Higher physical activity during middle age is associated with increased odds of survival without cognitive or physical impairments in older women
- Correspondence to Calvin Hirsch
Division of General Medicine, University of California, Davis Medical Center, 4150 V Street, PSSB Suite 2400, Sacramento, CA 95817, USA;
- Published Online First 8 June 2010
Sun and colleagues1 report on their analysis of data from the Nurses' Health Study, in which they evaluate the effects of a physically active lifestyle on a composite end point that is intended to model successful ageing. The authors define successful ageing as the absence of nine common age-associated medical conditions, cognitive impairment, disability and mental health limitations in women who reached their mid-70s. Prior studies have consistently shown a benefit of exercise in reducing the risk of the individual conditions that make up their successful-ageing construct. Strong evidence exists for an inverse relationship between level of physical activity and the development of hypertension, coronary heart disease, stroke, type 2 diabetes, depression, cancer of the lung or breast and functional decline.2,–,10 Moreover, a physically active lifestyle during middle age has been associated with a lower mortality risk in old age. In the Harvard Alumni Health Study, men (mean age 46 years) who exercised vigorously had a significantly lower risk of dying during the next 26 years than non-exercisers.11 Among women aged 50–74 years in the Framingham Heart Study, those in the upper two quartiles of physical activity were roughly two-thirds as likely to die in the following 16 years as women in the lowest activity quartile, after adjustment for comorbidity.12
The Nurses' Health Study, the Harvard Alumni Health Study and the Framingham Heart Study share the remarkable observation that a single measurement of physical activity can predict health outcomes many years later – 14 years in the case of the Nurses' Health Study. Does this mean that physical activity fosters physiological changes that confer long-lasting health benefits, or is it simply that a cross-sectional measurement of physical activity identified individuals with enduring exercise habits that provided ongoing health protection in the years that followed? The latter was most likely the case. In the Study of Osteoporotic Fractures, 7553 women (mean age 75 years) underwent activity assessment at baseline and 6 years later.13 Their vital status was tracked for up to 6.7 additional years. Compared with women who remained sedentary, those who became active had a hazard rate ratio (HRR) for mortality similar to that of the always-active, and those who became sedentary had a HRR that was statistically similar to that of the always-sedentary. These findings suggest that recent exercise behaviour trumps earlier exercise patterns in its influence on survival. Recent physical activity patterns also predict disability better than historical activity. In the Glastrup (Denmark) cohort study, neither physical inactivity at age 50 years nor cumulative physical inactivity from age 50 to 70 years was associated with disability at age 75 years; only inactivity at age 70 years predicted disability at age 75 years.14
Unlike randomised trials, cohort studies must control for baseline health by statistical adjustment. Because it is impossible to control for all potential confounders, there is a potential for bias, in that individuals who exercised more at baseline may have done so because they were healthier. Sun and colleagues attempted to reduce this bias by restricting their analysis to healthy women who, at baseline, were free of all the conditions subsumed under the composite end point. However, other unmeasured conditions, such as arthritis, could have affected the ability to exercise. Because the authors did not report the relative incidence of conditions that triggered women to transit from ‘successful’ to ‘usual’ ageing, we do not know how physical activity was protective in this cohort.
Sun and colleagues showed a clear dose–response relationship between overall exercise intensity and the odds ratio of successful ageing. They also reported that women who reported brisk walking (a moderate-intensity exercise) were 37–42% more likely to meet the criteria for successful ageing than sedentary women. Recommendations for moderate-intensity activity may be more realistic for currently inactive patients. Walking is the most common leisure-time physical activity among older adults. Studies have shown that women who walked the most could lower their risk of cardiovascular disease between 32% and 52% compared with women who walked the least.2 Healthcare providers should encourage the adoption and maintenance of a physically active lifestyle and should explore the reasons for their patients becoming less active.
Competing interests None.