current state. The main modifications proposed by participants to improve both intervention content and technological aspects were recorded and explored.

Secondary endpoints

Cognitive function

Assessed using the mean of a cognitive composite score (10) combining the Z-scores of four scales: MMSE (7) 10-items orientation, Digit Symbol Substitution Test of Wechsler Adult Intelligence Scale-Revised (DSST, WAIS-R) (11), total recall (up to 48 points) of the Free and Cued Selective Reminding test (FCSRT) (12), and Category Naming test (CNT). The following original scales were investigated separately: MMSE, DSST WAIS-R, total recall of FCSRT, CNT, and Controlled Oral Word Association Test (COWAT) (13).

Physical function

Measured using the SPPB (14), composed of three tests (4-meter usual pace gait speed, 5-repetition chair rise, and balance tests; score range from 0 to 12, higher is better), and the 4-meter gait speed (m/sec).

Depressive symptoms

Measured by the 15-item Geriatric Depression Scale (GDS-15)(15); scores vary from zero to 15, higher is worse.

Nutritional status

Assessed by the MNA (9); scores vary from zero to 30, higher is better.

Health-related quality of life (HRQOL)

Assessed using the Euro-QoL 5D-5L (16). We used two variables: index value (continuous variable varying from -1 to 1, higher is better) for the French population (17) calculated using the EuroQol calculator website ( and the visual analogue scale (VAS, varying from zero to 100, higher is better).

Physical Activity (PA)

This was assessed subjectively by a self-reported questionnaire (QAPPA (18, 19), continuous values of metabolic equivalent task/week (MET-min/week), higher means higher PA) and using step counts from the accelerometers. We used the mean steps/day from four waves of data collection: first week (baseline measure); first month (excluding the first week); month three; month six. Data on step count were considered valid if captured during ≥60% of the exposure (ie, ≥ 18 days/month) and ignoring values ≤1000 steps (20–22).

Leisure-time cognitive activities

Assessed through a 14-item self-reported questionnaire asking for the frequency of cognitive stimulating activities (eg, crosswords, cultural outings); scores range from 0 to 56, higher is better.

Food intake

Measured by a 21-item food frequency questionnaire (FFQ; scores vary from 0 to 21, higher is better) used in the clinical practice in the frailty day-hospital of the Toulouse University Hospital.

Statistical Analysis

The present pilot study was proposed to inform the development of a future, well-powered RCT of ICT multidomain lifestyle intervention; we estimated having 60 people per study arm, taking into account low adherence to multidomain interventions and 6% dropout, would allow us to have the needed data for a proper sample size calculation on clinical outcomes.
Analyses were performed on an intention-to-treat (ITT) basis. Descriptive statistics (mean (SD) and absolute numbers (%), as appropriate) were used. Content analysis of recorded semi-structured interviews performed at post-intervention were used to identify aspects to be improved in study procedures and tools. Baseline difference between MIG and CG were tested using Student t-test for independent samples and chi-square test, as appropriate. The effects of the intervention on the secondary outcome measures were assessed using mixed effect linear regressions, with group, time and group-by-time interaction as fixed terms, a random effect at participants’ level, and a random slope on time; participant’s age was used as a confounder in all analysis due to imbalance between groups.
Statistical significance was determined p < 0.05. All analyses were performed using Stata (v.14.0, Texas, USA).



Figure 2 show the flow of study participants. From the 120 participants randomized, 109 were assessed at 6-month (MIG, n=54; CG, n=55). Among the 58 adverse events registered, 12 were serious (eg, heart infarct), being four in MIG (n=3 subjects) and 8 in CG (n=6 subjects); among the 46 non-serious events (eg, intermittent dizziness), 32 occurred among MIG (n=24 subjects) and 14 in CG (n=10 subjects). No adverse event (both serious and non-serious) was related to study procedures according with study physicians.
Table 1 shows participants’ characteristics. Study groups were well-balanced, except that MIG participants were 2-year older than CG (p=0.052).

Figure 2. Flowchart of study participants


Table 1. Participants’ characteristics


Regarding feasibility, 58 (out of 60) participants in MIG connected to the multidomain platform at least once during the 6-month trial. Among them, adherers (≥75% of prescribed sessions) varied across the different components of the multidomain: 37 individuals (63.8%) for cognitive training, 35 (60.3%) for nutrition, and only three (5.2%) for exercise; these same three persons were adherers in all three multidomain components. Regarding exercise, 31 individuals (53.4%) followed ≥50% of the requested frequency (they connected once a week); for cognitive training and nutritional advices, they were 75.8% and 81%, respectively.
Regarding acceptability, we interviewed 53 participants from MIG: four (7.5%) said the multidomain web-platform was not ready to be used and needed major changes; three (5.7%) indicated it required minor changes; 18 (34%) said it was ready to be used, but minor modifications could render it easier to use; and 28 (52.8%) indicated the platform was ready to be used without any change. Among the strengths/weakness of the platform, although most participants interviewed (n=50, 94.3%) indicated the technical/technological aspects were simple, they reported technical interruptions during the cognitive training; one person indicated the letters were too small. Although all individuals interviewed indicated the content was clear, they found weaknesses, the most cited being the physical exercises were too easy (not challenging enough), repetitive (should vary more the exercises proposed), sometimes difficult to perform due to space limitations at home, and progressed slowly. To a lower extent, other weaknesses highlighted were: nutritional advices were lacking novelty; cognitive training was sometimes repetitive and too long. Some participants suggested that having more contacts with the research team would be helpful and a few indicated that the gamification (eg, motivating rewarding system, points) of the intervention could increase interest and adherence.
The effects of the intervention on clinical and lifestyle outcomes are displayed in Table 2. No statistically significant effects were found, except for the two variables (ie, index value and VAS) of HRQOL, showing MIG had an improved HRQOL compared to CG.
A set of recommendations for developing a large RCT on multidomain lifestyle training intervention is provided in Box 1. These recommendations were developed according with the lessons learnt from eMIND.

