Abstract
Background
The apolipoprotein E (APOE) e4 allele is a well-established genetic risk factor of brain aging. Vigorous physical activity may be particularly important in APOE-e4 carriers, but data have been inconsistent, likely due to differences in the timing of the physical activity assessment, definition of cognitive decline, and/or sample size.
Methods
We prospectively evaluated the association between vigorous physical activity and cognition assessed at least 9 years later, according to APOE-e4 carrier status. Biennially from 1986, Nurses’ Health Study participants reported their leisure-time physical activities. Starting in 1995–2001 and through 2008, participants (aged 70+ years) underwent up to 4 repeated cognitive telephone assessments (6 tasks averaged together using z-scores).
Results
Among 7252 women, latent process mixed models identified 3 major patterns of cognitive change over 6 years: high-stable, medium-stable, and decline. Taking the high-stable cognitive trajectory as the outcome reference in multinomial logistic regressions, highest tertile of vigorous physical activity (≥5.9 metabolic-equivalent [MET]-hours/wk) compared to lowest tertile (≤0.9 MET-hours/wk) was significantly associated with subsequent lower likelihood of the medium-stable trajectory in the global score (odds ratio [OR] [95% CI] = 0.72 [0.63, 0.82]), verbal memory (OR [95% CI] = 0.78 [0.68–0.89]), and telephone interview of cognitive status score (OR [95% CI] = 0.81 [0.70–0.94]). Vigorous physical activity was also associated with lower likelihood of decline in category fluency (OR [95% CI] = 0.72 [0.56, 0.92]). We observed some evidence (p-interaction = .07 for the global score) that the association was stronger among APOE-e4 carriers than noncarriers (OR [95% CI] = 0.60 [0.39, 0.92] vs 0.82 [0.59, 1.16]).
Conclusion
Midlife vigorous physical activity was associated with better cognitive trajectories in women in their seventies, with suggestions of stronger associations among APOE-e4 carriers.
Keywords: Cognition, Cohort studies, Epidemiology, Genotype, Latent process mixed models, Physical activity
Decrements in cognitive function predict dementia many years later and may be considered a marker of this highly disabling disease (1), whose prevalence is expected to increase dramatically in the near future with population aging (2). Given the absence of any curative treatment, epidemiologic studies play a critical role for the investigation of various risk factors and gene–environment interactions to delay the onset or reduce the risk of cognitive decline (3).
Among nonmodifiable risk factors, the apolipoprotein E (APOE)-e4 allele is a well-established genetic risk factor for cognitive decline that may account for 13% to 20% of dementia cases (4). APOE is a protein found in the plasma and the central nervous system, helping to regulate cholesterol and lipid metabolism and involved in cellular reparative processes (5). The e4 isoform of the APOE, however, reduces the clearance of soluble amyloid beta (Aβ) from the brain and increases its deposition in neuritic plaques (6). As the APOE-e4 allele is both common—20% to 25% in North-America and European countries (7)—and nonmodifiable, it is important to understand the variability in cognitive decline and possible modifiable risk factors for APOE-e4 carriers (4).
A recent study suggested an interaction between APOE-e4 status and physical activity on coronary artery disease, where an association between physical activity and a lower risk was observed among APOE-e4 carriers only, especially when vigorous physical activity such as running or bicycling was evaluated (5). Similarly, some studies have observed an inverse association on cognitive decline of high physical activity in APOE-e4 carriers only (8). A possible mechanism is that high physical activity in APOE-e4 carriers could reduce the neuronal level of APOE-e4 and thereby lower the total amount of APOE-e4 fragments in the brain and subsequently, the risk of developing cognitive decline (8,9). Other biological pathways involving APOE-e4, physical activity, and cognition such as vascular risk modulation, neurotrophic factors, and/or reduced inflammation, may also contribute to a better response of APOE-e4 carriers to the beneficial effects of physical exercise on cognitive function compared with noncarriers, as suggested by both observational and experimental studies (10–13). In fact, mechanisms linking physical activity and the brain are likely long term and multidimensional (14), making it more challenging to understand the potential interaction associated with the APOE-e4 allele. In addition to disentangling biologic mechanisms involved, it will be essential to determine the optimal time-windows for lifestyle intervention to prevent cognitive aging (15), especially among high-risk groups.
On the whole, studies investigating the interaction between physical activity and APOE-e4 in relation to cognitive aging have been inconsistent, likely due to differences in the timing of the assessment of physical activity, definition of cognitive decline, and/or sample size (16–20). In particular, studies may not have been able to address reverse causation bias, because of the short duration of follow-up. It is critically important to evaluate physical activity, particularly vigorous activity, in midlife in relation to cognitive change (14). Furthermore, most interaction studies assessed cognitive function only at one timepoint (18,19) rather than considering cognitive change over time, which can be heterogeneous and reflective of different underlying processes of decline (16,20). In group-based trajectory modeling (21), major patterns of cognitive change can be identified and evaluated; this approach has been particularly useful in the study of decline associated with APOE-e4 (22).
Thus, our aim was to evaluate the long-term association between midlife vigorous physical activity and later cognitive trajectories with sufficient delay between the 2 measures and the interaction with APOE-e4 status among older women in the Nurses’ Health Study (NHS).
Materials and Methods
The Nurses’ Health Study Cohort and the Cognitive Subcohort
The NHS began in 1976, when 121 700 U.S. registered nurses, aged 30–55 years, and residing in 11 U.S. States, completed a mailed questionnaire about their health and lifestyle. Follow-up questionnaires were mailed biennially. From 1995 to 2001, participants who were ≥ 70 years old and free of stroke (N = 22 103) were invited to participate in a telephone-based prospective study of cognitive function (Supplementary Figure 1). Of all eligible women, 92% participated in the first telephone cognitive interview. Three follow-up cognitive assessments were administered at approximately 2-year intervals (median cognitive follow-up time: 6 years; Supplementary Figure 1), and participation rate remained high (>90%) over time. The study was approved by the Institutional Review Board of Brigham and Women’s Hospital (Boston, MA).
Cognitive Assessment
Assessment of cognitive function via validated telephone interviews has been previously described in detail (23). Each initial and follow-up cognitive interview was administered by trained interviewers using (i) the Telephone Interview of Cognitive Status (TICS) (24), an adaptation of the Mini-Mental State Exam, as well as 5 other cognitive tests: (ii) immediate and (iii) delayed recalls of the East Boston Memory test (EBMT), (iv) delayed recall of the TICS 10-word list, (v) category fluency (animal naming test), and (vi) digit span-backward. To evaluate overall cognitive functioning, a global composite score was calculated as the average of all z-scores (standardized scores were used rather than raw score because their range varied across tests). A composite verbal memory score was also derived by averaging test-specific z-scores of the immediate and delayed recalls of the EBMT and of one component of the TICS—the TICS 10-word list.
