Physical activity and cognitive decline, the role of the... : Medicine & Science in Sports & Exercise (original) (raw)
It has been well-documented that growing old is accompanied with a gradual decline in cognitive functioning (3,4,6,26). Therefore, in our aging society, we may expect an increase in the number of people with cognitive impairment. In the last decade, age, sex, education, smoking, high alcohol consumption, head trauma, and family history have been identified as risk factors for cognitive impairment (8,11,16,28). Cardiovascular disease, atherosclerosis, and risk factors such as hypertension, diabetes mellitus, serum total cholesterol, and fibrinogen concentration have also emerged as risk factors (22,27,29). In this context, a preventive role of regular physical activity has been suggested.
Several epidemiological studies have investigated the association between physical activity and cognitive functioning (6,7,10). A meta-analysis demonstrated that physical activity was directly associated with reaction time, analytic capacity, and acuity (6). The benefits were greater for chronic physical activity than acute bouts of activity. However, nearly all of the studies were cross-sectional, precluding conclusions on temporality. Intervention studies report inconsistent results (9,19,31–33). Moreover, most of them lack a control group and have small sample sizes. The extent to which physical activity can affect the development of cognitive impairment remains debatable.
In this study, we investigated the association between physical activity and cognitive decline in a prospective study of a population-based cohort of elderly men in the Netherlands. As the apolipoprotein e4 allele is associated with a substantially increased risk for cognitive impairment (12,37), the association between physical activity and cognitive decline may differ between carriers and noncarriers of the APOE*4 allele. In other words, the importance of regular physical activity in people with a genetically susceptibility may be different from those who are not genetically predisposed. Therefore, we investigated whether the association varied across the carriers and noncarriers of the APOE*4 allele.
METHODS
Subjects.
The Zutphen Elderly Study is a longitudinal study of risk factors for chronic diseases in men, aged 65–84 at baseline, living in Zutphen, a town in the eastern part of the Netherlands (13). The Zutphen study is the Dutch contribution to the Seven Countries Study (25). In the spring of 1990, 560 men (78%) were examined. In 1993, 390 of 553 surviving men were reexamined. About 88% of the participants lived independently at home. In 1993, a nonresponse survey was carried out. Nonrespondents did not differ significantly from respondents in terms of age, but they had a lower socioeconomic status and a worse health status regarding the prevalence of stroke, disabilities, and poor self-rated health (20). Complete information on physical activity at baseline and cognitive functioning at follow-up was available for 347 subjects. The study has been approved by the medical ethical committee of the University of Leiden, the Netherlands. All participants had given their written informed consent.
Mini Mental State Examination.
Global cognitive functioning was tested with the Dutch version of the 30-point Mini Mental State Examination (MMSE). This examination includes questions on orientation to place and time, registration, attention and calculation, recall, language, and visual construction. The MMSE has been extensively used in epidemiological studies (30). In 1990, the MMSE took place in a hospital setting and was administered by two trained nurses. In 1993, the MMSE was administered at home by staff who were trained by the same instructors as in 1990. The MMSE consists of 20 items and yields a score (maximum of 30). A score of 25 or less is indicative for impaired cognitive functioning (15). Cognitive decline was defined as a decrease of more than three points in the MMSE score between 1990 and 1993 (more than one per year). This change would be significant at the 5% level for an individual, based on the standard error of measurement (1.96*SEM = 3.3 in this population). If individual items were missing, they were rated as errors with exception of those that could not be done due to physical disabilities, then a weighted total score was given (14).
Apolipoprotein E phenotype.
During the examination in 1990, blood samples were obtained and frozen at −20°C until determination of phenotype in 1993. For 19 subjects, samples from 1985 were used. The apolipoprotein E phenotype was determined by isoelectric focusing of either delipidated or guanidine-HC1-treated serum or plasma in a horizontal slab gel system followed by immunoblotting using either polyclonal or monoclonal anti-apolipoprotein E antibodies as first antibody. Apolipoprotein E phenotyping with this method in 200 serum samples that had been stored at −20°C for more than 1 yr gave exactly the same results as obtained with the conventional method based on isoelectric focusing of delipidated very low density lipoproteins isolated from fresh serum followed by protein staining. The validity of the use of stored serum for this purpose has been extensively described before (18). Because of a relatively small sample size, homozygotes (2.5%) and heterozygotes (20.7%) for the APOE*4 allele were considered together.
