The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 63:62-66 (2008)
© 2008 The Gerontological Society of America
Physical Exercise at Midlife and Risk of Dementia Three Decades Later: A Population-Based Study of Swedish Twins
Ross Andel,
Michael Crowe,
Nancy L. Pedersen,
Laura Fratiglioni,
Boo Johansson and
Margaret Gatz
1 School of Aging Studies and Florida Alzheimer's Disease Center, University of South Florida, Tampa.
2 Department of Psychology, University of Alabama at Birmingham.
3 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden.
4 Department of Psychology, University of Southern California, Los Angeles.
5 Aging Research Center, Karolinska Institutet, Stockholm, Sweden.
6 Department of Psychology, Göteborg University, Sweden.
Address correspondence to Ross Andel, PhD, School of Aging Studies, 4202 East Fowler Ave., MHC 1321, University of South Florida, Tampa, Florida 33620. E-mail: randel{at}cas.usf.edu
 |
Abstract
|
|---|
Background. With the number of people with dementia increasing, identifying potential protective factors has become more important. We explored the association between physical exercise at midlife and subsequent risk of dementia among members of the HARMONY study.
Methods. Measures of exercise were obtained by the Swedish Twin Registry an average of 31 years prior to dementia assessment. Dementia was diagnosed using a two-stage procedure—screening for cognitive impairment followed by full clinical evaluation. We used two study designs: case–control analyses included 264 cases with dementia (176 had Alzheimer's disease) and 2870 controls; co-twin control analyses included 90 twin pairs discordant for dementia.
Results. In case–control analyses, controlling for age, sex, education, diet (eating fruits and vegetables), smoking, drinking alcohol, body mass index, and angina, light exercise such as gardening or walking and regular exercise involving sports were associated with reduced odds of dementia compared to hardly any exercise (odds ratio [OR] = 0.63, 95% confidence interval [CI], 0.43–0.91 for light exercise; OR = 0.34, 95% CI, 0.16–0.72 for regular exercise). Findings were similar for Alzheimer's disease alone. In co-twin control analyses, controlling for education, the association between higher levels of exercise and lower odds of dementia approached significance (OR = 0.50, 95% CI, 0.23–1.06; p =.072).
Conclusions. Exercise at midlife may reduce the odds of dementia in older adulthood, suggesting that exercise interventions should be explored as a potential strategy for delaying disease onset.
Key Words: Exercise • Dementia • Twins • Sweden
PHYSICAL exercise has long been recognized as an effective strategy to promote physical health and reduce the risk of cardiovascular and all-cause mortality (1,2). Recently, reviews of animal and human studies have suggested that regular exercise may support cerebrovascular health, reduce oxidative stress and brain tissue loss, and facilitate normal cognitive function (3–6), offering biological mechanisms for the influence of exercise on risk of Alzheimer's disease (AD) and other types of dementia. With the number of dementia cases expected to grow in the coming decades (7), finding potential protective factors has become increasingly important.
So far, results from prospective longitudinal studies with older adults have been inconsistent, with some reporting lower risk of dementia and AD associated with regular exercise (8–10) and others reporting null findings (11,12). The follow-up time in these studies, which averaged between 4 and 7 years, leaves unanswered the possibility that even affirmative findings may be partially attributable to factors associated with preclinical dementia. Only two prospective studies have used a baseline assessment of activity at middle age (13,14). However, even these two studies yielded conflicting findings. Another issue that deserves attention is exactly how much physical activity is necessary to reap beneficial effects on cognitive health.
Finally, questions remain regarding the role of genetic influences on the association between exercise and dementia (9,13). Testing for genetic and familial influences may be important because dementia appears to be highly heritable (15) and the tendency to be physically active may be partially attributable to familial factors (16).