Our primary outcome was the global composite score, but to evaluate multiple other cognitive domains, we considered also as secondary outcomes the verbal memory score, the TICS score, and the category fluency score. Previous data confirmed the validity and reliability of the telephone cognitive interview (25). The performance on the telephone cognitive battery was strongly correlated (r = .81) with performance on a comprehensive battery of neuropsychological tests from a detailed, in-person interview conducted in 61 highly educated women aged 70 years or older. Inter-interviewer reliability was also high across 10 interviewers (r > .95 for each cognitive test).
Physical Activity Assessment
From 1986, participants were asked detailed information on their leisure-time physical activity repeatedly every 2–4 years in the biennial questionnaires. Women reported the average amount of time spent per week during the past year on the following activities: running (≤10 minutes/mile); jogging (>10 minutes/mile); walking or hiking outdoors; racquet sports; lap swimming; bicycling; and aerobic dance or use of exercise machines. Participants also indicated their usual outdoor walking pace: easy (>30 minutes/mile), normal (21–30 minutes/mile), brisk (16–20 minutes/mile), or very brisk (≤15 minutes/mile), and the number of flights of stairs they climbed daily. We assigned each activity a metabolic equivalent value (MET) according to accepted standards (26), where 1 MET is the energy expended while sitting quietly. MET values were 12 for running; 8 for stair climbing; 7 for jogging, racquet sports, lap swimming and bicycling; and 6 for aerobic dance and use of exercise machines. MET values for walking varied by reported pace, from 2.5 METs for easy pace to 4.5 METs for very brisk pace. For each activity, we estimated the energy expended in MET-hours/wk, by multiplying its MET value by the time spent performing it.
Our exposure of interest was vigorous physical activity (≥6 MET) following standard definitions of distinguishing between vigorous and lower-intensity physical activity (27). It was estimated here as the sum of the energy expended in all leisure-time physical activities inquired, except walking/hiking as this latter activity was assigned lower METs than 6 even for very brisk pace.
In a validation study among women in the NHS II (a similar cohort of nurses), physical activity estimated using our instrument correlated strongly with repeated past-week recalls of physical activity (r = .79) and with physical activity logged in diaries during the year (r = .62) (28). Moreover, participants’ responses to the physical activity questionnaire assessed 1-year apart were reasonably correlated (r = .59), given the expected true changes that might occur over a 1-year period.
To try to rule out reverse causation while maximizing robustness of the vigorous physical activity assessment, we estimated its level by averaging, as much as possible, estimations from questionnaires completed at least 9 years prior to the baseline cognitive assessment; such a time lag was recommended by 2 previous studies supporting a dropout of physical activity among cognitive decliners according to this time pattern (29,30). In our setting, this constraint of time lag led to an evaluation of vigorous physical activity based on 2 physical activity questionnaires for 85% participants, and 1 questionnaire for the 15% remaining. Mean age at physical activity assessment was 62 years. As “middle age” is defined as age 45–65 years, we hereafter designated our exposure of interest as “midlife” vigorous physical activity.
APOE-e4 Genetic Data
To obtain genetic data in the NHS, blood samples were collected from participants in 1989–1990. Among those who had not provided blood samples, participants were invited to provide buccal cell samples in 2002–2004. Genomic DNA was extracted using the ReturPureGene DNA Isolation Kit (Gentra Systems Minneapolis, MN), which was the genotyped using a Taqman Assay (Applied Biosystems, Foster City, CA) (31) or was imputed from various GWAS chips (32). Genotype data were available for half of the cognitive subcohort. Health status of included women was better than those without genotype data, but they did not differ by physical activity level. For the APOE genotypes, 63.5% had e3/e3; 20.2%, e3/e4; 0.6%, e2/e2; 12.4%, e2/e3; 1.8%, e2/e4; and 1.6%, e4/e4.
Covariates
We obtained information on multiple potential confounders factors determined a priori as cognitive risk factors possibly linked with physical activity (33,34), including demographic, socioeconomic, occupational, lifestyle, and health-related factors. In particular, we considered age at first cognitive assessment (continuous, in years), education (Associate’s degree or Nursing degree, Bachelor’s degree, Master’s or Doctorate degree), walking or hiking (continuous, in MET-hours/wk), smoking status (never smoker, past smoker, current smoker), alcohol consumption (0, 1–14, >15 g/day), body mass index (≤21, [22–24], [25–29], ≥30 kg/m2), the Alternate Healthy Eating Index score (<50, ≥50), which is a diet quality indicator (35), age at menopause (≤49, [50–52], ≥53 years), postmenopausal hormone use (never user, past user, current user), current vitamin E supplement use (nonuser, user, missing), current multivitamin supplement use (nonuser, user), current aspirin use (nonuser, 1–2/wk, 3–7/wk), current ibuprofen use (nonuser, user), SF-36 Mental Health Index (≤52, >52), diabetes diagnosis (yes, no), high blood pressure (yes, no), hypercholesterolemia (yes, no), myocardial infarction history (yes, no), and SF-36 physical function score (in tertiles). All confounders were updated as of initial cognitive interview. If missing values for covariates were <10%, they were imputed to the modal category or the median value; otherwise a missing category was created.
Study Population
Of the 8536 participants who completed the initial cognitive interview and had genotype data, we excluded women who did not complete any follow-up cognitive interviews (n = 459; 5.4%). We then excluded those without any data on their physical activity at least 9 years before the first cognitive assessment (n = 607; 7.1%). We further excluded those whose SF-36 physical function score < 20 (the lowest decile observed among UK women aged 60–74 years (36); n = 183; 2.1%), those unable to walk (n = 21; 0.2%), and those with Parkinson disease (n = 14; 0.2%), all of whom may have difficulty engaging in physical activity, leaving 7252 women in the analysis sample. Included versus excluded women did not differ significantly in age. Regarding vigorous physical activity, mean level was 7.6 MET-hours/wk among the former group versus 6.5 MET-hours/wk in the latter (p = .06).
Statistical Analysis
Typology of cognitive trajectories
To identify major patterns of cognitive trajectories up to 4 repeated assessments for global cognition, verbal memory, TICS, and category fluency, we used group-based trajectory modeling (21), also referred to as latent process mixed model or semiparametric mixture model (37). We used PROC TRAJ in SAS (21), for each of the cognitive score sets, with the duration since the first assessment as the time scale. The group-based trajectory modeling presumes that the data set comprises latent distinct groups of trajectories that best summarize the complex developmental information collected over a life-course (38).
