Dementia is a common and debilitating disease that is characterized by a decline in memory and at least one other aspect of cognitive function that is of sufficient severity to interfere with daily activities. For the next 50 years, it is anticipated that the number of people in the United States with Alzheimer disease-the most common type of dementia-will nearly triple, from 4.5 million to 13.2 million (10). However, more than 4 million cases could be prevented by an intervention that reduces dementia risk by 50% or delays disease onset by 5 yr (4).
There is converging evidence from animal and human studies that physical activity holds tremendous promise for dementia prevention. In this article, we review the evidence that physical activity is associated with a reduced risk of cognitive decline and dementia, with a focus on our own work. We then recommend that prevention trials be performed to determine whether these associations are causal.
LONGITUDINAL OBSERVATIONAL STUDIES
Much of our work has focused on longitudinal observational studies of risk factors for cognitive decline and dementia. Because cognitive decline is integral to the diagnosis of dementia, many longitudinal studies use cognitive decline as a proxy measure for dementia. An important advantage of longitudinal studies is that data on the subjects can be collected over a long period; a disadvantage is that they are more susceptible to confounding and bias. In the next sections, we discuss the results of longitudinal, observational studies of the association between physical activity and cognitive decline/dementia.
Physical Activity and Cognitive Decline
In a longitudinal study of 5925 older women participating in a study of osteoporotic fractures, we found that women who reported engaging in more exercise at baseline were less likely to experience cognitive decline during 6-8 years of follow-up (23). Women were asked how many blocks they walked per day for exercise or as part of their normal routine. In addition, the total number of kilocalories (energy) expended per week in physical activities, recreation, and sport was calculated using a modified Paffenbarger Scale. Cognitive function was measured at baseline and 6-8 years later using a modified Mini-Mental State Examination (mMMSE), and cognitive decline was defined as a drop of three or more points (~1.5 standard deviations greater than the mean). Cognitive decline occurred in 17%, 18%, 22%, and 24% of the women in the highest, third, second, and lowest quartiles of blocks walked per week (P < 0.001 for trend over quartiles) (Table 1). After adjustment for demographic and health-related covariates, the odds of cognitive decline were reduced by 37% and 34% in women in the third and highest exercise quartiles, respectively, compared with the lowest quartile.
Cardiorespiratory Fitness and Cognitive Decline
Self-reported physical activity measures may be associated with measurement error or reporting bias. Therefore, in a second study, we examined the association between objectively measured cardiorespiratory fitness and cognitive decline in older adults (3). The study population was 349 noninstitutionalized adults aged 55 years or older living in the rural town of Sonoma, California, who were part of a larger study. Treadmill exercise tests were performed at baseline using a standard exercise protocol, and cardiorespiratory fitness measures included peak oxygen consumption (peak V˙o2, in milliliters per kilogram per minute), treadmill exercise duration (in minutes), and oxygen uptake efficiency slope (milliliters per minute oxygen/milliliters per minute ventilation), a fitness measure that is independent of effort. Cognitive function was assessed at baseline using the mMMSE and after 6 years of follow-up using the mMMSE and a detailed cognitive battery that included measures of attention/executive function (the ability to focus on a task, plan, and make decisions), verbal memory (the ability to learn and recall a list of words), and verbal fluency (the ability to generate words beginning with the letter s or to name different animals).
We found that subjects with better cardiorespiratory fitness at baseline experienced less cognitive decline on the mMMSE over 6 years (Fig. 1). Scores on the mMMSE declined by 0.0 points (95% confidence interval (CI), −0.3 to 0.2) for subjects in the highest peak V˙o2 tertile at baseline compared with −0.2 (95% CI, −0.5 to 0.0) points in the middle tertile and −0.5 (95% CI, −0.8 to −0.3) points in the lowest tertile (P = 0.002 for trend over tertiles). Subjects with better cardiorespiratory fitness at baseline also performed better on cognitive tests conducted 6 years later, especially on tests of attention/executive function.