Physical activity.
Physical activity was assessed in 1990 with a self-administered questionnaire, originally designed for retired men by Professor J. N. Morris (London School of Hygiene and Tropical Medicine) and described by Caspersen et al. (5). A research assistant collected the questionnaires after administration and, if necessary, explained questions. This questionnaire is considered reliable and valid for measuring physical activity in elderly men, having demonstrated a substantial 4-month test-retest correlation (r = 0.93, P < 0.001) and having been validated by the doubly labeled water method (r = 0.61, _P_ < 0.01) in a subsample of Zutphen study participants (36). Questions were asked about the frequency and duration of walking and bicycling in the previous week, the average amount of time spent weekly on hobbies and gardening in both summer and winter; and the average amount of time spent monthly on odd jobs and sport. Time estimates were converted to minutes per week for each type of activity and summed to yield total weekly physical activity (1). When a man did not report engaging in a particular activity, it was coded as zero. Physical activity was categorized in three groups: ≤ 30 min·d-1, 31–60 min·d-1, and > 60 min·d-1.
Covariates.
Possible factors influencing the relationship between physical activity and cognitive functioning are age, education, smoking. and alcohol consumption. Other factors that may affect the relationship between physical activity and cognitive functioning are disability, depression, and self-rated health. Medical history of cardiovascular disease, particularly CVA, is probably an intermediate factor. During the examination in 1990, education, smoking, and alcohol consumption were assessed in a self-administered questionnaire. Education (years) was divided in tertiles: low (0–6 yr), medium (7–12 yr), and high (>12 yr). Smoking was stratified in current, ex-, and never smokers. Alcohol consumption was included as a continuous variate (g·wk-1). History of disease, i.e., myocardial infarction (MI), cerebrovascular accident (CVA), transient ischemic attack (TIA), angina pectoris (AP), type 2 diabetes mellitus, and hypertension, was dichotomized (yes/no). Disabilities in the “Activities of the Daily Life” (ADL) were based on self-report and defined as needing help with basic activities of daily living, mobility, and instrumental activities of daily living as described earlier (21). Self-rated health was defined by the answer to the following question “We would like to know what you think about your health. Please check what fits best in your case. Do you feel healthy, rather healthy, moderately healthy or not healthy?”
Statistical methods.
Mean (SD) baseline characteristics of the study population were calculated for the total group and separately for the physical activity categories. The differences in mean and prevalence of these characteristics between the activity groups were evaluated using Student’s _t_-test and chi-square, respectively. Logistic regression was used to calculate the risk of cognitive decline in the period 1990–1993 for subjects in the categories ≤ 30 min·d-1, 31–60 min·d-1, and > 60 min·d-1, using the last one as the reference category. In the models, we adjusted for age, education, smoking, alcohol consumption, and cognitive functioning in 1990. Two additional models were used: one in which also history of cardiovascular disease (CVA, TIA, MI, AP) and type 2 diabetes mellitus was included, and one in which disabilities in ADL and self-reported health was added. To investigate whether the association between physical activity and cognitive decline varied between carriers and noncarriers of APOE*4, the analyses were repeated separately for these groups. In these analyses, the physical activity categories ≤ 30 min·d-1 and 31–60 min·d-1’ were combined, because of small sample size. The relation between change in MMSE and change in physical activity was investigated using regression analyses with age, age, and baseline MMSE as covariates. SAS statistical programs were used for the analyses, version 6.12 (Statistical Analyses System, SAS Institute Inc. Cary, NC).
RESULTS
Our study population consisted of elderly men (mean age 74.6 yr), with a normal body mass index (mean 25.7 kg·m-2), moderate drinkers (mean 11.4 g alcohol·wk-1), and low prevalence of current smokers (19.5%). The prevalence of impaired cognitive functioning was significantly higher among inactive subjects as compared with those more active. In addition, physical activity was related to limitations in ADL, self-rated health, and age (P < 0.05) (Table 1).