We followed the process described by Andruff et al. (39), by selecting the number of cognitive trajectory groups meeting 2 criteria: (i) model that includes a reasonable proportion of persons (ideally no less than 5% (39)) and (ii) model with the smallest absolute Bayesian Information Criterion value that also minimized overlap in the CIs of adjacent trajectories while summarizing the distinctive features of the data as parsimoniously as possible. The shape (eg, linear vs quadratic) of each trajectory group was also identified.
After selecting the best fitting model, individuals were assigned to one of the trajectory groups based on their highest estimated group-membership probabilities (“p_i”). To further assess model adequacy, we checked that the trajectory group-membership probabilities was on average greater than the suggested threshold of .7 (38). Then, to refine our cognitive outcome measures, we excluded from the analysis women whose cognitive trajectories did not fit into any identified normative patterns (“max of p_i < .7”): 1116 (15.4%) women for the global cognition trajectory, 1382 (19.1%) for the verbal memory trajectory, 971 (13.4%) for the TICS trajectory, and 1196 (16.5%) for the category fluency trajectory.
Descriptive and association analyses
We used multivariable-adjusted multinomial logistic regressions to evaluate the association with our primary outcome—the cognitive trajectory group variable for the global score (dependent variable) —and (i) APOE-e4 status (carrier or noncarrier), and (ii) vigorous physical activity (in tertiles), in the whole sample or stratified by APOE-e4 status. We considered alternately as secondary outcomes, the trajectory group variables regarding verbal memory, TICS, and category fluency, respectively. We used multiple logistic regression and Wald tests to formally test the interaction between APOE-e4 status and higher level of vigorous physical activity for each cognitive outcome. Two sets of adjustment variables were considered each time: the first model was only adjusted for age and educational level, whereas the second model further controlled for all covariates previously listed as potential confounders.
All statistical analyses were conducted using SAS V.9.3 (SAS Institute, Cary, NC). The statistical significance level was set to α = .05 for analyses except for the interaction analyses (which tend to be less statistically powerful), where we set the significance level at α = .10.
Sensitivity/secondary analyses
We conducted several sensitivity analyses. First, to evaluate whether the results may be sensitive to the time scale choice, we used age as the time scale rather than calendar time. Second, we conducted analyses without the exclusion criteria of “max of p_i < .7” to verify the robustness of our trajectory modeling. Third, we restricted the analyses to women without any APOE-e2 allele (excluding 14.7% of participants), as some evidence suggest that this allele may counteract partly the deleterious cognitive effect associated with APOE-e4 allele (40). Lastly, and although our primary interest was in vigorous physical activity given results from prior studies, we also evaluated the interaction with total physical activity.
Results
Cognitive Trajectories
For each of the 4 cognitive outcomes considered, group-based trajectory modeling indicated 3 major and similarly shaped evolution patterns over a median of 6 years of follow-up: “high-stable” (representing 40.2% of women for the global score and 39.0%, 69.0%, and 8.5%, respectively for the verbal memory, TICS, and category fluency scores); “medium-stable” (52.3%, 54.6%, 27.8%, and 52.5%, respectively); and “decline” (7.5%, 6.4%, 3.2%, and 39.0%, respectively; Figure 1 and Supplementary Figure 2). Kappa coefficients between cognitive trajectory groups were .86 for global versus verbal memory scores, .47 for global versus TICS scores and .08 for global versus category fluency scores.
Figure 1.
Main patterns of global cognitive score trajectory indicated by group-based trajectory modeling. Global score is the average of the z-scores of Telephone Interview of Cognitive Status (TICS), immediate and delayed recalls of East Boston Memory Test, delayed recall of TICS 10-word list, test of category fluency and digit backwards test.
Characteristics by Midlife Vigorous Physical Activity
Compared to women who had, on average, a low level of midlife vigorous physical activity, those with higher vigorous physical activity were more likely to show high educational attainment and high AHEI index and to consume alcohol but less likely to smoke or to be overweight or obese (all p < .05; Table 1). Women with higher vigorous physical activity were also more likely to use vitamin E, vitamin C, or multivitamin but less likely to have high blood pressure. Vigorous physical activity was positively linked with physical functioning as well as walking (all p < .05) but not with APOE-e4 status.
Table 1.
Characteristics* of the Study Population by Level of Midlife† Vigorous Physical Activity (n = 7252)
| Midlife Vigorous Physical Activity | |||
|---|---|---|---|
| Tertile 1 (n = 2563) | Tertile 2 (n = 2268) | Tertile 3 (n = 2421) | |
| Mean (SD) age at first cognitive interview, year‡ | 74.3 (2.3) | 74.1 (2.3) | 74.2 (2.2) |
| Mean (SD) vigorous physical activity, MET-hours/wk§ | 0.4 (0.3) | 2.9 (1.5) | 19.5 (19.9) |
| Mean (SD) walking, MET-hours/wk§ | 7.3 (9.0) | 7.3 (8.1) | 10.4 (9.9) |
| Mean (SD) global score‖ at first cognitive interview, standard unit | 0.04 (0.5) | 0.09 (0.5) | 0.11 (0.6) |
| High physical function¶, % | 29 | 36 | 43 |
| APOE-e4 allele carrier, % | 24 | 22 | 24 |
| Highest attained education | |||
| RN degree, % | 80 | 77 | 73 |
| Bachelor’s degree, % | 15 | 16 | 19 |
| Master/doctorate degree, % | 5 | 7 | 8 |
| Current smoker, % | 8 | 8 | 5 |
| Alcohol drinker, % | 43 | 50 | 56 |
| Body mass index >30 kg/m2, % | 20 | 17 | 14 |
| Alternate Healthy Eating Index without alcohol <50, % | 65 | 63 | 52 |
| Age at menopause <50 y, % | 35 | 35 | 32 |
| Current postmenopausal hormone use, % | 40 | 40 | 42 |
| Aspirin use 3–7 times/wk, % | 39 | 41 | 39 |
| Ibuprofen use, % | 17 | 18 | 18 |
| Vitamin E use, % | 42 | 47 | 49 |
| Multivitamin use, % | 63 | 67 | 70 |
| SF-36 Mental Health Index ≤52, % | 4 | 4 | 3 |
| Diabetes, % | 10 | 9 | 8 |
| High blood pressure, % | 56 | 53 | 52 |
| High cholesterol, % | 67 | 67 | 65 |
| Myocardial infarction, % | 6 | 6 | 6 |
Notes: APOE = apolipoprotein E.