Physical Activity and Risk of Dementia
Several other investigators have examined the association between physical activity and risk of dementia. In the Canadian Study of Health and Aging, a longitudinal study of 4615 elders without dementia, physical activity was associated with a 42% reduction in the odds of cognitive impairment (odds ratio (OR), 0.58; 95% CI, 0.41-0.83), a 50% reduction in the odds of Alzheimer disease (OR, 0.50; 95% CI, 0.28-0.90), and a 37% reduction in the odds of dementia of any type (OR, 0.63; 95% CI, 0.40-0.98) after adjustment for age, sex, and education (14).
Similarly, in a study conducted in Stockholm, Sweden, of 776 subjects aged 75 years or older who had no dementia (12), the risk of dementia was 39% lower in those who were physically active compared with those who were not (adjusted relative risk (RR), 0.61; 95% CI, 0.42-0.87). In the Honolulu-Asia Aging Study (1), the risk of dementia over 6 years was 44% lower in older Japanese men who walked more than 2 miles·d−1 compared with those who walked less than ¼ mile·d-1 (age-adjusted hazard ratio (HR), 0.56; 95% CI, 0.33-0.96).
Although some longitudinal studies have found no significant association between physical activity and risk of cognitive decline or dementia, this may have been due to low statistical power. For example, in the study by Wilson et al. (22), the risk of Alzheimer disease was reduced by 39% (RR, 0.61; 95% CI, 0.35-1.05) for those in the highest physical activity quartile compared with those in the lowest, which, although not statistically significant, is quite similar in magnitude to the studies previously discussed.
Summary of Longitudinal Studies
Most longitudinal studies have found that high physical activity is associated with a 30%-50% reduction in the risk of cognitive decline and dementia. The prospective nature of these studies, combined with the consistency of their results in different study populations and using different measures of physical activity and different cognitive outcomes, provides strong evidence in support of a causal association between physical activity and reduced risk of cognitive decline and dementia.
There are several potential mechanisms by which physical activity could directly or indirectly reduce the risk of cognitive decline and dementia (Fig. 2), and these have been reviewed in detail by others (6,13,15). In the next sections, we review evidence from our work indicating that many of the beneficial effects of physical activity-such as reduced risk of cardiovascular disease, diabetes, hypertension, and obesity-are associated with increased risk of cognitive decline and dementia. We also present evidence that physical activity may directly enhance neuronal health and function.
Reduced Vascular Risk
A decade ago, the surgeon general concluded that physical activity reduces the risk of cardiovascular disease, diabetes, hypertension, and obesity (18). Our work and the work of others have shown that these factors, in turn, are associated with a reduced risk of cognitive decline and dementia. In a study of 8845 members of the Kaiser Permanente Medical Care Program of Northern California, we found that midlife (aged 40-44 years) cardiovascular risk factors were associated with increased risk of dementia later in life (21). After adjustment for age, race, education, and sex, the hazard of dementia was increased by 24% in adults with midlife hypertension, 46% in those with midlife diabetes, 42% in those with midlife high cholesterol level, and 26% in those who were smokers in midlife (Table 2). Furthermore, the risk of dementia increased in an additive manner when subjects had two or more of these risk factors combined.
In another study of Kaiser members, we prospectively examined the role of obesity in middle age (aged 40-45 years) and risk of dementia more than two decades later (20). Obesity was determined by both body mass index (BMI) and skinfold thickness. Subjects who were obese (BMI ≥30.0 kg·m−2) or overweight (BMI 25.0-29.9 kg·m−2) in middle age were significantly more likely to develop dementia later in life compared with those with normal BMI (18.5-24.9 kg·m−2), with HR of 1.74 (95% CI, 1.34-2.26) and 1.35 (95% CI, 1.14-1.60), respectively, after adjustment for demographics and comorbid conditions in midlife and late life. Similarly, the risk of dementia was significantly increased by 60%-70% for those in the top 20% of subscapular and triceps skinfold thickness compared with those in the bottom 20% after adjustment. Interestingly, these associations remained strong after adjusting for vascular disease and vascular risk factors. There is a growing realization that adipose tissue is not inert but is hormonally active with inflammatory and metabolic components. Therefore, adipose tissue associated with obesity may lead to dementia directly or by increasing inflammation.