Table 2. Effects of the web-based multidomain lifestyle training intervention on both clinical and lifestyle secondary outcomes using mixed-effect linear regression

a. For all variables, except GDS-15, positive values in the within-group adjusted mean difference indicate improvement over time. For GDS-15, negative values indicate improvement; b. For all variables, except GDS-15, positive values in the between-group adjusted mean difference favor the multidomain intervention. For GDS-15, negative values favor the multidomain intervention

Box 1. Recommendations for developing a large RCT on multidomain lifestyle training: lessons from eMINDa

a. These recommendations should be seen as additional elements to the eMIND contents and procedures.



This is the first RCT using a web-based multidomain lifestyle-training intervention for community-dwelling older adults. We showed that study procedures were well-accepted, but expectations regarding participant’s adherence to the intervention were not met, in particular for exercise. Moreover, compared to controls, participants of MIG have improved their HRQOL at the end of the 6-month intervention.
Although most participants were not compliant to the protocol to the expected extent, raising questions about the feasibility of a larger trial, a few considerations are worth discussing. We arbitrarily defined good adherence as connecting to ≥75% of intervention sessions, which is a high rate, in particular for demanding activities, such as exercise (23). Indeed, the FINGER study, a previous multidomain RCT (2) with supervised exercise training, showed that less than 60% of participants performed at least half of the exercise sessions (24), which is a similar rate than ours (53.4% of MIG connected to half of expected sessions). FINGER also had part of the cognitive intervention performed at home through a web-based training system; they found that 47.2% of participants did at least half of the training sessions (24), which is much less than 75.8% observed in eMIND. In the multidomain MAPT trial our team conducted (1, 24), 53.5% of participants attended ≥75% of multidomain sessions (PA and nutritional counselling and cognitive training performed in the same sessions), which is similar to what we found for nutritional advices and cognitive training in eMIND, but not for exercise. Several methodological differences of FINGER and MAPT, compared to eMIND, must be mentioned: FINGER and MAPT were long-term (2- and 3-year, respectively) trials, with larger samples, and different modalities and frequency for providing the multidomain interventions; eMIND and MAPT have similar population, but were much older than FINGER. In the HATICE (25) web-based multidomain trial (focused on counselling and motivation through a coaching system), participants in the intervention group connected in average 1.8 times per month, during almost 18 months; HATICE’s participants were not required to connect on a regular frequency as in eMIND. The fact the study was well-accepted by participants, with clear and, most of time, useful contents is a major result. Therefore, considering the good acceptability and the acceptable adherence rates for cognitive training and nutritional advice, it is plausible to suggest the eMIND findings are promising. Although adherence to exercise was low, compliance of once/week exercise training was acceptable; this is important because doing PA once a week may already bring several health benefits during aging (26, 27).
However, weaknesses must be mentioned and corrected if a larger, well-powered trial should be developed. Indeed, the exercise program should comprise a more diversified set of exercises, with different levels of difficulty in the execution and adapted progression. For this, one possibility is to use a coaching system, like in the HATICE study (25), which would allow participants to interact with the research team through the web-platform in a shorter time interval and adapt the intervention program according to participant’s current status; this might lead to providing the most appropriate content to participants, reducing loss of motivation due to repetitive and not challenging enough exercises. Providing participants with inexpensive exercise materials (therabands, calf weights) may facilitate exercise load progression. Adapting the intervention content to avoid demotivation would also improve cognitive training. The coaching system would further help in making more interactive the nutritional advices component. If connected devices are to be used, an important workload should be foreseen to solve technical problems. Problems with the synchronization between accelerometer and tablets/smartphones in eMIND were frequent, being most of time related with participants’ low digital literacy; sometimes, problems were related to technical issues (requiring the support of IT professionals).
The absence of noticeable differences between groups for clinical outcomes was expected, since eMIND has not been powered to test the effectiveness of the intervention. Why MIG participants had a better HRQOL than CG, even if no statistically significant between-group differences were found in any of the clinical and lifestyle outcomes, deserves further investigation. Woo and colleagues (28) showed that a 24-week exercise and nutritional supplementation, compared to controls, improved self-reported health in middle-aged and young older adults. The effects of the isolated components of multidomain are better known, especially exercise, which has shown to improve HRQOL in older adults (29-33). Nutritional support was shown to have a positive effect on quality of life in older hospitalized patients (34). It is possible that alterations in patient-reported outcomes, such as HRQOL, which are subjective and then potentially more sensitive to change in a short-time interval, precede changes in more robust clinical outcomes, such as mobility. However, since the effects of eMIND intervention on HRQOL is of arguable clinical meaningfulness, it is also possible that the HRQOL improvement found would be temporary. A longer and well-powered trial would shed light on this topic.
The main strengths of our study were: the mixed method approach, composed of a RCT design and qualitative semi-structured interviews (for MIG); the use of a wearable step count tracker; and the fact this is the first web-based multidomain lifestyle-training intervention for older adults. Although other web-based multidomain trials (5) were performed, they focused on counselling and motivation or were developed in middle-aged/young older populations; other trials are on-going (35, 36), in particular the large and long-term MYB trial (36), which will provide a training platform for people 55-77 years-old, but all components of the multidomain (performed in short blocks of 10 weeks) will not be available to all participants. eMIND’s weaknesses were: the small sample; and short intervention length.
Implementing a web-based multidomain lifestyle-training platform accessible to a large number of older people could have a major positive impact in a public health perspective. This is still more relevant in periods of prolonged social isolation and containment, such as during the COVID-19 pandemics. The preliminary findings of this pilot RCT support the development of a larger, well-powered, long-term RCT to test the effectiveness of an adapted version (in particular, for exercise) of the eMIND platform on clinical outcomes.


Funding: This study is supported by the Fondation pour la Recherche Médicale (FRM DOC20161136208) in the context of the call “Evaluation of the impact of connected devices on health 2016”. The Centre Hospitalier Universitaire of Toulouse (CHU-Toulouse) is the sponsor of the study (protocol ID: 17 0071).

Role of Funding Source: None.

Conflict of interest: All authors declare no conflicts of interest related to this article.