Baseline characteristics (mean ± SD) of men in 1990 who survived until 1993; Zutphen Elderly Study.
Table 2 presents the OR of cognitive decline (MMSE score decrease > 3 points) within physical activity categories, adjusted for potential confounders. The table shows that, after adjustment for age, education, smoking, alcohol consumption, and cognitive status in 1990, subjects who were physically active 1 h or less per day had an OR of 2 for cognitive decline over a 3-yr period (nonsignificant) as compared with the rest. Additional adjustment for disabilities in ADL and self-reported health or history of CVA, TIA, MI, and AP resulted in similar odds ratios. Correction for disability in mobility and basic ADL, instead of correction for disability in total ADL, did not attenuate either relationship. The age- and education-adjusted OR of subjects active ≤ 1 h·d-1 was 2.0 (95% CI: 0.9–4.5) as compared with the rest (data not shown in table). Change in physical activity (h·wk-1) in the 3-yr period was significantly associated with change in MMSE in that same period. The beta was 0.05 h·wk-1 for 1 increase in MMSE, adjusted for age, and education (P = 0.05). Additional adjustment for baseline MMSE attenuated the beta to 0.04 (P = 0.06).
Odds ratio (95% confidence interval) of cognitive decline between 1990–1993 of men in categories of physical activity in 1990.
Table 3 and Figure 1 show the association between physical activity and cognitive decline according to presence of the APOE*4 allele. The table shows that an active lifestyle is particularly important among carriers of the APOE*4 allele. Although the risk for cognitive decline was similar among active and inactive noncarriers, the risk of cognitive decline among inactive carriers of the APOE*4 allele was nearly four times the risk of active carriers. These differences could not be explained by a larger variation in physical activity between APOE*4 carriers and noncarriers (Table 3). Additional adjustment of alcohol consumption, smoking and MMSE score in 1990 resulted in somewhat higher ORs with larger confidence intervals, but significance remained. Risk of cognitive decline is particularly confined to individuals with low activity and the APOE*4 allele (Fig. 1). Carriers of the APOE*4 allele who were ≤ 1 h·d-1 active in physical activity had a 13.7-fold increased risk of cognitive decline (95% CI: 4.2–45.5) as compared with noncarriers active for more than 1 h.
The association between physical activity and cognitive decline according to presence of e4 allele.
Adjusted odds ratios for cognitive decline; adjusted for age and education, MMSE 1990.
DISCUSSION
In this study, we observed a higher prevalence of impaired cognitive functioning in subjects with low physical activity as compared with those with higher activity. In the prospective analyses, an OR for cognitive decline (over the total range) of 2 was observed for individuals active maximal 1 h·d-1 as compared with the others (nonsignificant). This OR persisted after adjustment for age, education, MMSE score in 1990, smoking, and alcohol consumption. The relationship was also independent of cardiovascular diseases, disabilities in ADL, self-rated health, and depressive symptoms. Additional analyses showed that the observed associations were particularly strong in, and seemed to be confined to, carriers of the APOE*4 allele. Increased risk of cognitive decline among carriers of the APOE*4 allele who were inactive was more than could be expected from the separate effects of these risk factors combined.
A few issues of internal validity should be considered. First, unlike most previous studies, our study employed a prospective design, reducing the possibility that physical activity was a result rather than a cause of cognitive impairment.
Second, our results are based on a physical activity questionnaire specifically designed for elderly men. The questionnaire was validated against the doubly labeled water method (Pearson correlation = 0.61) (36) and shows expected congruence with physiological indicators (5). We have used a division in two groups because the sample size and the number of people with a cognitive decline were relatively low and more categories would lead to insufficient power. The cut-off points of 30 min·d-1 and 1 h·d-1 were chosen to simplify the public health message. Third, the MMSE is a reliable and valid indicator of cognitive impairment (34). It has been used extensively for clinical and epidemiological purposes, and it has been shown that it adequately reflects the distribution of cognitive function in a population. Although the MMSE was not created to assess changes over time, we have used the change in MMSE score as a measure of cognitive decline, as was previously done (37). The cut-off point of a decline of more than 3 points was significant on both an individual and population level and can be considered clinically meaningful.