We examined the association between level of physical exercise and risk of dementia 31 years later using a population-based cohort from the Swedish Twin Registry (STR). We analyzed data with a case–control design, where cases are compared on exposure to unrelated (external) controls, and with a co-twin control design, where the control group consists of related (internal) controls (i.e., nondemented twin partners to the twins with dementia). In the case–control analyses, we tested whether light exercise such as walking or gardening, regular exercise involving sports, or hard physical training at midlife would be associated with reduced risk of dementia and AD later in life compared to little or no exercise. We also explored potential effect modification by other lifestyle indicators including smoking, alcohol consumption, diet, and body mass index. In co-twin control analyses, we examined twin pairs discordant for dementia to test whether the twin with dementia reported a lower level of physical exercise at midlife compared to the nondemented co-twin. When using co-twin controls, unmeasured genetic and familial factors such as family history of the disease, early-life habits, childhood socioeconomic status, and cognitive abilities are accounted for in the design.
|
METHODS
|
|---|
Participants
Participants were drawn from the STR, a population-based study of twins (17). The STR consists of several cohorts of twin pairs, the oldest of which consists of like-sexed twins born before 1926. This study used all members of the older cohort who completed a questionnaire in 1967 (or in 1970 for those who did not answer in 1967) and underwent dementia assessment in 1998 or later as part of the HARMONY study (18). HARMONY is taken from the Swedish words for "health" (Hälsa), "genes" (ARv), "environment" (Miljö), "and" (Och), and "new" (NY).
In 1998, the HARMONY study began follow-up of all living twins in the STR who were 65 years old or older. This follow-up involved telephone screening of the twin (or an informant if the twin performed poorly or was unable to be interviewed) followed by in-person clinical evaluations for dementia in those who screened positive for cognitive impairment as well as their co-twins. Potential participants were located by means of matching with a national address registry. Thus, for those in long-term care, either a family member or an aide served as an informant, and even severely demented individuals could be included in the analyses. A comprehensive discussion of the study design, diagnostic procedures, and participation rates can be found elsewhere (18).
In all, 4506 individuals with complete information from 1967 or 1970 were eligible to be contacted by HARMONY when the follow-up began. Of these, 3366 participated in telephone screening and/or clinical work-up for dementia. Among the dropouts, 730 refused to participate, 173 could not be reached, 155 could not be interviewed due to physical problems and an informant was not available, and 82 died before they could be interviewed. An additional 232 persons were screened as suspect for cognitive impairment but were not worked up. Therefore, 3134 individuals (including 655 twin pairs) were used in the analyses, reflecting a 70% response rate. The most common reason for having only one member of a twin pair in the study was death of the co-twin. Average age was 48.1 years (standard deviation [SD] = 4.9) at baseline and 79.5 (SD = 5.0) at the cognitive screening follow-up. The sample included more women (61%) than men.
Measures
Physical exercise.--
Physical exercise was assessed in a questionnaire from 1967 or 1970. Respondents were asked: "How much exercise have you had from age 25 to 50?" Level of exercise was measured on a 4-point scale as 0 (hardly any exercise), 1 (light exercise such as walking or light gardening), 2 (regular exercise involving sports), and 3 (hard physical training).
Covariates.--
Covariates included age (continuous), sex (men/women), education (basic/more than basic), and baseline values reported on the following variables: portion of fruits and vegetables in the diet (small or no part/medium or great part), body mass index (BMI; = weight [kg]/height squared [m2]; measured as BMI < 25/BMI 
25), alcohol drinking (no drinks per week/one or more drinks per week), current smoking status (yes/no), and angina pectoris (yes/no).
Data Analysis
In case–control analyses, we used logistic regression to calculate odds of dementia and AD for light exercise, regular exercise involving sports, and hard physical training with hardly any exercise used as a reference value. Only participants with complete data on all study variables were used. In model 1, we estimated unadjusted odds ratios (ORs) and 95% confidence intervals (CIs). In model 2, we adjusted the results for age, sex, and education. In model 3, we added the other covariates. Finally, we explored effect modification due to lifestyle by adding exercise as an ordinal variable and the product of exercise and each lifestyle variable (eating fruits and vegetables, alcohol consumption, and BMI) separately into a fully adjusted model. All CI values were adjusted for data dependence due to including two members of the same family in the analyses (19).