1. Andrieu S, Guyonnet S, Coley N, Cantet C, Bonnefoy M, Bordes S, et al. Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): a randomised, placebo-controlled trial. Lancet Neurol. 2017 May;16(5):377–89.
2. Ngandu T, Lehtisalo J, Solomon A, Levälahti 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 Lond Engl. 2015 Jun 6;385(9984):2255–63.
3. Moll van Charante EP, Richard E, Eurelings LS, van Dalen J-W, Ligthart SA, van Bussel EF, et al. Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): a cluster-randomised controlled trial. Lancet Lond Engl. 2016 Aug 20;388(10046):797–805.
4. Fourteau P, Virecoulon Giudici K, Rolland Y, Vellas B, de Souto Barreto P. Associations between Multidomain lifestyle Interventions and Intrinsic Capacity domains during Aging: A Narrative Review. 2020;In Press.
5. Wesselman LM, Hooghiemstra AM, Schoonmade LJ, de Wit MC, van der Flier WM, Sikkes SA. Web-Based Multidomain Lifestyle Programs for Brain Health: Comprehensive Overview and Meta-Analysis. JMIR Ment Health [Internet]. 2019 Apr 9 [cited 2020 Sep 8];6(4). Available from:
6. Pothier K, Soriano G, Lussier M, Naudin A, Costa N, Guyonnet S, et al. A web-based multidomain lifestyle intervention with connected devices for older adults: research protocol of the eMIND pilot randomized controlled trial. Aging Clin Exp Res. 2018 Sep;30(9):1127–35.
7. 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.
8. Bherer L, Gagnon C, Langeard A, Lussier M, Desjardins-Crépeau L, Berryman N, et al. Synergistic effects of cognitive training and physical exercise on dual-task performance in older adults. J Gerontol B Psychol Sci Soc Sci. 2020 Aug 17;
9. Guigoz Y, Vellas B, Garry PJ. Assessing the nutritional status of the elderly: The Mini Nutritional Assessment as part of the geriatric evaluation. Nutr Rev. 1996 Jan;54(1 Pt 2):S59-65.
10. Donohue MC, Sperling RA, Salmon DP, Rentz DM, Raman R, Thomas RG, et al. The preclinical Alzheimer cognitive composite: measuring amyloid-related decline. JAMA Neurol. 2014 Aug;71(8):961–70.
11. Wechsler D. Wechsler adult intelligence scale—revised. New York: New York: Psychological Corp; 1981.
12. Grober E, Buschke H, Crystal H, Bang S, Dresner R. Screening for dementia by memory testing. Neurology. 1988 Jun;38(6):900–3.
13. 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.
14. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, et al. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994 Mar;49(2):M85-94.
15. Brink TL, Yesavage JA, Lum O, Heersema PH, Adey M, Rose TL. Screening Tests for Geriatric Depression. Clin Gerontol. 1982 Oct 14;1(1):37–43.
16. Herdman M, Gudex C, Lloyd A, Janssen MF, Kind P, Parkin D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res. 2011 Dec;20(10):1727–36.
17. Andrade LF, Ludwig K, Goni JMR, Oppe M, de Pouvourville G. A French Value Set for the EQ-5D-5L. PharmacoEconomics. 2020;38(4):413–25.
18. Barreto P de S, Ferrandez A-M, Saliba-Serre B. Questionnaire d’activité physique pour les personnes âgées (QAPPA) : validation d’un nouvel instrument de mesure en langue française. Sci Sports. 2011;26(1):11–8.
19. de Souto Barreto P. Construct and convergent validity and repeatability of the Questionnaire d’Activité Physique pour les Personnes Âgées (QAPPA), a physical activity questionnaire for the elderly. Public Health. 2013 Sep;127(9):844–53.
20. Patel MS, Benjamin EJ, Volpp KG, Fox CS, Small DS, Massaro JM, et al. Effect of a Game-Based Intervention Designed to Enhance Social Incentives to Increase Physical Activity Among Families: The BE FIT Randomized Clinical Trial. JAMA Intern Med. 2017 01;177(11):1586–93.
21. Patel MS, Small DS, Harrison JD, Fortunato MP, Oon AL, Rareshide CAL, et al. Effectiveness of Behaviorally Designed Gamification Interventions With Social Incentives for Increasing Physical Activity Among Overweight and Obese Adults Across the United States: The STEP UP Randomized Clinical Trial. JAMA Intern Med. 2019 Sep 9;1–9.
22. Patel MS, Asch DA, Rosin R, Small DS, Bellamy SL, Heuer J, et al. Framing Financial Incentives to Increase Physical Activity Among Overweight and Obese Adults: A Randomized, Controlled Trial. Ann Intern Med. 2016 Mar 15;164(6):385–94.
23. de Souto Barreto P, Rolland Y, Vellas B, Maltais M. Association of Long-term Exercise Training With Risk of Falls, Fractures, Hospitalizations, and Mortality in Older Adults: A Systematic Review and Meta-analysis. JAMA Intern Med. 2019 Mar 1;179(3):394–405.
24. Coley N, Ngandu T, Lehtisalo J, Soininen H, Vellas B, Richard E, et al. Adherence to multidomain interventions for dementia prevention: Data from the FINGER and MAPT trials. Alzheimers Dement J Alzheimers Assoc. 2019;15(6):729–41.
25. Richard E, Moll van Charante EP, Hoevenaar-Blom MP, Coley N, Barbera M, van der Groep A, et al. Healthy ageing through internet counselling in the elderly (HATICE): a multinational, randomised controlled trial. Lancet Digit Health. 2019 Dec 1;1(8):e424–34.
26. de Souto Barreto P, Cesari M, Andrieu S, Vellas B, Rolland Y. Physical Activity and Incident Chronic Diseases: A Longitudinal Observational Study in 16 European Countries. Am J Prev Med. 2017 Mar;52(3):373–8.
27. de Souto Barreto P, Delrieu J, Andrieu S, Vellas B, Rolland Y. Physical Activity and Cognitive Function in Middle-Aged and Older Adults: An Analysis of 104,909 People From 20 Countries. Mayo Clin Proc. 2016 Nov;91(11):1515–24.
28. Woo J, Chan R, Ong S, Bragt M, Bos R, Parikh P, et al. Randomized Controlled Trial of Exercise and Nutrition Supplementation on Physical and Cognitive Function in Older Chinese Adults Aged 50 Years and Older. J Am Med Dir Assoc. 2020 Mar;21(3):395–403.
29. Raafs BM, Karssemeijer EGA, Van der Horst L, Aaronson JA, Olde Rikkert MGM, Kessels RPC. Physical Exercise Training Improves Quality of Life in Healthy Older Adults: A Meta-Analysis. J Aging Phys Act. 2020 01;28(1):81–93.
30. Tulloch A, Bombell H, Dean C, Tiedemann A. Yoga-based exercise improves health-related quality of life and mental well-being in older people: a systematic review of randomised controlled trials. Age Ageing. 2018 01;47(4):537–44.
31. Fukuta H, Goto T, Wakami K, Ohte N. Effects of drug and exercise intervention on functional capacity and quality of life in heart failure with preserved ejection fraction: A meta-analysis of randomized controlled trials. Eur J Prev Cardiol. 2016 Jan;23(1):78–85.
32. Zampogna B, Papalia R, Papalia GF, Campi S, Vasta S, Vorini F, et al. The Role of Physical Activity as Conservative Treatment for Hip and Knee Osteoarthritis in Older People: A Systematic Review and Meta-Analysis. J Clin Med. 2020 Apr 18;9(4).
33. Oguchi H, Tsujita M, Yazawa M, Kawaguchi T, Hoshino J, Kohzuki M, et al. The efficacy of exercise training in kidney transplant recipients: a meta-analysis and systematic review. Clin Exp Nephrol. 2019 Feb;23(2):275–84.
34. Rasmussen NML, Belqaid K, Lugnet K, Nielsen AL, Rasmussen HH, Beck AM. Effectiveness of multidisciplinary nutritional support in older hospitalised patients: A systematic review and meta-analyses. Clin Nutr ESPEN. 2018;27:44–52.
35. Belleville S, Cuesta M, Bieler-Aeschlimann M, Giacomino K, Widmer A, Mittaz Hager AG, et al. Rationale and protocol of the StayFitLonger study: a multicentre trial to measure efficacy and adherence of a home-based computerised multidomain intervention in healthy older adults. BMC Geriatr. 2020 Aug 28;20(1):315.
36. Heffernan M, Andrews G, Fiatarone Singh MA, Valenzuela M, Anstey KJ, Maeder AJ, et al. Maintain Your Brain: Protocol of a 3-Year Randomized Controlled Trial of a Personalized Multi-Modal Digital Health Intervention to Prevent Cognitive Decline Among Community Dwelling 55 to 77 Year Olds. J Alzheimers Dis JAD. 2019;70(s1):S221–37.