Another point of consideration is the correction for possible confounders. Disabilities in the ADL and presence of depressive symptoms and cardiovascular diseases may be confounders as well as intermediate factors in the relation between physical activity and cognitive functioning. Adjustment for these factors did not markedly change the relation.
Previous studies have observed that nonresponse is related to cognitive status. Selective nonresponse can lead to bias in estimates of risk. We have compared physical activity and MMSE in subjects who responded to the 1993 survey and those who did not. Nonrespondents had a lower mean baseline MMSE score than the respondents (24.8 vs 26.4) and a lower mean physical activity level (438 min·wk-1 vs 588 min·wk-1 respectively). Although it is not clear to what extent this selective response has affected the association, we may assume that the presented association is more likely to be underestimated than overestimated.
Several other studies have demonstrated that physically fit older individuals process cognitive information more efficiently than less-fit individuals of the same chronological age. The strongest association is observed for tasks that require rapid or effortful cognitive processing rather than self-paced or automatic processing (6). Intervention studies have mainly been performed in young (17,35) or middle-aged subjects (9,10), and results are generally inconsistent (2,9,10). Moreover, not all of them were well controlled (2,10). Blumenthal and Madden (2) and Panton et al. (32) have both investigated the effect of training on cognitive function in elderly subjects. Despite substantial changes in aerobic capacity, no changes on cognitive parameters were observed. Hawkins et al. (19), on the other hand, did find a significant effect of 10-wk aerobic training on dual-task procedures but not on single-task procedures. Most of the intervention studies performed have a duration of 4 to 8 months. Although this period is mostly sufficient to observe changes in cardiovascular risk profile, it may be insufficient to observe changes in cognitive function.
The stronger association between physical activity and cognitive decline in carriers of the APOE*4 allele as described in our study has not been reported before. Our study indicates that genetic predisposition may be an important risk factor to consider in future research investigating the effect of physical activity on cognitive functioning.
There are several hypotheses that might explain an association of physical fitness to cognitive functioning. In a review (6), evidence in support of some mechanisms was summarized. It has been proposed that exercise maintains cerebrovascular integrity by enhancing oxygen transport to the brain. Impaired cerebral circulation may have adverse consequences for cognitive performance. From animal models, there is evidence that regular physical activity may promote neurotrophic changes, such as nerve cell regeneration or neurotransmitter repletion. Improvement in cognitive functioning by exercise could also be mediated through psychological factors (6). Physical activity may beneficially influence risk factors of CVD and type 2 diabetes mellitus, which are also associated with cognitive functioning. Finally, physical activity may prevent stress and its associated elevated cortisol levels, which, in turn, are associated with impaired cognitive functioning (23). Reasons why physical activity level was more strongly related to cognitive function in APOE*4 allele carriers are unclear. Possibly, physical activity only has an effect on cognition when the brain tissue is already altered by the APOE*4 allele. Other studies have reported an interaction between APOE*4 and cardiovascular disease or atherosclerosis in relation to cognitive functioning (24).
To conclude, our results suggest that elderly individuals who are relatively inactive have an increased risk of cognitive decline as compared with individuals who are active. The impact of low activity is particularly strong in individuals carrying the APOE*4 allele. Promoting physical activity at older age may therefore not only reduce the risk of cardiovascular disease or type 2 diabetes mellitus but also reduce the risk of cognitive decline particularly in genetically susceptible subjects. The existence of subgroups with a genetically considerable higher risk may have important implications for preventive intervention strategies.
The authors thank Dr Sandra Kalmijn for her contribution to the paper.
Address for correspondence: A. J. Schuit, Ph.D., Department of Chronic Diseases Epidemiology, National Institute of Public Health and the Environment, P.O. Box 1, 3720 BA Bilthoven, The Netherlands; E-mail: [email protected].