*Values are means (SD) for continuous variables; percentages for categorical variables and are standardized to the age distribution of the study population.
†Participants’ mean age at physical activity estimation was 62 years.
‡Value is not age adjusted.
§MET = metabolic equivalent of task.
‖Global score is the average of the z-scores of TICS, immediate and delayed recalls of the East Boston Memory Test, delayed recall of TICS 10-word list, test of category fluency and digit backwards test.
¶Defined as the highest tertile of the physical function score (subscale of the SF-36 questionnaire).
Midlife Vigorous Physical Activity and Later Cognitive Trajectories, by APOE-e4 Status
Compared to women with no APOE-e4 allele, APOE-e4 carriers had a higher probability of following a “medium-stable” or a “decline” cognitive trajectory rather than a high-stable pattern for global score (p < .001; Figure 1); results were very similar for verbal memory, TICS, and category fluency (data not shown in figures).
Midlife vigorous physical activity was consistently associated with better cognitive outcomes in later life (Table 2). In multivariable-adjusted models where the high-stable cognitive trajectory was the outcome reference, higher vigorous physical activity was associated with lower likelihood of mediums-stable cognitive trajectories in global cognitive score (ORtertile3 vs tertile 1 [95% CI] = 0.72 [0.63–0.82]), verbal memory (0.78 [0.68–0.89]), and TICS (0.81 [0.70–0.94]) and of decline trajectory in category fluency score (0.72 [0.56–0.92]). The associations with other cognitive tests did not attain signification level but they were all 3 in the same direction (lower likelihood of decline with higher vigorous physical activity).
Table 2.
Association Between Midlife Vigorous PA and Later Worse Cognitive Trajectories*
| Global† (N = 6136) | Verbal Memory† (N = 5870) | TICS‡ (N = 6281) | Category Fluency (N = 6056) | |||
|---|---|---|---|---|---|---|
| Trajectories* | Vigorous PA tertile | OR [95% CI] | OR [95% CI] | OR [95% CI] | OR [95% CI] | |
| Model 1§ | Medium-stable (52.2%) | 1 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 2 | 0.78 [0.69–0.90] | 0.86 [0.75–0.98] | 0.78 [0.68–0.89] | 0.80 [0.64–1.02] | ||
| 3 | 0.68 [0.60–0.77] | 0.74 [0.64–0.84] | 0.76 [0.66–0.87] | 0.82 [0.65–1.03] | ||
| p-trend | <.0001 | <.0001 | .002 | .2 | ||
| Decline (7.4%) | 1 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | |
| 2 | 0.88 [0.69–1.13] | 0.96 [0.74–1.25] | 0.97 [0.69–1.36] | 0.73 [0.57–0.93] | ||
| 3 | 0.73 [0.57–0.93] | 0.75 [0.57–0.98] | 0.81 [0.57–1.14] | 0.69 [0.55–0.88] | ||
| p-trend | .01 | .02 | .2 | .01 | ||
| Model 2‖ | Medium-stable (52.2%) | 1 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 2 | 0.81 [0.71–0.92] | 0.88 [0.77–1.01] | 0.81 [0.70–0.93] | 0.80 [0.63–1.02] | ||
| 3 | 0.72 [0.63–0.82] | 0.78 [0.68–0.89] | 0.81 [0.70–0.94] | 0.81 [0.64–1.03] | ||
| p-trend | <.0001 | .0006 | .03 | .2 | ||
| Decline (7.4%) | 1 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | |
| 2 | 0.94 [0.73–1.21] | 1.02 [0.78–1.33] | 1.02 [0.72–1.44] | 0.74 [0.58–0.95] | ||
| 3 | 0.80 [0.61–1.02] | 0.81 [0.61–1.07] | 0.84 [0.59–1.21] | 0.72 [0.56–0.92] | ||
| p-trend | .08 | .09 | .3 | .04 |
Notes: OR = odds ratio; PA = physical activity; TICS = Telephone Interview of Cognitive Status.
*The high-stable cognitive trajectory is considered as the reference.
†Global score is the average of the z-scores of TICS, immediate and delayed recalls of the East Boston Memory Test, delayed recall of TICS 10-word list, test of category fluency and digit backwards test; verbal memory score is the average of the z-scores of the immediate and delayed recalls of the East Boston Memory Test and the immediate and delayed recalls of the TICS 10-word list.
‡Telephone Interview of Cognitive Status.
§Adjusted for age at first cognitive assessment (continuous, in years) and education (Associate’s degree or Nursing degree, Bachelor’s degree, Master’s or Doctorate degree).
‖Adjusted for age at first cognitive assessment (continuous, in years), education (Associate’s degree or Nursing degree, Bachelor’s degree, Master’s or Doctorate degree), walking (continuous, in MET-hours/wk), smoking status (never smoker, past smoker, current smoker), alcohol consumption (0, 1–14, >15 g/d), body mass index (≤21, [22–24], [25–29], ≥30 kg/m2), the Alternate Healthy Eating Index score (<50, ≥50), age at menopause (≤49, [50–52], ≥53 y), postmenopausal hormone use (never, past, current use), vitamin E supplement use (nonuse, current use, missing), multivitamin supplement use (nonuse, current use), current aspirin use (nonuse, 1–2 times/wk, 3–7 times/wk), ibuprofen use (nonuse, current use), SF-36 Mental Health Index (≤52, >52), diabetes (yes, no), high blood pressure (yes, no), hypercholesterolemia (yes, no), myocardial infarction history (yes, no), and SF-36 physical function score (in tertiles).
Midlife Vigorous Physical Activity and Later Cognitive Trajectory Stratified by APOE-e4 Status and Corresponding Interactions
We observed some evidence of an interaction by APOE-e4 status in the association between midlife vigorous physical activity and cognitive decline, specifically, when modeling the global score trajectories (p-interaction = .07 in the logistic model contrasting “decline” versus “high-stable” outcome) and the TICS score trajectories (p-interaction = .09 for the same contrast). When analyses were stratified by APOE-e4 status, the association between vigorous physical activity and better cognitive outcomes tended to be stronger among APOE-e4 carriers than among noncarriers, although the ORs were not systematically significant (Table 3). In particular, the OR [95% CI] of “decline” versus “high-stable” global score cognitive trajectory for tertile 3 compared to tertile 1 of vigorous physical activity were 0.60 [0.39–0.92] among APOE-e4 carriers versus 0.82 [0.59–1.16] among noncarriers and for category fluency score, 0.63 [0.36–1.10] among APOE-e4 carriers versus 0.74 [0.56–0.98] among noncarriers.