P. de Souto Barreto1,2, S. Andrieu1,2, Y. Rolland1,2

1. Gerontopole of Toulouse, Institute of Ageing, Toulouse University Hospital (CHU Toulouse), Toulouse, France; 2. UMR INSERM 1027, University of Toulouse III, Toulouse, France.

Corresponding Author: Dr. Philipe de Souto Barreto. Gérontopôle de Toulouse, Institut du Vieillissement. 37, Allées Jules Guesde. 31000 Toulouse, France. Phone number: (+33) 561 145 668, Fax: (+33) 561 145 640, e-mail:

J Prev Alz Dis 2015;2(1):56-63
Published online Januay 13, 2015,


Physical activity (PA) contributes to brain health and plasticity, which suggests that PA would protect against the development of Alzheimer’s disease (AD). However, research on PA and AD biomarkers is very scarce. The objective of the present study was to perform a systematic review of studies that investigated the associations between PA and β-amyloid brain deposition in humans. Electronic searches were performed in PubMed, Cochrane Library, SportDiscus, PEDro, and PsychInfo databases. Articles were eligible if they have assessed both PA and β-amyloid brain deposition in humans. Five articles, published between 2010 and 2013, met eligibility criteria (study population varied across studies from 54 to 515, according with the β-amyloid measure. All five studies assessed both PA and PET-amyloid; among them, two studies also assessed CSF Aβ42 levels). All studies were based on cross-sectional data, from non-demented populations. Among the five included studies, three found significant associations between PA and β-amyloid brain deposition, and the other two did not find any significant association; limited evidence suggests that PA-amyloid plaques associations would be APOE ε4 allele-specific. In sum, no solid conclusions can be drawn on the associations between PA and human β-amyloid brain deposition currently. Future research on this topic should particularly pay attention to the operationalisation of clinically relevant and valid PA variables and should include important confounders in multivariate analysis. More information is needed on the potential interactions between PA and other AD risk factors (e.g., cognitive activities, APOE ε4 status, nutrition, smoking) and their combined effects on AD biomarkers.

Key words: Amyloid plaques, exercise, Alzheimer’s disease, physical fitness.



There is currently some evidence supporting that physical activity (PA) plays a protective role against the onset of dementia (1-6). Although the knowledge on the pathophysiology of Alzheimer’s disease (AD) has improved in the last decades, (7,8) the exact AD pathophyiology is still unknown. Despite this, physical inactivity and low PA levels are now considered as risk factors for AD (9).        

The research field that investigates the associations between PA, exercise, physical fitness, and brain health during the aging process is a promising field in expansion that may bring about some insight on how PA can protect subjects’ brain against AD (10). For example, at the molecular level, exercise (a subtype of PA that is systematically planned and purposeful, generally developed to improve subjects’ physiological and physical capacities), was found to increase the production and release of several molecules associated to angiogenesis and/or neurogenesis and cell survival, such as insulin-growth factor 1, vascular endothelial growth factor and brain-derived neurotrophic factor (10-12). Regarding brain structure, PA was found to be associated with increased grey matter volumes (13). Moreover, exercise was associated  with the size of (14) and possibly the blood flow for (15) the hippocampus, and with a better connectivity between the hippocampus and the anterior cingulated cortex and between aspects of the frontal, posterior, and temporal cortices (16). Therefore, PA appears to contribute to brain plasticity during aging (2, 17, 18).