Reduced Levels of Inflammatory Markers
Adults who engage in regular physical activity also have lower serum levels of inflammatory markers (9), and our studies have found that high levels of inflammatory markers, in turn, are associated with increased risk of cognitive decline (25). In the Health, Aging, and Body Composition Study, which is a prospective study of more than 3000 white and African American older persons, we found that those in the highest inflammatory marker tertile were more likely to experience cognitive decline than those in the lowest tertile for interleukin 6 (26% vs 20%; age-adjusted OR, 1.34; 95% CI, 1.06-1.69) and for C-reactive protein (24% vs 19%; OR, 1.41; 95% CI, 1.10-1.79) but not for tumor necrosis factor α (23% vs 21%; OR, 1.12; 95% CI, 0.88-1.43).
Furthermore, we observed an interaction in this study population between high inflammation and metabolic syndrome-a clustering of cardiovascular risk factors that include abdominal obesity, high triglycerides, low high-density lipoprotein, hypertension, and hyperglycemia (24). The risk of cognitive impairment was 30% in those with metabolic syndrome and high inflammation compared with 23% in those with metabolic syndrome and low inflammation and 21% in those without metabolic syndrome, regardless of level of inflammation (Table 3).
Enhanced Neuronal Function
There is also compelling evidence from animal studies that physical activity may lead directly to enhanced neuronal health and function, including reduced brain damage after injury, angiogenesis (formation of new blood vessels), neurogenesis (formation of new neurons), synaptogenesis (formation of new synapses), and increased levels of neurotrophic factors (8). In one study, mice that were raised in cages with running wheels, as opposed to standard cages without running wheels, experienced enhanced neurogenesis and synaptic plasticity in the dentate gyrus of the hippocampus, an area of the brain that is critical for learning and memory, and also learned to perform a water maze task more quickly (19). In another study using a transgenic mouse model, mice raised with running wheels experienced decreased deposition of amyloid-β, which is a pathological hallmark of Alzheimer disease (2). There is evidence that some of these effects may be caused by increases in levels of brain-derived neurotrophic factor, which increases strongly in response to running in mice and is important for the growth and survival of neurons (8).
These animal studies are supported by a handful of studies in humans. A study of retirees found that those who became inactive after retirement experienced significant declines in cerebral blood flow over 4 years, whereas those who continued to work or engaged in regular physical activity did not (16). Another study found that, after adjusting for age, higher levels of aerobic fitness were correlated cross-sectionally with higher brain tissue densities (7).
Summary of Potential Mechanisms
There are several potential mechanisms that may underlie the association between physical activity and reduced risk of cognitive decline and dementia. Physical activity could lead to reduced vascular risk, obesity, or levels of inflammatory markers, all of which are interrelated and have been associated with reduced risk of cognitive decline and dementia. Physical activity could also directly lead to enhanced neuronal health and function, which could minimize the clinical impact of neuronal loss that may occur early in the dementia process. The existence of plausible biological mechanisms provides further evidence in support of a causal association between physical activity and reduced risk of cognitive decline and dementia in older adults.
RANDOMIZED CONTROLLED TRIALS
Despite the evidence presented in the preceding sections, randomized controlled trials are usually considered the gold standard for proving that an association is causal. In the next sections, we describe the results of short-term exercise intervention studies in older adults and argue that longer-term prevention trials are recommended.
Short-term Impact of Exercise Interventions
Several randomized, controlled trials during the past three decades have evaluated the short-term impact of exercise interventions on cognition in healthy and sedentary elders without cognitive impairment. Most of these trials were 6 months or less in duration. A meta-analysis of 18 trials found that, on average, exercise groups experienced greater improvement in cognitive function than control groups (0.478 SD vs 0.164 SD, respectively; P < 0.05) (5). In addition, the benefits of exercise were most pronounced in the domains of executive function, visuospatial function, and mental processing speed, which is consistent with our study of cardiorespiratory fitness and cognitive decline described previously (3) and supportive of a vascular mechanism.