Table 3.
Association Between Midlife Vigorous PA and Later Worse Cognitive Trajectory* Stratified by APOE-e4 Status†
| Global‡ (APOE-e4- n = 4691, APOE-e4+ n = 1445) |
Verbal memory‡ (APOE-e4-n = 4481, APOE-e4+ n = 1389) | TICS§ (APOE-e4- N = 4802, APOE-e4+ n = 1479) |
Category fluency (APOE-e4- n = 4621, APOE-e4+ n = 1435) | |||
|---|---|---|---|---|---|---|
| Trajectories* | Vigorous PA tertile | OR [95% CI] | OR [95% CI] | OR [95% CI] | OR [95% CI] | |
| p-interactions‖ | Medium-stable | p = .7 | p = .9 | p = .4 | p = .8 | |
| Decline | p = .07 | p = .8 | p = .09 | p = .5 | ||
| APOE-e4- | Medium-stable | 1 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 2 | 0.80 [0.69–0.93] | 0.88 [0.75–1.02] | 0.78 [0.66–0.92] | 0.82 [0.63–1.07] | ||
| 3 | 0.75 [0.64–0.87] | 0.79 [0.67–0.92] | 0.83 [0.70–0.98] | 0.81 [0.62–1.06] | ||
| p-trend | .002 | .005 | .1 | .3 | ||
| Decline | 1 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | |
| 2 | 1.11 [0.81–1.52] | 1.38 [0.96–1.96] | 1.34 [0.84–2.12] | 0.76 [0.58–1.00] | ||
| 3 | 0.82 [0.59–1.16] | 0.87 [0.58–1.29] | 0.98 [0.59–1.61] | 0.74 [0.56–0.98] | ||
| p-trend | .1 | .1 | .6 | .1 | ||
| APOE-e4+ | Medium-stable | 1 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) |
| 2 | 0.78 [0.58–1.05] | 0.89 [0.66–1.21] | 0.91 [0.68–1.21] | 0.76 [0.44–1.32] | ||
| 3 | 0.58 [0.43–0.78] | 0.72 [0.53–0.97] | 0.74 [0.56–0.99] | 0.80 [0.46–1.39] | ||
| p-trend | .0006 | .03 | .04 | .6 | ||
| Decline | 1 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | |
| 2 | 0.71 [0.46–1.11] | 0.69 [0.43–1.09] | 0.75 [0.42–1.33] | 0.71 [0.41–1.24] | ||
| 3 | 0.60 [0.39–0.92] | 0.65 [0.42–1.01] | 0.66 [0.38–1.15] | 0.63 [0.36–1.10] | ||
| p-trend | .04 | .1 | .2 | .2 |
Notes: APOE = apolipoprotein E; OR = odds ratio; PA = physical activity; TICS = Telephone Interview of Cognitive Status.
*The high-stable cognitive trajectory is considered as the reference.
†Adjusted for age at first cognitive assessment (continuous, in years), education (Associate’s degree or Nursing degree, Bachelor’s degree, Master’s or Doctorate degree), walking (continuous, in MET-hours/wk), smoking status (never smoker, past smoker, current smoker), alcohol consumption (0, 1–14, >15 g /d), body mass index (≤21, [22–24], [25–29], ≥30 kg/m2), the Alternate Healthy Eating Index score (<50, ≥50), age at menopause (≤49, [50–52], ≥53 years), postmenopausal hormone use (never, past, current use), vitamin E supplement use (nonuse, current use, missing), multivitamin supplement use (nonuse, current use), current aspirin use (nonuse, 1–2 times/wk, 3–7 times/wk), ibuprofen use (nonuse, current use), SF-36 Mental Health Index (≤52, >52), diabetes (yes, no), high blood pressure (yes, no), hypercholesterolemia (yes, no), myocardial infarction history (yes, no), and SF-36 physical function score (in tertiles).
‡Global score is the average of the z-scores of TICS, immediate and delayed recalls of the East Boston Memory Test, delayed recall of TICS 10-word list, test of category fluency, and digit backwards test; verbal memory score is the average of the z-scores of the immediate and delayed recalls of the East Boston Memory Test and the immediate and delayed recalls of the TICS 10-word list.
§Telephone Interview of Cognitive Status.
‖Interaction between vigorous physical activity and APOE-e4 status was formally tested using a binary logistic regression model. The dependent variable was alternately the “medium-stable vs high-stable” cognitive trajectory or the “decline vs high-stable” cognitive trajectory.
Sensitivity Analyses
When age was used as the time scale, results were virtually unchanged. Similar results were also observed when including women with group-membership probabilities below than .70 or excluding women with any APOE-e2 allele although some results became nonsignificant, probably due to misclassification or loss of statistical power (data not shown). Lastly, we did not observe any significant interactions between APOE-e4 and total physical activity (Supplementary Table 1).
Discussion
In this longitudinal study with at least a 9-year lag between the physical activity assessment and the outcome evaluation, we observed 3 main patterns of cognitive evolution over the later life: a high and stable-, a medium and stable-, and a decline-type trajectory. We confirmed that APOE-e4 carriers had higher likelihood of less favorable cognitive trajectories and observed that lower midlife vigorous physical activity was associated with poorer cognition later in life. We observed some evidence of an interaction between APOE-e4 genotype and midlife vigorous physical activity with more marked inverse associations for decline with vigorous physical activity in APOE-e4 carriers compared to noncarriers.
Using a novel group-based trajectory modeling approach, our study highlighted 3 main types of cognitive evolution over 6 years of follow-up among women in their seventies: high-stable, medium-stable, and decline. This categorization complements other approaches considering cognitive decline as a binary variable or modeling it as a linear process (16,18–20).
Interestingly, the association between midlife vigorous physical activity and lower odds of cognitive decline remained significant after adjustment for several vascular risk factors, supporting the multidimensionality of the mechanisms linking physical activity and the brain (14).