In co-twin control analyses, we used conditional logistic regression to estimate OR and 95% CI values based on within-pair comparisons. The measure of exercise was dichotomized so that a twin pair member with a higher level of exercise was assigned a 1 and the less active twin was assigned a 0. When twins from the same pair reported the same level of exercise, both twins were assigned a 0 because these twin pairs did not contribute to the calculation of the point estimate. Results show whether the twin with dementia had a higher level of physical exercise compared to his or her co-twin. In model 2, we adjusted for education, but age and sex were controlled by design. Because of limited sample size, we included all disease-discordant pairs with data for education and physical exercise. In model 3, analyses were restricted to pairs with complete data on all study variables. We used SAS software, version 9 (20) and a two-tailed significance level of 0.05 in all analyses.
|
RESULTS
|
|---|
Case–Control Analyses
When the 3134 participants were compared to the 1372 dropouts, the dropouts were on average about 1.5 years older (t = 7.36, p <.001), more likely to be women (69% of the dropouts vs. 61% of the participants, 
2 = 26.03, p <.001), less likely to have attained a high level of education (23% vs. 37%, 2 = 18.60, p <.001), and less likely to drink alcohol (62% vs. 70%, 2 = 22.72, p <.001). The dropouts also differed from the participants with respect to exercise (2for trend = 4.49, p =.03) such that 9% of the dropouts versus 11% of the participants engaged in hardly any exercise, 73% versus 71% engaged in light exercise, 6% versus 9% engaged in regular exercise, and 11% versus 8% engaged in hard physical training. The dropouts were similar to the participants with respect to BMI, smoking status, and diet (p >.05).
Characteristics of participants in case–control analyses by level of exercise are shown in Table 1. The majority of participants engaged in light exercise such as walking or gardening, whereas only 18% participated in regular exercise involving sports or in hard physical training. There were more women engaged in no or light exercise and more men who exercised regularly or hard. Follow-up time differed significantly across levels of exercise, although the greatest between-group difference was <4 months, and there was no between-group difference in age at baseline or age at follow-up.
Characteristics of the case–control sample by dementia status are shown in Table 2. Of the 3134 participants, 264 were diagnosed with dementia (176 with AD) and 2870 were nondemented controls. Cases were older than controls, but the groups did not differ in time of follow-up.
Results for the association between physical exercise at midlife and dementia or AD are summarized in Table 3. Light and regular exercise were associated with reduced odds of dementia and AD compared to hardly any exercise. There was only a small change in the effect of exercise when results were adjusted for covariates. In the full models, the odds of dementia were 37% lower for light and 66% lower for regular exercise compared to hardly any exercise. The results were similar for AD, although the association between light exercise and AD was marginal. In the unadjusted models, there was a significant linear trend across levels of exercise, but this result was explained by the covariates. Tests of potential effect modification (results not shown) by smoking, alcohol consumption, eating fruits and vegetables, or BMI yielded no significant results, but the interaction of Exercise x BMI approached significance (p =.101 for dementia and.089 for AD). Post hoc analysis of data stratified by BMI indicated that there was a stronger protective effect for exercise among overweight participants (OR = 0.61, 95% CI, 0.35–1.08 for dementia; OR = 0.52, 95% CI, 0.24–1.14 for AD) than among normal weight participants (OR = 0.92, 95% CI, 0.73–1.16 for dementia; OR = 0.90, 95% CI, 0.69–1.19 for AD).
View this table:
[in this window]
[in a new window]
|
Table 3. Case–Control Analyses of the Association between Exercise at Midlife and Dementia or Alzheimer's Disease (AD).
|
|
Co-Twin Control Analyses
There were 90 twin pairs discordant for dementia, of which 32 pairs were monozygotic. Characteristics of the co-twin control sample are presented in Table 4.