Although these findings support that increasing PA levels contribute to the maintenance of brain health and optimal cognitive function, the exact mechanisms (if any) through which PA might protect the brain against the development of AD are still unknown. The lack of research on the associations between PA and specific AD biomarkers may explain this research gap. Lack of research seems to be of particular importance regarding β-amyloid brain deposition, which is a major pathological hallmark of AD (7, 8). The purpose of the present study was, therefore, to perform a systematic review of studies that investigated the associations between PA and β-amyloid brain deposition in humans; a particular interest was provided to highlight the heterogeneity among included studies to inform future directions for research in this field. 


This systematic review has been registered in the PROSPERO database (registration number CRD42014009805) and is publicly available (19). As indicated in the review protocol (19), and after the exploration of the included studies (particularly regarding heterogeneity), the authors of this systematic review consensually decided not to undertake a quantitative analysis (meta-analysis). The main reason for this was related to the small number of included studies, study design (all of them were based on cross-sectional data), and the high heterogeneity regarding the operationalisation of the PA variable; taken together, these limiting aspects would have probably led to spurious results if a meta-analysis had been undertaken. The reporting of this review follows the PRISMA guidelines (20). 

Main outcome

ß-amyloid brain deposition
We expected that most studies would assess this using  cerebrospinal fluid (CSF) Aβ42 and/or Positron Emission Tomography (PET), and that investigators of the original studies using PET would operationalise a kind of β-amyloid brain retention, such as a standardised uptake value ratio (SUVR). Therefore β-amyloid brain deposition variable can be continuous (CSF Aβ42 levels measured in pg/ml and SUVR or similar variable for PET-amyloid studies). Moreover, continuous β-amyloid variables may have been used to categorise participants as “presence” (+) or “absence” (-) of significant β-amyloid brain deposition according to specific cut-offs. For PET-amyloid studies, it is also possible that the investigators of the original articles have established the “presence” or “absence” of significant β-amyloid brain deposition through visual inspection of the imagery data. Therefore, our main outcome may potentially have been operationalised in four different variables; this will be taken into account in the present review by emphasising studies’ results according with the variable type (i.e., continuous, categorical, or categorical by visual inspection) in the following order of importance: 1) β-amyloid brain retention index (for PET-amyloid studies) and CSF Aβ42 levels, both of them operationalised as continuous variable, will constitute the preferable variables since they are probably more sensitive to detect associations (even weak associations) between amyloid plaques and PA; 2) presence” (+) or “absence” (-) of amyloid plaques according with a cut-off for the continuous β-amyloid variables; and, then, 3)  presence” (+) or “absence” (-) of amyloid plaques according to visual inspection.

Search strategy and eligibility criteria

Electronic searches were performed between May 9 and 12, 2014, in PubMed, the Cochrane Library, SportDiscus, PEDro, and PsychInfo. The full search strategy can be seen in appendix 1. No language, publication date, or study design restrictions were applied. Review articles retrieved by the electronic searches, as well as the studies included in the present review, were screened to find other potentially eligible studies that had escaped from our search strategy. Moreover, track tools available in the databases (e.g., Related Citations option in PubMed) were explored to further retrieve potentially eligible studies.

One rater screened the title/abstract of all retrieved articles. Potentially eligible articles were then accessed in full by one author, and eligibility was then confirmed by a second author. There was no divergences between the two raters on articles’ eligibility.

Articles were eligible if they met all the following criteria: 1) the study must have assessed both PA and β-amyloid brain deposition; 2) the study must have been performed in humans. No restrictions were applied with regard to age, gender, cognitive status, or characteristics of the PA type(s) evaluated (e.g., exercise, leisure-time PA, household PA, occupational PA, and/or PA for transportation).

Data extraction 

For each study, one author extracted the information from the original studies using a standard data collection form specifically designed for this study; a second author then confirmed the exactness of the data extracted. Any divergence between raters was resolved in a meeting between the two raters (100% consensus regarding data extraction was reached).   

Risk of bias 

The risk of bias was assessed using a modified version of the Downs and Black scale (21). The Downs and Black scale was reported to be useful to assess the risk of bias of studies included in systematic reviews (22) and has been used extensively (23-25). Although the Downs and Black scale is useful for systematic reviews, its use often requires adaptations (22) according with the characteristics (e.g., design of included studies, etc.) of the systematic review that is to be undertaken (23-25). It is important to highlight that other systematic reviews in the field of PA (23,24) have successfully used adapted versions of the Downs and Black scale. In the present work, we used a 14-item modified Downs and Black scale (scores can vary from 0 to 14, with higher scores indicating lower risk of bias), with the following items from the original scale being retained: 1-3, 5-7, 10-12, 16-18, 20, and 25. The statement of the item 17 from the original scale was adapted to take into account the risk related to the non-adjustment of statistical analysis for the time interval between β-amyloid assessment (PET-scan and/or CSF puncture) and PA assessment, which in some cases may constitute a potential bias for this review.   


Electronic searches retrieved 155 records; from this, only five articles (26-30) met the eligibility criteria (a total of 915 participants with assessment of both PA and PET-amyloid and 221 participants (n=2 studies (27, 29)) with assessment of both PA and CSF Aβ42 levels); articles were published very recently, between 2010 and 2013. Figure 1 displays the flowchart of article selection in the different phases of the screening process. One autopsy study (31) (n=135) examining the associations between diabetes-related factors and AD-related pathologic outcomes (including amyloid plaques) met all the inclusion criteria; however, PA in that study was used just as a confounder and the investigators did not report the effect size and the probability of the PA-amyloid plaques association, nor even mentioned in the original manuscript if this association was significant or not. Therefore, this study (31) was not included in this review.

Table 1 describes the main characteristics of the five studies included in this review. Most of them come from the United States (27-30) and all were based on cross-sectional data. Study population was composed of non-demented people in all studies.