Regarding the interaction between APOE-e4 genotype and physical activity on cognitive decline, the data have been inconsistent, probably due to major differences in study design (including sample size and statistical power), the way the outcome of cognitive decline was defined, the age of participants, and how, when, and what type of physical activity was evaluated. Three previous prospective studies have reported no interaction with cognitive function (18–20); however, 2 assessed cognitive function at one time (18,19), 2 had a short follow-up period (≤5 years) (18,19), none investigated vigorous physical activity specifically, and one assessed physical activity when participants were at a mean age of 70 years (20). One study of Alzheimer’s disease (41) investigated the potential APOE-e4 and physical activity interaction and observed no significant findings, but the follow-up of 5 years, particularly for Alzheimer’s disease, may have been too short (42). More in line with our results, 5 studies of various cognitive outcomes suggested an interaction between physical activity and APOE-e4, with a stronger direct association of physical activity among APOE-e4 carriers compared to noncarriers (16,43–46). Schuit and colleagues were the first to report among 560 Dutch men an interaction between physical activity and APOE genotype on cognitive decline, with a cognitive follow-up of 3 years but with concurrent assessment of physical activity (16). Rovio et al. in a longitudinal study with a follow-up of 20 years, and Luck et al. in the cross-sectional AgeCoDe German study observed an interaction between APOE-e4 and physical activity in dementia and Alzheimer’s disease, respectively (17,44). Etnier et al. (43), in a cross-sectional study of 94 women found that greater cardiorespiratory fitness (VO2peak assessed by maximal aerobic fitness test) was associated with better cognitive performance only in APOE-e4 carrier women. This result was consistent with APOE-e4–physical activity interaction studies of coronary artery disease, which suggested that high level of physical activity may be particularly important in APOE-e4 carriers (5) possibly via cardiovascular protection processes.
One point to keep in mind is that vigorous physical activity could be a good indicator of long-term overall physical activity. Such a measure would be less affected by memory bias than the assessment of less strenuous activities. Indeed, a recent mechanistic study observed that higher cardiorespiratory fitness was consistently associated with better neurocognitive function and preclinical decline (46,47). Two recent reviews and several mechanistic studies observed that higher levels of physical activity (that can be achieved in practice through vigorous activities in particular) may be more effective in reducing amyloid burden and activating the semantic memory-related neural circuits among APOE-e4 carriers than among APOE-e4 noncarriers (8,48–52). Overall, the relation between physical activity and cognitive function or decline have been mixed (29,53–61), due to a variety of reasons that have been already discussed (29).
A major strength of our study was the prospective evaluation of vigorous physical activity assessed about a decade before cognitive assessments. Some methodological limitations should be considered. First, our measure of vigorous physical activity level was based on self-report. Although our instrument has shown good properties as compared to past-week recalls or diaries during the year, it is less reliable than an objective measure. For most participants however, we were able to average 2 assessments of physical activity some years apart, making our indicator of midlife physical activity more robust. Second, our results are based on observational data and interpretation in terms of causality requires appropriate caution, particularly as we performed several tests, and we cannot rule out that findings, despite their robustness, may have been due to chance. Third, statistical power might have been limited, in particular since the group corresponding to the cognitive decline trajectories was small. Lastly, because the generalizability of results observed among mostly White well-educated women whose genotype data were available may be limited, further studies are needed in other populations.
To conclude, midlife vigorous physical activity was associated with better cognitive evolution in women in their seventies, with suggestions of stronger associations among APOE-e4 carriers.
Supplementary Material
Acknowledgments
The authors thank all the participants of the Nurses’ Health Study for their continued contributions.
Funding
This work was supported by the National Institutes of Health (R29 AG013482, R01AG015424, R01 AG036755, UM1 CA186107, and R01 CA49449). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Author Contributions
P.F.: Methodology, Software, Figures and tables, Writing—Original draft preparation; J.H.K.: Resources, Conceptualization, Methodology, Writing- Reviewing and Editing; I.-M.L.: Writing—Reviewing and Editing; F.G.: Resources, Writing—Reviewing and Editing; M.-N.V.: Conceptualization, Methodology, Supervision, Writing—Original draft preparation. All authors had final approval of the submitted manuscript.
Conflict of Interest
None declared.
References
- 1. Linn RT, Wolf PA, Bachman DL, et al. The ‘preclinical phase’ of probable Alzheimer’s disease. A 13-year prospective study of the Framingham cohort. Arch Neurol. 1995;52:485–490. doi:10.1001/archneur.1995.00540290075020 [DOI] [PubMed] [Google Scholar]
- 2. Arvanitakis Z, Shah RC, Bennett DA. Diagnosis and management of dementia: review. JAMA. 2019;322(16):1589–1599. doi:10.1001/jama.2019.4782 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Winblad B, Amouyel P, Andrieu S, et al. Defeating Alzheimer’s disease and other dementias: a priority for European science and society. Lancet Neurol. 2016;15(5):455–532. doi:10.1016/S1474-4422(16)00062-4 [DOI] [PubMed] [Google Scholar]
- 4. Haan MN, Shemanski L, Jagust WJ, Manolio TA, Kuller L. The role of APOE epsilon4 in modulating effects of other risk factors for cognitive decline in elderly persons. JAMA. 1999;282:40–46. doi:10.1001/jama.282.1.40 [DOI] [PubMed] [Google Scholar]
- 5. Raichlen DA, Alexander GE. Exercise, APOE genotype, and the evolution of the human lifespan. Trends Neurosci. 2014;37:247–255. doi:10.1016/j.tins.2014.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Verghese PB, Castellano JM, Garai K, et al. ApoE influences amyloid-β (Aβ) clearance despite minimal apoE/Aβ association in physiological conditions. Proc Natl Acad Sci USA. 2013;110:E1807–E1816. doi:10.1073/pnas.1220484110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Smith JD. Apolipoprotein E4: an allele associated with many diseases. Ann Med. 2000;32:118–127. doi:10.3109/07853890009011761 [DOI] [PubMed] [Google Scholar]
