Abstract
Background: Physical activity may be associated with better cognition.
Objective: To investigate whether change in duration and intensity of physical activity is associated with 10-year cognitive decline in elderly men.
Methods: Data of 295 healthy survivors, born between 1900 and 1920, from the Finland, Italy, and the Netherlands Elderly (FINE) Study were used. From 1990 onward, physical activity was measured with a validated questionnaire for retired men and cognitive functioning with the Mini-Mental State Examination (maximum score 30 points).
Results: The rates of cognitive decline did not differ among men with a high or low duration of activity at baseline. However, a decrease in activity duration of >60 min/day over 10 years resulted in a decline of 1.7 points (p < 0.0001). This decline was 2.6 times stronger than the decline of men who maintained their activity duration (p = 0.06). Men in the lowest intensity quartile at baseline had a 1.8 (p = 0.07) to 3.5 (p = 0.004) times stronger 10-year cognitive decline than those in the other quartiles. A decrease in intensity of physical activity of at least half a standard deviation was associated with a 3.6 times stronger decline than maintaining the level of intensity (p = 0.003).
Conclusions: Even in old age, participation in activities with at least a medium-low intensity may postpone cognitive decline. Moreover, a decrease in duration or intensity of physical activity results in a stronger cognitive decline than maintaining duration or intensity.
Physical activity has a beneficial effect on risk factors for cardiovascular diseases and thereby improves cerebral blood flow and reduces risk of stroke, which may subsequently diminish the risk of dementia and cognitive decline.1 In addition, physical activity may directly stimulate neurogenesis in the hippocampus, providing cognitive reserves against decline.2–4⇓⇓
Several cross-sectional and intervention studies and a few longitudinal studies have investigated the effect of physical activity on the risk of cognitive decline and dementia, but the results are inconsistent. Some studies reported that increased levels of physical activity are (moderately) associated with better cognitive functions.5–9⇓⇓⇓⇓ Other studies showed a weak or no relationship between physical activity and cognitive functioning.10,11⇓ The results of previous studies are difficult to compare owing to different and frequently global operationalizations of both physical activity and cognition. In most studies, detailed information about the duration and the intensity of physical activity was lacking. Therefore, it is not yet clear whether duration and intensity of physical activity are independently related to cognitive decline. Moreover, not only does cognitive functioning decrease with increasing age, physical activity behavior changes with age. Generally, participation in physical activities decreases, and activities will be performed at a lower pace at increasing age.
We sought to investigate the independent association of duration and intensity of physical activity at baseline with 10-year cognitive decline in elderly men. We also studied the effect of 10-year change in duration and intensity of physical activity on 10-year change in cognitive functioning. Data collected in the Finland, Italy, and the Netherlands Elderly (FINE) Study were used.
Materials and methods.
Study population.
The FINE Study consists of the surviving cohorts of the Seven Countries Study12: East Finland, West Finland, Zutphen, the Netherlands, Montegiorgio, Italy, and Crevalcore, Italy. This study started in 1984 among 2,285 men born between 1900 and 1920; follow-up examinations of the cohorts were carried out around 1990, 1995, and 2000. In 1990, 1,416 Finnish, Dutch, and Italian men were still alive and participated in the survey. Complete baseline information on cognitive functioning, physical activity, and possible confounders was available for 1,149 of these men in 1990 and 384 survivors in 2000. Of those survivors, 295 were classified as healthy (without myocardial infarction, stroke, diabetes, or cancer and with Mini-Mental State Examination [MMSE] score above 18 in 1990 [not severely cognitively impaired]).13,14⇓ We hypothesized that among healthy participants, the contribution of physical activity would be strongest because their cardiovascular system is generally fitter than that of participants with a chronic disease. For this reason, analyses were performed on healthy participants only. In 2000, information on cognition was available for all 295 healthy Dutch, Finnish, and Italian men. Information on physical activity was available in 2000 for Dutch and Italian men only. The data collected in the FINE Study population have been described in detail elsewhere.15
Approval of the medical ethics committee in the different countries was obtained for each participating center. Participants have given their written informed consent.