Table 1. Overall characteristics of the studies selected for the review and their study population

Note. AD, Alzheimer’s disease; CDR, clinical dementia rating; CSF, cerebrospinal fluid; MCI, mild cognitive impairment; MET, metabolic equivalent; MMSE, mini-mental state examination: PA, physical activity; PET, positron emission tomography; SUVR, standard uptake value ratio; *Descriptive information of the subpopulation of 116 individuals with measurements of both PA and brain β-amyloid deposition was not reported.

Table 2 summarises the PA and amyloid-β brain assessments and variables, and reports the main findings for each study with regard to the associations between PA and β-amyloid brain deposition. All studies assessed β-amyloid brain deposition using PET imaging performed with the radiotracer 11C-Pittsburgh compound B (PiB); two studies (27, 29), from the same research team, also assessed CSF Aβ42 levels. PA assessment tools and variables highly varied among studies. All studies operationalised at least one continuous variable regarding β-amyloid brain deposition. Overall, three studies (26, 27, 29) found a significant association between PA and β-amyloid brain deposition, and the other two did not find any significant associations (28, 30).

Table 2. Characteristics of physical activity and β-amyloid brain assessments and variables, as well as the main findings of each selected article regarding the association between physical activity and brain amyloid-β deposition

Note. AD, Alzheimer’s disease; CI, confidence interval; CSF, cerebrospinal fluid; IPAQ, international physical activity questionnaire; MET, metabolic equivalent; PA, physical activity; PET, positron emission tomography; SUVR, standard uptake value ratio; *After adjustments to confounders (sex and the presence or history of diabetes, heart problems, or depression, and the delay between the PET scan and exercise assessment), all the associations reported for PiB-PET and CSF Aβ42 levels remained significant, but the exact coefficients were not reported.


Table 3 shows the assessment of the risk of bias for each included study. Using the 14-item modified Downs and Black scale, the risk of bias varied among studies from 10 to 13, which can be considered as a moderate-to-low risk overall.

Table 3. Risk of bias assessment evaluated by using a 14-item modified Downs and Black scale


Figure 1. Flowchart of article selection during the screening process



This review gathered for the first time all the available information to date about the associations between PA and β-amyloid brain deposition in humans. Very sparse information is available on this topic, with only five studies being included in this review. The main finding from this work indicates that the results of the original studies were mixed regarding the associations between PA and amyloid plaques, with three studies showing significant associations and the other two presenting non-significant relationships. Moreover, these associations appear to be partially APOE ε4 allele-specific, with greater PA-amyloid plaques associations in ε4 carriers (26, 27) than in ε4 non-carriers. A few methodological shortcomings that may have affected the PA- β-amyloid association were detected and will be discussed hereafter.

Firstly, all included studies used cross-sectional data, which exclude any possibility of teasing out the direction of the associations between PA and β-amyloid brain deposition. Although designing randomised controlled trials (RCT) to disentangle the effects of PA on amyloid plaques would be difficult to operationalise since such a RCT would probably need a very long-time length (maybe several years to detect relevant changes in β-amyloid brain deposition), with the associated increased risk of attrition and behavioural changes in terms of subjects’ daily PA (discontinuing PA is common among older people (32 ,33) a population frequently targeted in studies investigating β-amyloid brain deposition), long-term prospective observational studies would shed light on the potential causal relationship between PA and amyloid plaques. Indeed, as indicated by Jack et al. (8), long-term within-subject longitudinal data on AD biomarkers is currently lacking. Although that type of study design would probably be economically expensive (especially when PET-scan are used) and potentially burdensome to participants (especially when CSF Aβ42 measures are involved), it would provide information without precedent and would help to fill up some gaps in research on AD pathophysiology.

Second, all included studies examined PET-amyloid using PiB-PET. 18F-labeled ligands, such as the Florbetapir F 18 (previously called 18F-AV-45), are particularly interesting for research purposes since they have longer radioactive half-lives Imagery studies using different PET ligands are then needed to examine the associations between amyloid plaques and lifestyle behaviours.

Third, PA assessment tools used and the PA variables operationalised highly varied across studies. Although this high variation can be explained, at least partially, by the lack of “golden standard” in the assessment of PA (34), it affects the comparability among studies negatively. For example, the five studies included in this review assessed PA through: (1) interview focusing on the last 2-week PAs; (2) questionnaire focusing on last 7-day PA; (3) questionnaire focusing on last 10-year for a few and specific PAs; (4) PA frequency questionnaire focusing on last 12-month PA. It is important to highlight that any of the included studies did not assess PA using objective measures, such as doubly-labelled water or accelerometers, which constitute an important research gap. Moreover, a potential dose-response relationship between PA and β-amyloid brain deposition was rarely investigated. A preliminary attempt to approach this aspect was undertaken only by Brown et al., (26) who split their sample in PA tertiles; however, overall PA volumes were extremely high in that study and even the lowest PA tertile exceeded by far the current public health PA recommendations of 150min/week (35, 36) of at least moderate PA (i.e., at least 600 MET-min/week compared with the mean 1212 MET-min/week obtained by the lowest tertile in Brown et al. (26)); this suggests the presence of a selection bias towards the recruitment of highly active people. Future research on this topic should use both objective and subjective measures of PA and should pre-define meaningful cut-offs in PA volumes to test the potential dose-response relationship between PA and β-amyloid brain deposition. For example, testing the associations between increases of 30-minute/week of brisk walking and amyloid plaques would provide clinically relevant information for healthcare professionals.

Fourth, another gap related to the operationalisation of PA variables regards the fact that any study did not operationalise PA variables according to PA types and /or main domains (i.e., leisure-time, transportation, occupational/work, household), which would allow to investigate if the PA-amyloid plaques associations are type- or domain-dependent. Indeed, an increasing body of evidence suggests that PA-associated health promotion is probably domain-dependent (37-40), although, as far as we know, no study has examined the PA domain-dependent associations regarding cognitive outcome. Therefore, it is possible that the association between PA and β-amyloid brain deposition is PA domain-dependent and future research should take this into account by operationalising different variables according with PA domains and, if possible, by operationalising a variable accounting only for subjects’ exercise training (e.g., strength training, aerobic training, etc.).