REFERENCES
1. Bijnen, F. C. H., E. J. M. Feskens, C. J. Caspersen, et al. Physical activity and cardiovascular risk factors among elderly men in Finland, Italy and the Netherlands. Am. J. Epidemiol. 143: 553–561, 1996.
2. Blumenthal, J. A., and D. J. Madden. Effects of aerobic exercise training, age, and physical fitness on memory-search performance. Psychol. Aging 3: 280–285, 1988.
3. Brayne, C., C. Gill, E. S. Paykel, F. Huppert, and D. W. O’Connor. Cognitive decline in an elderly population a two wave study of change. Psychol. Med. 25: 673–683, 1995.
4. Breteler, M. M., J. J. Claus, D. E. Grobbee, and A. Hofman. Cardiovascular disease and distribution of cognitive function in elderly people: the Rotterdam Study. Br. Med. J. 308: 1604–1608, 1994.
5. Caspersen, C. J., B. P. Bloemberg, W. H. M. Saris, R. Merritt, and D. Kromhout. The prevalence of selected physical activities and their relation with coronary heart disease risk factors in elderly men: the Zutphen Study. Am. J. Epidemiol. 133: 1078–1092, 1991.
6. Chodzko-Zajko, W. J., and K.A. Moore. Physical fitness and cognitive functioning in aging. Exerc. Sport Sci. Rev. 22: 195–220, 1994.
7. Christensen, H., A. Korten, A. F. Jorm, A. S. Henderson, R. Scott, and A. J. Mackinnon. Activity levels and cognitive functioning in an elderly community sample. Age Ageing 25: 72–80, 1996.
8. D uijn, C. M. van. Epidemiology of the dementias: recent developments and new approaches. J. Neurol. Neurosurg. Psychiatry 60:478–488, 1996.
9. Dustman, R. E., R. O. Ruhling, E. M. Russel, et al. Aerobic exercise training and improved neuropsychological function of older individuals. Neurobiol. Aging 5: 35–42, 1984.
10. Elsayed, M., A. H. Ismail, and R. J. Young. Intellectual differences of adult men related to age and physical fitness before and after an exercise program. J. Gerontol. 35: 383–387, 1980.
11. Farmer, M. E., S. J. Kittner, D. S. Rae, J. J. Bartho, and D. A. Regier. Education and change in cognitive function: the epidemiological catchment area study. Ann. Epidemiol. 5: 1–7, 1995.
12. F eskens, E. J. M., L. M. H avekes , S. K almijn , P. K nijff de , L. J. L auner , and D. K romhout . Apolipoprotein e4 allele and cognitive decline in elderly men. Br. Med. J. 309:1202–1206, 1994.
13. Feskens, E. J. M., M. P. Weijenberg, and D. Kromhout. A longitudinal study on diet, risk factors and cardiovascular disease in an aging cohort: the Zutphen Study. Neth. J. Cardiol. 4: 200–204, 1993.
14. Fillenbaum, G. G., L. K. George, and D. G. Blazer. Scoring non response on the Mini-Mental State Examination. Psychol. Med. 18: 1021–1025, 1988.
15. Folstein, M. F., S. E. Folstein, and P. R. McHugh. Mini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res. 12: 189–198, 1975.
16. Freidl, W., R. Schmidt, W. J. Stronegger, A. Irmler, B. Reinhart, and M. Koch. Mini mental state examination: influence of sociodemographic, environmental and behavioral factors and vascular risk factors. J. Clin. Epidemiol. 49: 73–78, 1996.
17. Office of the Surgeon General. Effect of training on reflex and reaction time. In:Research report DA-49–007-MD-889, D. Gibson, P. V. Karpovich, and P.D. Gollnick, 1961.
18. H avekes, L., P. K nijf de , U. B eisiegel , J. H avinga , M. S mit , and E. K lasen . A rapid micro-method for apolipoprotein E phenotyping directly in serum. J. Lipid Res. 28:445–463, 1987.