Assessment of physical activity.
Physical activity was assessed each survey round in the Netherlands and Italy and only in 1990 in Finland. A validated self-administered questionnaire, originally designed for retired men, was used.16 The questionnaire consisted of questions on the frequency, duration, and pace of walking and bicycling during 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 sports. The kinds of odd jobs, sports, and hobbies were assessed as well. In both the Finnish and the Italian questionnaire, questions about the average amount of time spent weekly on farming or forestry (for Finland only) in both summer and winter were added. Estimated times were converted into minutes per week. All types of activity of an intensity of >2 kcal/kg-h (e.g., fishing and billiards), reflecting multiples of resting oxygen consumption, were summed to obtain total weekly duration of physical activities.15 Participants with more than two activities missing were excluded from the analyses (n = 53). If one or two activities were missing, the duration and intensity of this activity were assumed to be zero. Total daily duration of physical activity in 1990 and 2000 was calculated and categorized into four groups: ≤30, 31 to 60, 61 to 120, and >120 min/day.
A mean intensity score for each participant was calculated as follows: For each person, the intensity code of each activity reported was multiplied by its duration; these values were summed and divided by the total time spent by that individual on all activities to yield a mean intensity score. Mean intensity scores were classified in quartiles based on the mean intensity in 1990. Activities in the lowest quartile are, for example, playing billiards and walking at lower pace than 3 mi/h (converted to the Ainsworth compendium of physical activities).17 Activities in the second quartile are playing volleyball and walking at about 3 mi/h. Activities in the third quartile are, for example, gymnastics and walking at about 3.3 mi/h. Activities in the highest quartile are swimming and walking at >3.5 mi/h.
Change in duration of physical activities between 1990 and 2000 was categorized as follows: Increased duration was defined as an increase of >60 min/day (half a standard deviation), decreased duration equaled a decrease of >60 min/day, and stable duration was defined as no change or a maximal change of 60 min/day.
Change in mean intensity of physical activity between 1990 and 2000 was divided into the following categories: Increased intensity was defined as an increase in intensity of >0.8 point (half a standard deviation, which corresponds, e.g., to the difference between playing volleyball and swimming or with a change in walking velocity of 0.5 mi/h), decreased intensity was classified as a decrease in intensity of >0.8 point, and stable intensity was defined as no change or a maximal change of 0.8 point.
Assessment of cognitive function.
The MMSE was used to assess cognitive function and includes questions on orientation to time and place, registration, attention and calculation, recall, language, and visual construction.18 Originally, this screening test was created for clinical use, but now it is extensively used in epidemiologic studies, has proven to be a reliable and valid indicator of cognitive impairment, and has a good test–retest reliability.13,19,20⇓⇓
The maximum score on the MMSE is 30 points; a higher score indicates better cognitive performance. If a subject did not answer ≥4 individual items (of a total of 20), the total MMSE score was considered missing (n = 4). If less than four items were missing, missing items were rated as errors, and a total MMSE score was still calculated.21
Assessment of other variables.
Demographic, lifestyle, and other information was obtained with standardized questionnaires. Education was assessed as the number of years of education. Smoking status was categorized into never, former, and current smoking. Alcohol consumption and participation in mental activities (like puzzling) were both categorized into yes/no. Height and weight were measured while men were wearing light clothing and no shoes. Body mass index (BMI) was calculated by dividing weight by the square of height (kg/m2). Information about the history of myocardial infarction, stroke, diabetes, and cancer was collected by standardized questionnaires and validated by hospital registries. Functional status was registered with a self-administered questionnaire, which measures activities of daily living (ADL),22 and depressive symptoms were assessed with a self-rating depression scale.23 Information on the use of antihypertensive drugs was collected by a standardized questionnaire, and serum high-density lipoproteins (HDL) cholesterol, serum total cholesterol, and blood pressure were determined in standardized laboratories.15
Statistical analyses.