Fifth, any study to date has not investigated the associations of the time spent in sedentary activities, i.e., activities during waking hours that do not increase energy expenditure above 1.5 metabolic equivalents (e.g., watching TV, reading or talking on the phone while sitting), with amyloid plaques. Indeed, sedentary time is associated to adverse health outcomes independently of subjects’ PA levels(41), including with regard to cognitive function (42). Similarly, no study has investigated the potential mediation effect played by physical fitness (performance-based physical tests) in the association between PA and β-amyloid brain deposition. Therefore, beside investigating the direct associations of sedentary time and physical fitness with AD biomarkers, to better appreciate the magnitude of the association between PA and amyloid plaques it would be useful to adjust this association to both sedentary time and physical fitness (when the information is available).

Finally, attention must be paid to the adjustments of the statistical analysis between PA, or any other modifiable AD risk factor, and β-amyloid brain deposition. β-amyloid brain deposition and clinical/lifestyle outcomes are commonly assessed in different dates, which means that there is a time interval between these assessments. For example, among the included studies of this review, the time interval between PA assessment and PET-scan was: almost one year (27), 1.5 years (30), or even 2.8 years (29) (no indication about time interval between PET and PA assessments was reported in the studies by Brown et al. (26) and Landau et al. (28)). The time interval between PA assessment and CSF Aβ42 assessments for the two studies with CSF measurements was: 1.627 and 1.729 years. Since we do not know the real implications of these time intervals for the deposition of amyloid plaques, investigators should adjust their analysis for this time variable (this adjustment was performed and clearly reported only by one study in this review (27)).

Although no solid conclusions can be drawn on the associations between PA and human β-amyloid brain deposition, particularly with regard to an eventual causal relationship, this review evidenced important methodological gaps in this research field that must be taken into account in future research. Future research on this topic should particularly pay attention to the following aspects: 1) the PA assessment tool must be valid and, preferentially, allow the construction of domain-specific PA variables; 2) the eventual dose-response relationship should be approached using clinically relevant cut-offs for PA; 3) sedentary time and physical fitness should be evaluated and controlled for in the statistical analysis; 4) other potentially important confounders (e.g., time interval between the assessments of AD biomarkers and clinical/PA variables) should not be neglected in multivariate analysis. Other important aspects for the prevention paradigm of AD (43) that deserve further attention regard the potential interactions between PA and other risk factors for AD (e.g., cognitive activities, APOE ε4 status, nutrition, smoking) (44) and their combined effects on AD biomarkers (with a particular attention to the potential synergistic effects of PA plus cognitive activities (45, 46)), as well as the associations between sarcopenia and AD biomarkers according with PA levels.

Funding/support: There was no external funding for this study.

Role of the sponsor: None

Conflicts of Interest: The authors declare no conflict of interest. There were no financial relationships with any organizations that might have an interest in the submitted work, no other relationships or activities that could appear to have influenced the submitted work.


1. Paterson DH, Warburton DE. Physical activity and functional limitations in older adults: a systematic review related to Canada’s Physical Activity Guidelines. Int J Behav Nutr Phys Act. 2010;7:38

2. Kramer AF, Erickson KI. Capitalizing on cortical plasticity: influence of physical activity on cognition and brain function. Trends Cogn. Sci. (Regul. Ed.). 2007;11(8):342‑348. 

3. Rockwood K, Middleton L. Physical activity and the maintenance of cognitive function. Alzheimers Dement. 2007;3(2 Suppl):S38‑44.

4. Larson EB, Wang L, Bowen JD, McCormick WC, Teri L, Crane P, et al. Exercise is associated with reduced risk for incident dementia among persons 65 years of age and older. Ann. Intern. Med. 2006;144(2):73‑81. 

5. Taaffe DR, Irie F, Masaki KH, Abbott RD, Petrovitch H, Ross GW, et al. Physical activity, physical function, and incident dementia in elderly men: the Honolulu-Asia Aging Study. J. Gerontol. A Biol. Sci. Med. Sci. 2008;63(5):529‑535.

6. Laurin D,Verreault R, Lindsay J, MacPherson K, Rockwood K. Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch Neurol 2001;58:498–504.

7. Holtzman DM, Morris JC, Goate AM. Alzheimer’s disease: the challenge of the second century. Sci Transl Med. 2011;3(77):77sr1. 

8. Jack CR Jr, Knopman DS, Jagust WJ, Petersen RC, Weiner MW, Aisen PS, et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12(2):207-216. 

9. Thies W, Bleiler L, Alzheimer’s Association. 2013 Alzheimer’s disease facts and figures. Alzheimers Dement J Alzheimers Assoc. 2013;9(2):208-245. 

10. Lautenschlager NT, Cox K, Cyarto EV. The influence of exercise on brain aging and dementia. Biochim. Biophys. Acta. 2012;1822(3):474‑481. 

11. Hillman CH, Erickson KI, Kramer AF. Be smart, exercise your heart: exercise effects on brain and cognition. Nat Rev Neurosci. 2008;9(1):58-65. 

12. 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):3017-3022. 

13 Erickson KI, Raji CA, Lopez OL, Becker JT, Rosano C, Newman AB, et al. Physical activity predicts gray matter volume in late adulthood: the Cardiovascular Health Study. Neurology. 2010;75(16):1415‑1422. 

14. 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):3017‑3022. 

15. Burdette JH, Laurienti PJ, Espeland MA, Morgan A, Telesford Q, Vechlekar CD, et al. Using network science to evaluate exercise-associated brain changes in older adults. Front Aging Neurosci. 2010;2:23. 

16. Voss MW, Prakash RS, Erickson KI, Basak C, Chaddock L, Kim JS, et al. Plasticity of brain networks in a randomized intervention trial of exercise training in older adults. Front Aging Neurosci. 2010;2. 

17. Kramer AF, Erickson KI, Colcombe SJ. Exercise, cognition, and the aging brain. J. Appl. Physiol. 2006;101(4):1237‑1242. 