More recently, several randomized controlled trials have evaluated the impact of exercise interventions in older adults with cognitive impairment, such as those living in a nursing home. A meta-analysis of 10 studies found that, on average, improvement in cognitive function was 0.57 SD (P < 0.001) higher in the exercise groups compared with the control group (11).
Taken together, these studies provide strong evidence that physical activity results in short-term improvements in cognitive function in sedentary elders, regardless of their level of cognitive impairment. However, to date, no trials have directly addressed the question of whether physical activity prevents or slows cognitive decline and dementia over longer periods. This is primarily due to the logistical difficulty of carrying out long-term behavioral interventions. It remains unknown whether physical activity results primarily in small increases in cognitive function or whether it fundamentally alters the brain to prevent or minimize the impact of the neurodegenerative processes that underlie dementia.
The Need for Prevention Trials
We have presented evidence that older adults who are physically active are less likely to experience cognitive decline and dementia and that sedentary elders who begin exercise programs experience short-term gains in cognitive function. In addition, we have described several plausible biological mechanisms by which physical activity could lower dementia risk. Is this enough to establish causality?
The Women's Health Initiative Memory Study (WHIMS) (17) is an important reminder that randomized controlled trials are desirable to confirm the results of observational and animal studies whenever feasible. A multitude of studies had suggested that short-term estrogen therapy improved cognitive function in older women and was associated with a reduced risk of cognitive decline and dementia, and animal studies had suggested several plausible biological mechanisms. However, the WHIMS trial found that the risk of dementia was actually doubled in older women treated with estrogen compared with those treated with placebo.
There is still much debate about the reasons for the discrepancy between WHIMS and other studies. Some argue that the observational and animal studies got the wrong answer because of bias or inherent differences between humans and animals; others argue that the WHIMS trial got the wrong answer because estrogen was administered later in women's lives rather than during the "critical period" after menopause. However, there is no doubt that the results of the trial differed dramatically from what was expected. We may never know whether physical activity truly prevents cognitive decline and dementia until prevention trials are performed to test this specific hypothesis.
Dementia is a common and debilitating disease that will affect an increasingly large percentage of our population for the next 50 years unless a prevention or treatment is found. We have provided an overview of the evidence that physical activity is associated with a reduced risk of cognitive decline and dementia in older adults. There is consistent evidence from longitudinal studies that older adults who are physically active or have better cardiorespiratory fitness are less likely to experience cognitive decline and dementia. In addition, several plausible biological mechanisms have been identified, including reduced vascular risk, obesity, and levels of inflammatory markers and enhanced neuronal health and function. Exercise interventions are associated with short-term improvements in cognitive function in sedentary elders. Longer-term trials are recommended to determine whether physical activity truly prevents cognitive decline and dementia in the elderly.
1. Abbott, R.D., L.R. White, G.W. Ross, K.H. Masaki, J.D. Curb, and H. Petrovitch. Walking and dementia in physically capable elderly men. JAMA
2. Adlard, P.A., V.M. Perreau, V. Pop, and C.W. Cotman. Voluntary exercise decreases amyloid load in a transgenic model of Alzheimer's disease. J. Neurosci.
3. Barnes, D.E., K. Yaffe, W.A. Satariano, and I.B. Tager. A longitudinal study of cardiorespiratory fitness and cognitive function in healthy older adults. J. Am. Geriatr. Soc.
4. Brookmeyer, R., S. Gray, and C. Kawas. Projections of Alzheimer's disease in the United States and the public health impact of delaying disease onset. Am. J. Public Health
5. Colcombe, S., and A.F. Kramer. Fitness effects on the cognitive function of older adults: a meta-analytic study. Psychol. Sci.
6. Colcombe, S.J., A.F. Kramer, E. McAuley, K.I. Erickson, and P. Scalf. Neurocognitive aging and cardiovascular fitness: recent findings and future directions. J. Mol. Neurosci.