Results of conditional logistic regression analyses are summarized in Table 5. Using all disease-discordant twin pairs, there was a statistical trend indicating that twins who exercised more than their co-twins had reduced odds of dementia in analyses with and without controlling for education. When we restricted the sample in these analyses to twin pairs with complete data on all study variables and controlled for all covariates, the point estimates were in the same direction but not significant. No other results approached the 0.05 level of significance. Findings were similar for twin pairs discordant for AD.
|
DISCUSSION
|
|---|
We examined the association between level of exercise in midlife and dementia or AD an average of 31 years later. In case–control analyses, we found that light exercise such as gardening or walking and regular exercise involving sports were associated with reduced odds of dementia and AD compared to hardly any exercise, controlling for age, sex, education, baseline smoking, drinking, diet (eating fruits and vegetables), BMI, and angina. The association was particularly strong for regular exercise. Hard physical training was not associated with significantly reduced odds of disease, although OR values were below 1.0. For the most part, covariates did not appear to have a substantial influence on the results, with the possible exception of BMI, where the association between higher levels of exercise and lower odds of dementia was stronger for overweight than for normal weight participants.
In co-twin control analyses, point estimates indicated that twins who reported engaging in higher levels of exercise than their co-twins had reduced odds of dementia and AD. However, the results did not reach statistical significance, probably due to limited sample size. Statistical power to detect significant differences at the 0.05 level in co-twin control analyses was approximately 63% for an OR of 0.5 (in univariate analyses with all 90 available twin pairs). Limited sample size for the co-twin design also precluded a statistical test of a dose effect of exercise or a test of effect by specific levels of exercise. Still, the findings were consistent with the assertion that exercise may be protective against dementia even when genetic and familial factors shared by twins are controlled.
Our findings provide additional evidence that midlife exercise may influence risk of dementia decades later. The length of time between assessment of exercise and dementia diagnosis makes it less likely that the reported level of exercise reflected changes due to oncoming dementia. Our findings are consistent with the notion that exercise may support cerebrovascular health (4), although we were unable to test biological mechanisms directly. Similar to our study, one recent report found reduced risk of dementia in participants who exercised regularly 21 years earlier (13). In contrast, another study found no effect of physical activity on dementia 25–30 years later (14). However, the latter study combined leisure time physical activity with work-related physical activity, which may have introduced certain confounds specific to occupation. Unique features of our study include the follow-up time of >30 years, measurement of exercise across four levels, and use of two designs within one study: case–control and co-twin control.
Little is known about the relationship between dose and response with respect to exercise and dementia. In our study, it is worth noting that the participants who reported engaging in hard physical training were more likely than those who did light or regular exercise to have a low level of education, a low proportion of fruits and vegetables in their diet, and a high BMI, all of which may reflect lower socioeconomic status, which has been shown to be a risk factor for dementia (21).
Several limitations should be mentioned. First, physical exercise was measured with self-report data obtained at one time point. With self-report, questions may be interpreted differently by different responders. Social desirability bias is also possible in self-reports, especially on items such as body weight and height. Objective data collected over time would likely provide a better assessment of exercise during midlife. Second, we used prevalent cases of dementia, which can sometimes lead to bias due to differential survival. However, if physical inactivity increases the risk of mortality, then our results would likely underestimate the association with dementia. Third, we could not control for specific physical conditions other than angina that may have affected ability to engage in exercise. Finally, we were only able to identify a small number of twin pairs discordant for dementia.
Conclusion
We found that participants who engaged in light or regular exercise had reduced odds of dementia and AD approximately 31 years later. These results were not explained by age, sex, education, smoking, alcohol consumption, proportion of fruits and vegetables in the diet, BMI, or angina. Despite a limited sample size, co-twin control analyses yielded a similar pattern of results, suggesting that the exercise–dementia association may not be explained by factors shared by family members. Overall, these findings suggest that physical exercise may provide a widely available strategy to delay the onset of dementia. Further research is needed regarding potential benefit from specific types of exercise and the potential biological mechanisms underlying the association between exercise and cognitive health.
|
Acknowledgments
|
|---|
This work was supported by National Institute on Aging (NIA) grants R01 AG08724 and P30 AG17265, and by an Alzheimer's Association/Zenith Fellows Award.