Differences in baseline characteristics between countries were tested using Kruskal–Wallis tests for not normally distributed continuous variables. Categorical data were tested for differences with χ2 tests. Adjusted differences in baseline cognitive functioning were tested with multivariate linear regression analysis.
The association between physical activity duration and intensity at baseline and 10-year cognitive decline was assessed with multivariate linear regression analysis. The association between 10-year change in physical activity duration and intensity with 10-year cognitive decline was assessed with multivariate linear regression analysis in the Netherlands and Italy only. Categories of (change in) duration and intensity of physical activity were entered as class variables in the model and the outcome cognitive decline (MMSE 2000 − MMSE 1990) as a continuous outcome variable. We adjusted for the following confounders: age, education, country, alcohol consumption, smoking status, mental activities, and alternately for physical activity intensity or duration. Additional adjustments were made for ADL, depression, BMI, use of antihypertensive drugs, HDL, total cholesterol, blood pressure, and baseline cognitive functioning. We also examined whether an interaction was present between duration and intensity of physical activity by adding the product term of these variables to the model.
All statistical analyses were carried out using SAS software (version 8.2; SAS Institute, Cary, NC). A two-sided p value of ≤0.05 was considered to be significant.
Results.
Healthy Italian survivors were older than the Finnish and Dutch survivors (table). The number of years spent on education as well as cognitive functioning were highest for the Dutch men. Cognitive functioning decreased between 1990 and 2000 for all men. In 1990, Italian men were most physically active. Finnish participants spent most of their time on walking, Dutch men on bicycling, and Italian men on gardening. Mean duration of physical activity decreased for both Dutch and Italian men between 1990 and 2000. Mean intensity spent on physical activities did not differ between the countries and decreased between 1990 and 2000. The percentage of men who consumed alcohol or who smoked cigarettes decreased between 1990 and 2000 as well. Average BMI was stable.
Table Characteristics in 1990 of the 295 healthy participating Finnish, Dutch, and Italian male survivors (without myocardial infarction, stroke, diabetes, or cancer and with MMSE above 18), born between 1900 and 1920
Men who did not participate in our study (n = 854) were men who did not survive till 2000, who gave incomplete answers, or who were unhealthy in 1990. These men were overall older (p < 0.001), had worse cognitive functioning in 1990 (p < 0.001), spent less time on physical activities per day (p < 0.001), performed activities with a lower intensity (p < 0.001), drank less alcohol (p < 0.01), and more often smoked (p < 0.05) than the men who did participate in our study.
Baseline duration of physical activity.
Among 295 healthy Finnish, Dutch, and Italian survivors, cognitive functioning in 1990 did not differ among the four baseline duration of activity groups, after adjustment for intensity and confounders (figure 1A). In addition, the rates of the 10-year cognitive decline of the activity groups were also not significantly different (see figure 1A).
Figure 1. (A) Ten-year cognitive decline per category of baseline duration of physical activity for healthy Finnish, Dutch, and Italian survivors. Mean change in cognitive functioning between 1990 and 2000 (95% CI), adjusted for age, education, smoking status, alcohol consumption, country, mental activities, and intensity. aReference group; bnot different from reference group (p ≥ 0.3). MMSE = Mini-Mental State Examination; (dot–dash line/squares) = ≤30 min/day; (solid line/triangles) = 31 to 60 min/day; (dotted line/diamonds) = 61 to 120 min/day; (dashed line/circles) = >120 min/day. (B) Ten-year cognitive decline per quartile of baseline intensity of physical activity for healthy Finnish, Dutch, and Italian survivors. Mean change in cognitive functioning between 1990 and 2000 (95% CI), adjusted for age, education, smoking status, alcohol consumption, country, mental activities, and duration. aReference group; bborderline different from reference group (p = 0.07); cdifferent from reference group (p = 0.02); ddifferent from reference group (p = 0.004). (Dot–dash line/squares) = lowest quartile; (solid line/triangles) = second quartile; (dotted line/diamonds) = third quartile; (dashed line/circles) = highest quartile.