19. Hawkins, H., A. F. Kramer, and D. Capaldi. Aging, exercise and attention. Psychol. Aging 7: 643–653, 1992.
20. Hoeymans, N., E. J. Feskens, G. A. Van Den Bos, and D. Kromhout. Non-response bias in a study of cardiovascular diseases, functional status and self-rated health among elderly men. Age Ageing 27: 35–40, 1998
21. Hoeymans, N., E. J. M. Feskens, G. A. M. Van de Bos,and D. Kromhout. Measuring functional status: cross-sectional and longitudinal associations between performance and self-report (Zutphen Elderly Study 1990–1993). J. Clin. Epidemiol. 49: 1103–1110, 1996.
22. Kalmijn, S., E. J. M. Feskens, L. J. Launer, T. Stijnen, and D. Kromhout. Glucose intolerance, hyperinsulinaemia and cognitive function in a general population of elderly men. Diabetologia 38: 1096–1102, 1995.
23. Kalmijn, S., L. J. Launer, R. P. Stolk, et al. A prospective study on cortisol, dehydroepiandrosterone sulfate, and cognitive function in the elderly. J. Clin. Endocrinol. Metab. 83: 3487–92, 1998.
24. Kalmijn, S., E. J. M. Feskens, L. J. Launer, and D. Kromhout. Cerebrovascular disease, the apolipoprotein e4 allele and cognitive decline in a community-based study of elderly men. Stroke 27: 2230–2235, 1996.
25. K eys, A. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press, 1980, pp. 1–138.
26. Korten, A. E., A. S. Henderson, H. Christensen, et al. A prospective study of cognitive function in the elderly. Psychol. Med. 24: 919–930, 1997.
27. Kuusisto, J., K. Koivisto, L. Mykkänen, et al. Essential hypertension and cognitive function: the role of hyperinsulinemia. Hypertension 22: 771–779, 1993.
28. Launer, L. J., E. J. M. Feskens, S. Kalmijn, and D. Kromhout. Smoking, drinking and thinking: the Zutphen Elderly Study. Am. J. Epidemiol. 143: 219–227, 1996.
29. Launer, L. J., K. Masaki, H. Petrovich, D. Foley, and R. J. Havlink. The association between midlife blood pressure levels and data-life cognitive function. JAMA 274: 1846–1851, 1995.
30. Launer, L. J. Overview of incidence studies of dementia conducted in Europe. Neuroepidemiology 11 (Suppl. 1): 2–13, 1992.
31. Madden, D. J., J. A. Blumenthal, P. A. Allen, and C. F. Emery. Improving aerobic capacity for healthy older adults does not necessarily lead to improved cognitive performance. Psychol. Aging 4: 307–320, 1989.
32. Panton, L. B., J. E. Graves, M. L. Pollock, J. M. Hagberg, and W. Chen. Effect of aerobic resistance training on fractionated reaction time and speed of movement. J. Gerontol. 45: 26–31, 1990.
33. Powell, R. R. Effects of exercise on mental functioning. J. Sports Med. Phys. Fitness 15: 125–131, 1975.
34. Tombaugh, T. N., and N. J. McIntyre. The Mini-Mental State Examination: a comprehensive review. J. Am. Geriatr. Soc. 40: 922–935, 1992.
35. Tweit, A. H., P. D. Gollnick, and G. R. Hurn. Effects of a training program on total body reaction time of individuals of low fitness. Res. Q. 34: 508–513, 1963.
36. Westerterp, K. R., W. H. M. Saris, B. P. M. Bloemberg, K. Kempen, C. J. Caspersen, and D. Kromhout. Validation of the Zutphen Physical Activity Questionnaire for the Elderly with doubly labeled water. Med. Sci. Sports Exerc. 24: S68, 1992.
37. Yaffe, K., J. Cauley, L. Sands, and W. Browner. Apolipoprotein E phenotype and cognitive decline in a prospective study of elderly community women. Arch. Neurol. 53: 1110–1104, 1997.
Keywords:
AGED,; COGNITIVE DECLINE,; COHORT STUDY,; EXERCISE
© 2001 Lippincott Williams & Wilkins, Inc.