18. Cotman CW, Berchtold NC. Exercise: a behavioral intervention to enhance brain health and plasticity. Trends Neurosci. 2002;25(6):295-301. 

19. De Souto Barreto P, Vellas B, Rolland Y, et al. The associations of physical activity with beta-amyloid brain deposition in humans: a systematic review. (last accessed May 19, 2014).

20. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ. 2009;339:b2535. 

21. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. 

22. Deeks JJ, Dinnes J, D’Amico R, Sowden AJ, Sakarovitch C, Song F, et al. Evaluating non-randomised intervention studies. Health Technol Assess Winch Engl. 2003;7(27):iii-x, 1-173. 

23. Prince SA, Adamo KB, Hamel ME, Hardt J, Connor Gorber S, Tremblay M. A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. Int J Behav Nutr Phys Act. 2008;5:56. 

24. Warburton DE, Charlesworth S, Ivey A, Nettlefold L, Bredin SS. A systematic review of the evidence for Canada’s Physical Activity Guidelines for Adults. Int J Behav Nutr Phys Act. 2010;7:39. 

25. MacLehose RR, Reeves BC, Harvey IM, Sheldon TA, Russell IT, Black AM. A systematic review of comparisons of effect sizes derived from randomised and non-randomised studies. Health Technol Assess Winch Engl. 2000;4(34):1-154. 

26. Brown BM, Peiffer JJ, Taddei K, Lui JK, Laws SM, Gupta VB, et al. Physical activity and amyloid-β plasma and brain levels: results from the Australian Imaging, Biomarkers and Lifestyle Study of Ageing. Mol Psychiatry. 2013;18(8):875-881. 

27. Head D, Bugg JM, Goate AM, Fagan AM, Mintun MA, Benzinger T, et al. Exercise Engagement as a Moderator of the Effects of APOE Genotype on Amyloid Deposition. Arch Neurol. 2012;69(5):636-643. 

28. Landau SM, Marks SM, Mormino EC, Rabinovici GD, Oh H, O’Neil JP, et al. Association of lifetime cognitive engagement and low β-amyloid deposition. Arch Neurol. 2012;69(5):623-629. 

29. Liang KY, Mintun MA, Fagan AM, Goate AM, Bugg JM, Holtzman DM, et al. Exercise and Alzheimer’s disease biomarkers in cognitively normal older adults. Ann Neurol. 2010;68(3):311-318. 

30. Vemuri P, Lesnick TG, Przybelski SA, Knopman DS, Roberts RO, Lowe VJ, et al. Effect of lifestyle activities on Alzheimer disease biomarkers and cognition. Ann Neurol. 2012;72(5):730-738. 

31. Matsuzaki T, Sasaki K, Tanizaki Y, Hata J, Fujimi K, Matsui Y, et al. Insulin resistance is associated with the pathology of Alzheimer disease: the Hisayama study. Neurology. 2010;75(9):764-770. 

32. Cohen-Mansfield J, Shmotkin D, Goldberg S. Predictors of longitudinal changes in older adults’ physical activity engagement. J Aging Phys Act. 2010;18(2):141-157. 

33. Hirvensalo M, Lintunen T, Rantanen T. The continuity of physical activity–a retrospective and prospective study among older people. Scand J Med Sci Sports. 2000;10(1):37-41. 

34. Lamonte MJ, Ainsworth BE. Quantifying energy expenditure and physical activity in the context of dose response. Med Sci Sports Exerc. 2001;33(6 Suppl):S370-378; discussion S419-420. 

35. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee I-M, et al. American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc. 2011;43(7):1334-1359. 

36. American College of Sports Medicine, Chodzko-Zajko WJ, Proctor DN, Fiatarone Singh MA, Minson CT, Nigg CR, et al. American College of Sports Medicine position stand. Exercise and physical activity for older adults. Med Sci Sports Exerc. 2009;41(7):1510-1530.

37. Johnsen NF, Ekblond A, Thomsen BL, Overvad K, Tjønneland A. Leisure time physical activity and mortality. Epidemiol Camb Mass 2013;24(5):717-725.

38. Clays E, Lidegaard M, De Bacquer D, Van Herck K, De Backer G, Kittel F, et al. The combined relationship of occupational and leisure-time physical activity with all-cause mortality among men, accounting for physical fitness. Am J Epidemiol 2014;179(5):559-566. 

39. Holtermann A, Marott JL, Gyntelberg F, Søgaard K, Suadicani P, Mortensen OS, et al. Occupational and leisure time physical activity: risk of all-cause mortality and myocardial infarction in the Copenhagen City Heart Study. A prospective cohort study. BMJ Open 2012;2(1):e000556.

40. Bellavia A, Bottai M, Wolk A, Orsini N. Physical activity and mortality in a prospective cohort of middle-aged and elderly men – a time perspective. Int J Behav Nutr Phys Act 2013;10:94 

41. Van der Ploeg HP, Chey T, Korda RJ, Banks E, Bauman A. Sitting time and all-cause mortality risk in 222 497 Australian adults. Arch Intern Med. 2012;172(6):494-500.

42. Wang JYJ, Zhou DHD, Li J, Zhang M, Deng J, Tang M, et al. Leisure activity and risk of cognitive impairment: the Chongqing aging study. Neurology. 2006;66(6):911-913

43. Whitehouse P, George DR. Prevention: how the paradigm is shifting. J Prev Alz Dis. 2014 ; 1(1):51-55.

44. De Souto Barreto P. Alzheimer’s disease: learning from the past, looking to the future. Am J Alzheimers Dis Other Demen. 2013;28(4):304-305

45. Vellas B, Carrie I, Gillette-Guyonnet S, et al. MAPT study: a multidomain approach for preventing Alzheimer’s disease: design and baseline data. J Prev Alz Dis. 2014 ; 1(1):13-22.

46. Richard E, Khachaturiani AS. Multi-domain prevention for multi-factorial disease. J Prev Alz Dis. 2014 ; 1(1):4-5.