Baseline intensity of physical activity.
There were no differences in baseline cognitive functioning among the four quartiles for baseline intensity of activity in 1990, adjusted for duration and confounders (see figure 1B). Men in the lowest intensity quartile at baseline showed the strongest 10-year cognitive decline of 2.7 points. This decline was 1.8 (p = 0.07) to 3.5 (p = 0.004) times stronger than the decline among the other quartiles (see figure 1B).
Ten-year change in duration of physical activity.
Among the 243 healthy Dutch and Italian survivors for whom information was present on change in physical activity, there were no differences in MMSE scores in 1990 among the groups for change in activity duration after adjustment for intensity and confounders (figure 2A). However, cognitive decline was strongest (1.7 points) for those whose duration of activity decreased >60 min/day (half a standard deviation) during 10 years. This decline was 2.6 times stronger than for those whose duration of activity remained stable (p = 0.06) (see figure 2A). There was no cognitive decline among men who increased their duration of activity. The relationship between change in duration of physical activity and cognitive decline showed a linear trend (p = 0.02).
Figure 2. (A) Ten-year cognitive decline per category of 10-year change in duration of physical activity for healthy Dutch and Italian survivors. Mean change in cognitive functioning between 1990 and 2000 (95% CI), adjusted for age, education, smoking status, alcohol consumption, country, mental activities, intensity, and baseline duration. aReference group; bborderline different from reference group (p = 0.06); cdifferent from reference group (p = 0.05). MMSE = Mini-Mental State Examination; (dot–dash line/squares) = increased duration; (solid line/triangles) = stable duration; (dotted line/diamonds) = decreased duration. (B) Ten-year cognitive decline per category of 10-year change in intensity of physical activity for healthy Dutch and Italian survivors. Mean change in cognitive functioning between 1990 and 2000 (95% CI), adjusted for age, education, smoking status, alcohol consumption, country, mental activities, duration, and baseline intensity. aReference group; bdifferent from reference group (p = 0.003); cdifferent from reference group (p = 0.01). (Dot–dash line/squares) = increased intensity; (solid line/triangles) = stable intensity; (dotted line/diamonds) = decreased intensity.
Ten-year change in intensity of physical activity.
Cognitive functioning in 1990 did not differ among the change in activity intensity groups, after adjustment for duration and confounders (see figure 2B). However, men whose intensity of activity decreased more than half a standard deviation (0.8 point) during 10 years had the strongest cognitive decline of 2.3 points (see figure 2B). This decline was 3.6 times stronger than the decline of men whose intensity remained stable (p = 0.003). There was no cognitive decline in the increased intensity group. A linear trend (p = 0.002) was present for the relation between change in intensity and cognitive decline.
Additional adjustments for baseline MMSE, possible intermediates like ADL, depression, BMI, use of antihypertensive drugs, HDL, total cholesterol, and blood pressure did not influence the results (results not shown). Also, there was no interaction between duration and intensity in relation to cognitive decline (p = 0.8) (results not shown).
Discussion.
This study showed that in old age, physical activities of at least medium-low intensity at baseline were associated with less cognitive decline than physical activities of lowest intensity. Duration of activity at baseline was not associated with the rate of cognitive decline. However, maintaining or increasing the duration or intensity of physical activity over 10 years resulted in a smaller cognitive decline than decreasing the duration or intensity.
Before interpreting the results, several methodologic issues should be addressed. The advantages of this study are its longitudinal design, a long follow-up period, the ability to adjust for a large number of potential confounders, and the possibility of analyzing duration and intensity of physical activity separately and independently in relation to cognitive decline. The longitudinal design of the FINE Study makes it less likely that reduced physical activity was a consequence instead of a cause of cognitive decline. Also, the postulated underlying mechanisms are in favor of a causal relationship. However, in the analyses on change in physical activity and cognitive decline, the direction of the association is less obvious, and one cannot draw firm conclusions regarding causality.