- 8. Bos MM, Noordam R, Blauw GJ, Slagboom PE, Rensen PCN, Van Heemst D. The ApoE ϵ4 isoform: can the risk of diseases be reduced by environmental factors? J Gerontol A Biol Sci Med Sci. 2019;74(1):99–107. doi:10.1093/gerona/gly226 [DOI] [PubMed] [Google Scholar]
- 9. Brecht WJ, Harris FM, Chang S, et al. Neuron-specific apolipoprotein e4 proteolysis is associated with increased tau phosphorylation in brains of transgenic mice. J Neurosci. 2004;24:2527–2534. doi:10.1523/JNEUROSCI.4315-03.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Wang R, Fratiglioni L, Laukka EJ, et al. Effects of vascular risk factors and APOE ε4 on white matter integrity and cognitive decline. Neurology. 2015;84:1128–1135. doi:10.1212/WNL.0000000000001379 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Smith JC, Lancaster MA, Nielson KA, et al. Interactive effects of physical activity and APOE-ε4 on white matter tract diffusivity in healthy elders. Neuroimage. 2016;131:102–112. doi:10.1016/j.neuroimage.2015.08.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Nichol K, Deeny SP, Seif J, Camaclang K, Cotman CW. Exercise improves cognition and hippocampal plasticity in APOE ε4 mice. Alzheimer’s Dement. 2009;5(4):287–294. doi:10.1016/j.jalz.2009.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Fukao K, Shimada K, Naito H, et al. Voluntary exercise ameliorates the progression of atherosclerotic lesion formation via anti-inflammatory effects in apolipoprotein E-deficient mice. J Atheroscler Thromb. 2010;17:1226–1236. doi:10.5551/jat.4788 [DOI] [PubMed] [Google Scholar]
- 14. Spartano NL, Pantoni L. What can longitudinal observational studies of physical activity teach us about prevention of dementia? Neurology. 2021;96(7):297. doi:10.1212/WNL.0000000000011376 [DOI] [PubMed] [Google Scholar]
- 15. Seshadri S. Prevention of dementia-thinking beyond the age and amyloid boxes. JAMA Neurol. 2020;77:160–161. doi:10.1001/jamaneurol.2019.3785 [DOI] [PubMed] [Google Scholar]
- 16. Schuit AJ, Feskens EJ, Launer LJ, Kromhout D. Physical activity and cognitive decline, the role of the apolipoprotein e4 allele. Med Sci Sports Exerc. 2001;33:772–777. doi:10.1097/00005768-200105000-00015 [DOI] [PubMed] [Google Scholar]
- 17. Luck T, Riedel-Heller SG, Luppa M, et al. Apolipoprotein E epsilon 4 genotype and a physically active lifestyle in late life: analysis of gene-environment interaction for the risk of dementia and Alzheimer’s disease dementia. Psychol Med. 2014;44:1319–1329. doi:10.1017/S0033291713001918 [DOI] [PubMed] [Google Scholar]
- 18. Lyall DM, Celis-Morales C, Lyall LM, et al. Assessing for interaction between APOE ε4, sex, and lifestyle on cognitive abilities. Neurology. 2019;92:e2691–e2698. doi:10.1212/WNL.0000000000007551 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Sabia S, Kivimaki M, Kumari M, Shipley MJ, Singh-Manoux A. Effect of Apolipoprotein E epsilon4 on the association between health behaviors and cognitive function in late midlife. Mol Neurodegener. 2010;5:23. doi:10.1186/1750-1326-5-23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Stringa N, van Schoor NM, Milaneschi Y, et al. Physical activity as moderator of the association between apoe and cognitive decline in older adults: results from three longitudinal cohort studies. J Gerontol A Biol Sci Med Sci. 2020;75:1880–1886. doi:10.1093/gerona/glaa054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Jones BL, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociol Methods Res. 2001;29(3):374–393. doi:10.1177/0049124101029003005. [Google Scholar]
- 22. Pietrzak RH, Lim YY, Ames D, et al. Trajectories of memory decline in preclinical Alzheimer’s disease: results from the Australian Imaging, Biomarkers and Lifestyle Flagship Study of ageing. Neurobiol Aging. 2015;36:1231–1238. doi:10.1016/j.neurobiolaging.2014.12.015 [DOI] [PubMed] [Google Scholar]
- 23. Devore EE, Grodstein F, van Rooij FJ, et al. Dietary antioxidants and long-term risk of dementia. Arch Neurol. 2010;67:819–825. doi:10.1001/archneurol.2010.144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Brandt J, Spencer M, Folstein M. The telephone interview for cognitive status. Neuropsych Neuropsychol Behav Neurol. 1988;1(2):111–117. [Google Scholar]
- 25. Kang JH, Grodstein F. Regular use of nonsteroidal anti-inflammatory drugs and cognitive function in aging women. Neurology. 2003;60:1591–1597. doi:10.1212/01.wnl.0000065980.33594.b7 [DOI] [PubMed] [Google Scholar]
- 26. Ainsworth BE, Haskell WL, Leon AS, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc. 1993;25:71–80. doi:10.1249/00005768-199301000-00011 [DOI] [PubMed] [Google Scholar]
- 27.WHO. https://www.who.int/dietphysicalactivity/physical_activity_intensity/en/ Global Strategy on Diet, Physical Activity and Health. What is Moderate-intensity and Vigorous-intensity Physical Activity?
- 28. Wolf AM, Hunter DJ, Colditz GA, et al. Reproducibility and validity of a self-administered physical activity questionnaire. Int J Epidemiol. 1994;23:991–999. doi:10.1093/ije/23.5.991 [DOI] [PubMed] [Google Scholar]
- 29. Wagner M, Grodstein F, Proust-Lima C, Samieri C. Long-term trajectories of body weight, diet, and physical activity from midlife through late life and subsequent cognitive decline in women. Am J Epidemiol. 2020;189:305–313. doi:10.1093/aje/kwz262 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Sabia S, Dugravot A, Dartigues JF, et al. Physical activity, cognitive decline, and risk of dementia: 28 year follow-up of Whitehall II cohort study. BMJ. 2017;357:j2709. doi:10.1136/bmj.j2709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Kang JH, Logroscino G, De Vivo I, Hunter D, Grodstein F. Apolipoprotein E, cardiovascular disease and cognitive function in aging women. Neurobiol Aging. 2005;26:475–484. doi:10.1016/j.neurobiolaging.2004.05.003 [DOI] [PubMed] [Google Scholar]
- 32. Lindström S, Loomis S, Turman C, et al. A comprehensive survey of genetic variation in 20,691 subjects from four large cohorts. PLoS One. 2017;12:e0173997. doi:10.1371/journal.pone.0173997 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Weuve J, Kang JH, Manson JE, Breteler MM, Ware JH, Grodstein F. Physical activity, including walking, and cognitive function in older women. JAMA. 2004;292:1454–1461. doi:10.1001/jama.292.12.1454 [DOI] [PubMed] [Google Scholar]