|
Footnotes
|
|---|
Decision Editor: Luigi Ferrucci, MD, PhD
Received October 27, 2006
Accepted March 31, 2007
|
References
|
|---|
- Kesaniemi YA, Danforth E, Jr, Jensen MD, Kopelman PG, Lefebvre P, Reeder BA. Dose-response issues concerning physical activity and health: an evidence-based symposium. Med Sci Sport Exerc. 2001;33:S351-S358.[Medline]
- Thompson PD, Buchner D, Pi
a IL, et al. Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular disease. Circulation. 2003;107:3109-3116.[Free Full Text] - Churchill JD, Galvez R, Colcombe S, Swain RA, Kramer AF, Greenough WT. Exercise, experience and the aging brain. Neurobiol Aging. 2002;23:941-955.[Medline]
- Dishman RK, Berthoud HR, Booth FW, et al. Neurobiology of exercise. Obesity. 2006;14:345-356.[Medline]
- Cotman CW, Berchtold NC. Exercise: behavioral intervention to enhance brain health and plasticity. Trends Neurosci. 2002;25:295-301.[Medline]
- Kiraly MA, Kiraly SJ. The effect of exercise on hippocampal integrity: review of recent research. Int J Psychiatry Med. 2005;35:75-89.[Medline]
- Ferri CP, Prince M, Brayne C, et al. Global prevalence of dementia: a Delphi consensus study. Lancet. 2005;366:2112-2117.[Medline]
- Larson EB, Wang L, Bowen JD, et al. Exercise is associated with reduced risk of incident dementia among persons 65 years of age and older. Ann Intern Med. 2006;144:73-81.[Abstract/Free Full Text]
- Podewils LJ, Guallar E, Kuller LH, et al. Physical activity, APOE genotype, and dementia risk: findings from the Cardiovascular Health Cognition Study. Am J Epidemiol. 2005;161:639-651.[Abstract/Free Full Text]
- Laurin D, Verreault R, Lindsay J, et al. Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch Neurol. 2001;58:498-504.[Abstract/Free Full Text]
- Verghese J, Lipton RB, Hall CB, et al. Low blood pressure and the risk of dementia in very old individuals. Neurology. 2003;61:1667-1672.[Abstract/Free Full Text]
- Wilson RJ, Bennett DA, Bienias JL, et al. Cognitive activity and incident AD in a population-based sample of older persons. Neurology. 2002;59:1910-1914.[Abstract/Free Full Text]
- Rovio S, Kåreholt I, Helkala E, et al. Leisure-time physical activity at midlife and the risk of dementia and Alzheimer's disease. Lancet Neurol. 2005;4:705-711.[Medline]
- Yamada M, Kasagi F, Sasaki H, Masunari N, Mimori Y, Suzuki G. Association between dementia and midlife risk factors: the Radiation Effects Research Foundation Adult Health Study. J Am Geriatr Soc. 2003;51:410-414.[Medline]
- Gatz M, Reynolds CA, Fratiglioni L, et al. Role of genes and environments for explaining Alzheimer's disease. Arch Gen Psychiatry. 2006;63:168-174.[Abstract/Free Full Text]
- Carlsson S, Andersson T, Lichtenstein P, Michaëlsson K, Ahlbom A. Genetic effects on physical activity: results from the Swedish Twin Registry. Med Sci Sports Exerc. 2006;38:1396-1401.
- Lichtenstein P, De Faire U, Floderus B, et al. The Swedish Twin Registry: a unique source for clinical, epidemiological, and genetic studies. J Intern Med. 2002;252:184-205.[Medline]
- Gatz M, Fratiglioni L, Johansson B, et al. Complete ascertainment of dementia in the Swedish Twin Registry: the HARMONY study. Neurobiol Aging. 2005;26:439-447.[Medline]
- Moradi T, Hans-Olov A, Ekrom A, Wendrén TP, Floderus B, Lichtenstein P. Physical activity and risk for breast cancer: a prospective cohort study among Swedish twins. Int J Cancer. 2002;100:76-81.[Medline]
- SAS Institute. SAS System for Microsoft Windows (Version 9). Cary, NC: SAS Institute Inc.; 2003.
- Qiu C, Karp A, von Strauss E, Winblad B, Fratiglioni L, Bellander T. Lifetime principal occupation and risk of Alzheimer's disease in the Kungsholmen project. Am J Ind Med. 2003;43:204-211.[Medline]
Top
Abstract
Methods