A disadvantage of this study is the small number of healthy participants. However, despite this small number, we found consistent and significant results. The 854 nonparticipants in our study were less active and had lower cognitive test scores. Therefore, exclusion of these men probably led to an underestimation of the strength of the relationship between physical activity and cognition. We used the MMSE to assess cognitive decline; although it is a reliable and valid indicator of cognitive impairment and has a good test–retest reliability, it is just a screening test.13,19⇓ Future studies should therefore include a more extensive cognitive test battery.
We did not observe differences in the rates of cognitive decline among the baseline duration categories. Possibly, all participating elderly already achieved a minimal duration needed for an effect on cognitive decline. In addition, we found that men who performed activities with lowest intensity had the strongest cognitive decline. The rates of cognitive decline among the three other (higher) intensity quartiles did not differ significantly from each other. An activity with at least a medium-low intensity (like playing volleyball and walking at about 3 mi/h) showed already a borderline significantly smaller cognitive decline. The benefit of the positive effect of an activity with a medium-low intensity is that participation in these kinds of activities will be easier and requires less effort than a more vigorous activity and will therefore be easier to implement in our society.
Precise mechanisms by which physical activity may be beneficial for cognition are unknown. The following mechanisms may be involved.24,25⇓ First, physical activity may have a beneficial effect on cardiovascular risk factors, such as a reduced risk of hypertension and arteriosclerosis, and may thereby maintain cardiovascular fitness. This stimulates cerebral circulation, resulting in an increased oxygen transport to the brain.1 In addition, the risk of stroke and white matter lesions will be reduced.26 These effects will eventually lead to less cognitive decline.
Second, experimental animal research shows that being physically active directly stimulates trophic factors and neuronal networks, which leads to neurogenesis in the hippocampus and regulation of synaptic plasticity and neurotransmitter synthesis, possibly providing cognitive reserves against cognitive decline and dementia.2–4⇓⇓ Another possible mechanism could be that physical activity increases the levels of brain serotonin, thereby reducing stress and stress-induced hypercortisolemia and consequently stimulating regeneration of neurons within the hippocampus.24
Finally, physically active participants may have a healthier lifestyle, such as more adequate nutrition, which may reduce the risk of cardiovascular diseases and consequently of cognitive decline.27 It may also be possible that the mental aspect needed to perform a certain level of physical activity or a social aspect, like walking with a friend, stimulates cognition. Indeed, frequent participation in cognitively stimulating activities was found to be associated with a reduced risk of cognitive decline.28,29⇓
Several studies have found an association between physical activity and cognitive functioning, although they did not investigate duration and intensity of physical activities independently of each other.6,8,9⇓⇓ For example, a prospective study in the United States showed that women who walked more blocks a week and who spent more energy per week were less likely to develop cognitive decline.9 In the Canadian Study of Health and Aging, it was shown that regular physical activity (combination of frequency and intensity of an activity) had a significant protective effect on the risk of cognitive decline in the elderly.8 However, other studies found a weak or no association and did not investigate the independent effect of duration and intensity of physical activity on cognitive decline.10,11⇓
For developing specific treatment programs and recommendations for healthy aging, it is necessary to know which aspects of physical activity are associated with cognition. The results of this study suggest that stimulating elderly to be physically active with at least a medium-low intensity or becoming even more physically active (in duration or intensity) could be important for keeping their brains fit.
Acknowledgments
The FINE Study is a part of the HALE (Healthy Ageing: Longitudinal Study in Europe) Project and supported by a grant from the European Union (QLK6-CT-2000-00211) (D.K.).
The authors thank Dr. Jantine Schuit for sharing her expertise on physical activity.
- Received February 10, 2004.
- Accepted August 2, 2004.
References
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.