- 34. Vercambre MN, Grodstein F, Manson JE, Stampfer MJ, Kang JH. Physical activity and cognition in women with vascular conditions. Arch Intern Med. 2011;171(14):1244–1250. doi:10.1001/archinternmed.2011.282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Wang DD, Leung CW, Li Y, et al. Trends in dietary quality among adults in the United States, 1999 through 2010. JAMA Intern Med. 2014;174(10):1587–1595. doi:10.1001/jamainternmed.2014.3422 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Syddall HE, Martin HJ, Harwood RH, Cooper C, Aihie Sayer A. The SF-36: a simple, effective measure of mobility-disability for epidemiological studies. J Nutr Health Aging. 2009;13:57–62. doi:10.1007/s12603-009-0010-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Nagin DS. Analyzing developmental trajectories: a semi-parametric, group-based approach. Psychol Methods. 1999;4(2):139–157. doi:10.1037/1082-989x.4.2.139 [DOI] [PubMed] [Google Scholar]
- 38. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109–138. doi:10.1146/annurev.clinpsy.121208.131413 [DOI] [PubMed] [Google Scholar]
- 39. Andruff H, Carraro N, Thompson A, Gaudreau P, Louvet B. Latent class growth modelling: a tutorial. Tutor Quant Methods Psychol. 2009;5(1):11–24. doi:10.20982/tqmp.05.1.p011 [Google Scholar]
- 40. Insel PS, Hansson O, Mattsson-Carlgren N. Association between apolipoprotein E ε2 vs ε4, Age, and β-Amyloid in adults without cognitive impairment. JAMA Neurol. 2021;78(2):229. doi:10.1001/jamaneurol.2020.3780 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Lindsay J, Laurin D, Verreault R, et al. Risk factors for Alzheimer’s disease: a prospective analysis from the Canadian Study of Health and Aging. Am J Epidemiol. 2002;156:445–453. doi:10.1093/aje/kwf074 [DOI] [PubMed] [Google Scholar]
- 42. Baumeister SE, Karch A, Bahls M, Teumer A, Leitzmann MF, Baurecht H. Physical activity and risk of Alzheimer disease. Neurology. 2020;95(13):e1897–e1905. doi:10.1212/WNL.0000000000010013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Etnier JL, Caselli RJ, Reiman EM, et al. Cognitive performance in older women relative to ApoE-epsilon4 genotype and aerobic fitness. Med Sci Sports Exerc. 2007;39:199–207. doi:10.1249/01.mss.0000239399.85955.5e [DOI] [PubMed] [Google Scholar]
- 44. Rovio S, Kåreholt I, Helkala EL, et al. Leisure-time physical activity at midlife and the risk of dementia and Alzheimer’s disease. Lancet Neurol. 2005;4:705–711. doi:10.1016/S1474-4422(05)70198-8 [DOI] [PubMed] [Google Scholar]
- 45. Kivipelto M, Rovio S, Ngandu T, et al. Apolipoprotein E epsilon4 magnifies lifestyle risks for dementia: a population-based study. J Cell Mol Med. 2008;12(6B):2762–2771. doi:10.1111/j.1582-4934.2008.00296.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Deeny SP, Poeppel D, Zimmerman JB, et al. Exercise, APOE, and working memory: MEG and behavioral evidence for benefit of exercise in epsilon4 carriers. Biol Psychol. 2008;78:179–187. doi:10.1016/j.biopsycho.2008.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Aggio D, Smith L, Fisher A, Hamer M. Context-specific associations of physical activity and sedentary behavior with cognition in children. Am J Epidemiol. 2016;183:1075–1082. doi:10.1093/aje/kww031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Lautenschlager NT, Cox KL, Flicker L, et al. Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: a randomized trial. JAMA. 2008;300:1027–1037. doi:10.1001/jama.300.9.1027 [DOI] [PubMed] [Google Scholar]
- 49. John LW, Michael AS, Kristy AN, et al. Lifestyle and genetic contributions to cognitive decline and hippocampal structure and function in healthy aging. Curr Alzheimer Res. 2012;9(4):436–446. doi:10.2174/156720512800492477 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Head D, Bugg JM, Goate AM, et al. Exercise engagement as a moderator of the effects of APOE genotype on amyloid deposition. Arch Neurol. 2012;69:636–643. doi:10.1001/archneurol.2011.845 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Smith JC, Nielson KA, Woodard JL, et al. Physical activity reduces hippocampal atrophy in elders at genetic risk for Alzheimer’s disease. Front Aging Neurosci. 2014;6:61. doi:10.3389/fnagi.2014.00061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. de Frutos-Lucas J, Cuesta P, Ramírez-Toraño F, et al. Age and APOE genotype affect the relationship between objectively measured physical activity and power in the alpha band, a marker of brain disease. Alzheimers Res Ther. 2020;12:113. doi:10.1186/s13195-020-00681-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Baumgart M, Snyder HM, Carrillo MC, Fazio S, Kim H, Johns H. Summary of the evidence on modifiable risk factors for cognitive decline and dementia: a population-based perspective. Alzheimers Dement. 2015;11:718–726. doi:10.1016/j.jalz.2015.05.016 [DOI] [PubMed] [Google Scholar]
- 54. Blondell SJ, Hammersley-Mather R, Veerman JL. Does physical activity prevent cognitive decline and dementia?: A systematic review and meta-analysis of longitudinal studies. BMC Public Health. 2014;14:510. doi:10.1186/1471-2458-14-510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Xu W, Tan L, Wang HF, et al. Meta-analysis of modifiable risk factors for Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2015;86:1299–1306. doi:10.1136/jnnp-2015-310548 [DOI] [PubMed] [Google Scholar]
- 56. Ahlskog JE, Geda YE, Graff-Radford NR, Petersen RC. Physical exercise as a preventive or disease-modifying treatment of dementia and brain aging. Mayo Clin Proc. 2011;86(9):876–884. doi:10.4065/mcp.2011.0252 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential for primary prevention of Alzheimer’s disease: an analysis of population-based data. Lancet Neurol. 2014;13:788–794. doi:10.1016/S1474-4422(14)70136-X [DOI] [PubMed] [Google Scholar]
- 58. Morgan GS, Gallacher J, Bayer A, Fish M, Ebrahim S, Ben-Shlomo Y. Physical activity in middle-age and dementia in later life: findings from a prospective cohort of men in Caerphilly, South Wales and a meta-analysis. J Alzheimers Dis. 2012;31:569–580. doi:10.3233/JAD-2012-112171 [DOI] [PubMed] [Google Scholar]
- 59. Sofi F, Valecchi D, Bacci D, et al. Physical activity and risk of cognitive decline: a meta-analysis of prospective studies. J Intern Med. 2011;269:107–117. doi:10.1111/j.1365-2796.2010.02281.x [DOI] [PubMed] [Google Scholar]
- 60. Guure CB, Ibrahim NA, Adam MB, Said SM. Impact of physical activity on cognitive decline, dementia, and its subtypes: meta-analysis of prospective studies. Biomed Res Int. 2017;2017:9016924. doi:10.1155/2017/9016924 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Palta P, Sharrett AR, Gabriel KP, et al. Prospective analysis of leisure-time physical activity in midlife and beyond and brain damage on MRI in older adults. Neurology. 2021;96:e964–e974. doi:10.1212/WNL.0000000000011375 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.