7. Colcombe, S.J., K.I. Erickson, N. Raz, A.G. Webb, N.J. Cohen, E.McAuley, and A.F. Kramer. Aerobic fitness reduces brain tissue loss in aging humans. J. Gerontol. A. Biol. Sci. Med. Sci.
8. Cotman, C.W., and N.C. Berchtold. Exercise: a behavioral intervention to enhance brain health and plasticity. Trends Neurosci.
9. Ford, E.S. Does exercise reduce inflammation? Physical activity and C-reactive protein among U.S. adults. Epidemiology
10. Hebert, L.E., P.A. Scherr, J.L. Bienias, D.A. Bennett, and D.A. Evans. Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch. Neurol.
11. Heyn, P., B.C. Abreu, and K.J. Ottenbacher. The effects of exercise training on elderly persons with cognitive impairment and dementia: a meta-analysis. Arch. Phys. Med. Rehabil.
12. Karp, A., S. Paillard-Borg, H.X. Wang, M. Silverstein, B. Winblad, and L. Fratiglioni. Mental, physical and social components in leisure activities equally contribute to decrease dementia risk. Dement. Geriatr. Cogn. Disord.
13. Kramer, A.F., S.J. Colcombe, E. McAuley, P.E. Scalf, and K.I. Erickson. Fitness, aging and neurocognitive function. Neurobiol. Aging
26(suppl 1):124-127, 2005.
14. Laurin, D., R. Verreault, J. Lindsay, K. MacPherson, and K. Rockwood. Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch. Neurol.
15. McAuley, E., A.F. Kramer, and S.J. Colcombe. Cardiovascular fitness and neurocognitive function in older adults: a brief review. Brain Behav. Immun.
16. Rogers, R.L., J.S. Meyer, and K.F. Mortel. After reaching retirement age physical activity sustains cerebral perfusion and cognition. J. Am. Geriatr. Soc.
17. Shumaker, S.A., C. Legault, S.R. Rapp, L. Thal, R.B. Wallace, J.K. Ockene, S.L. Hendrix, B.N. Jones 3rd, A.R. Assaf, R.D. Jackson, J.M. Kotchen, S. Wassertheil-Smoller, J. Wactawski-Wende, and WHIMS Investigators. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women's Health Initiative Memory Study: a randomized controlled trial. JAMA
18. U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General
. Atlanta, GA: U.S. Department of Health and Human Services, CDC, National Center for Chronic Disease Prevention and Health Promotion; 1996.
19. van Praag, H., B.R. Christie, T.J. Sejnowski, and F.H. Gage. Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc. Natl. Acad. Sci. U. S. A.
20. Whitmer, R.A., E.P. Gunderson, E. Barrett-Connor, C.P. Quesenberry Jr., and K. Yaffe. Obesity in middle age and future risk of dementia: a 27 year longitudinal population based study. Br. Med. J.
21. Whitmer, R.A., S. Sidney, J. Selby, S.C. Johnston, and K. Yaffe. Midlife cardiovascular risk factors and risk of dementia in late life. Neurology
22. Wilson, R.S., C.F. Mendes De Leon, L.L. Barnes, J.A. Schneider, J.L. Bienias, D.A. Evans, and D.A. Bennett. Participation in cognitively stimulating activities and risk of incident Alzheimer disease. JAMA
23. Yaffe, K., D. Barnes, M. Nevitt, L.Y. Lui, and K. Covinsky. A prospective study of physical activity and cognitive decline in elderly women: women who walk. Arch. Intern. Med.
24. Yaffe, K., A. Kanaya, K. Lindquist, E.M. Simonsick, T. Harris, R.I. Shorr, F.A. Tylavsky, and A.B. Newman. The metabolic syndrome, inflammation, and risk of cognitive decline. JAMA
25. Yaffe, K., K. Lindquist, B.W. Penninx, E.M. Simonsick, M. Pahor, S. Kritchevsky, L. Launer, L. Kuller, S. Rubin, and T. Harris. Inflammatory markers and cognition in well-functioning African-American and white elders. Neurology