There has been considerable observational epidemiological data on the association between physical activity and cognition, but our confidence in the association (which may lead to future recommendations) can come only from interventional approaches. I would like to congratulate the authors for this significant effort that contributes to our knowledge expansion on the topic.

The exercise regimen that these middle-aged and elderly subjects followed is really impressive: 4 days/week, almost one hour per session at such HR levels for half a year constitutes a very intensive schedule. This is not very different than schedules used for marathon participation preparation and requires extremely high motivation and persistence (in particular for previously sedentary adults)! This and other exclusion criteria (in particular the absence of cardiovascular disease) limit the generalizability of the findings, but an intervention of high intensity is necessary in order to detect potential differences in such a short time. This is additionally demonstrated by the lack of an effect in 3...  Read more

There has been considerable observational epidemiological data on the association between physical activity and cognition, but our confidence in the association (which may lead to future recommendations) can come only from interventional approaches. I would like to congratulate the authors for this significant effort that contributes to our knowledge expansion on the topic.

The exercise regimen that these middle-aged and elderly subjects followed is really impressive: 4 days/week, almost one hour per session at such HR levels for half a year constitutes a very intensive schedule. This is not very different than schedules used for marathon participation preparation and requires extremely high motivation and persistence (in particular for previously sedentary adults)! This and other exclusion criteria (in particular the absence of cardiovascular disease) limit the generalizability of the findings, but an intervention of high intensity is necessary in order to detect potential differences in such a short time. This is additionally demonstrated by the lack of an effect in 3 months and for such a small sample.

The study included amnestic (only or plus) MCI subjects but, interestingly, no effect on memory tests (or Aβ) were noted. On the contrary, the associations were mostly noted for attention, speed, executive, set shifting, flexibility, more frontal types of tasks. Along these lines, and given the effects on the hypothalamic-pituitary axis, it may be important to consider differences in depression and dysthymia (mild depression.) Significant psychiatric diseases were excluded in this study.

Such types of studies cannot be double-blinded by design since the subjects know their intervention. Also, given the multiplicity of neuropsychological tests, clinical and biomarker information included, it may be important to know which of the above exceed a multiple comparison adjusted type I error threshold. I am not sure what to make of the reduced plasma BDNF in exercising women.

View all comments by Nikolaos Scarmeas

This is a very interesting study examining the effects of exercise on multiple variables, both cognitive and physiological. Although the sample sizes are small, the strict exclusion criteria allow for more control. The authors use several cognitive measures, which is very important for a more complete analysis of which memory functions are sensitive to the effects of aerobic exercise. The finding that women appear to benefit more from this kind of intervention is particularly interesting and similar to observations by Pietropaolo et al., 2008 in 3xTg-AD mice, where females showed improved cognitive performance after voluntary exercise, but males did not. Interactions among aerobic exercise, Alzheimer disease, and hormonal status in females will be important to dissect. It will be interesting in future studies, both human and animal, to examine hormonal status of females in these groups, as some may not be fully post-menopausal.

View all comments by Carla Yuede

This study presents cross-sectional associations between midlife and late-life physical activities and MCI diagnosis in a population-based study. The associations were significant for moderate-intensity exercise but not for light or vigorous exercise.

The study adds to an accumulating literature of the association between physical activity and cognition in the elderly. It is derived from a well-known population-based study and the methodology and presentation are exceptional.

The major limitation of the study is its cross-sectional character, which raises questions about causality-directionality of the associations. It is possible that MCI subjects just exercise less at present and hence report less moderate exercise in the past, too, which raises the possibility of reporting bias. It would be important to partially examine the validity of the responses using informant sources, too. Most likely the authors plan to perform longitudinal analyses in the near future (that will partially address the above issue). The absence of a dose-response effect, i.e. even lower MCI...  Read more

This study presents cross-sectional associations between midlife and late-life physical activities and MCI diagnosis in a population-based study. The associations were significant for moderate-intensity exercise but not for light or vigorous exercise.

The study adds to an accumulating literature of the association between physical activity and cognition in the elderly. It is derived from a well-known population-based study and the methodology and presentation are exceptional.

The major limitation of the study is its cross-sectional character, which raises questions about causality-directionality of the associations. It is possible that MCI subjects just exercise less at present and hence report less moderate exercise in the past, too, which raises the possibility of reporting bias. It would be important to partially examine the validity of the responses using informant sources, too. Most likely the authors plan to perform longitudinal analyses in the near future (that will partially address the above issue). The absence of a dose-response effect, i.e. even lower MCI rates in the vigorous physical activity group, is of some concern; its presence would further increase confidence in the finding.

View all comments by Nikolaos Scarmeas

This study correlates exercise with decreased odds of developing MCI, which suggests a protective effect of exercise, especially moderate exercise. The observation that moderate exercise performed at either mid- or late life decreased the odds of having MCI is similar to studies with animal models showing that exercise beginning late in life is still beneficial for improving/protecting cognitive function (Nichol et al., 2007). These data are important for showing that exercise can be a valuable intervention in the treatment of cognitive impairment even after symptoms have begun. It will also be important to determine the impact of exercise in combination with dietary factors, which may also have beneficial or detrimental influences.

View all comments by Carla Yuede