In older adults, as in younger individuals, habitual moderate-intensity physical activity is associated with a reduced risk of various chronic health conditions, including certain types of cardiovascular and musculoskeletal disease and certain forms of cancer. However, the pattern of physical activity associated with such benefits remains unclear. One problem is that most investigators have examined patterns of physical activity using either subjective questionnaires or accelerometer or pedometer measurements limited to a single week, despite clear evidence of both the unreliability/invalidity of questionnaires and seasonal changes in activity patterns.
Since 2000, we have thus conducted an interdisciplinary study examining the habitual physical activity and health of elderly people living in a medium-sized Japanese town (the Nakanojo Study). In about one-tenth of some 5000 available subjects aged ≥65 years, physical activity has already been assessed continuously for 24 h/day for >8 years using a specially adapted pedometer/accelerometer. This device has a storage capacity of 36 days and can distinguish >10 intensities of physical activity (expressed in metabolic equivalents [METs]). Data have to date been summarized as daily step counts and daily durations of activity of <3 and >3 METs, averaged over a 1-year period. This article provides a detailed overview of both factors influencing habitual physical activity and relationships between such activity and health in an elderly population.
To date, analyses have been cross-sectional in type. Substantial associations have been noted between the overall health of participants and both the daily duration of effort undertaken at an intensity of >3 METs and the daily step count. In men, the extent of health is associated more closely with the daily duration of activity of >3 METs than with the daily step count, whereas in women, the association is closer for the step count than for the duration of activity >3 METs. In both sexes, the threshold amount of physical activity associated with better health is greater for physical than for mental benefits: >8000 versus >4000 steps/day and/or >20 versus >5 min/day at an intensity >3 METs, respectively. In other words, better physical health is seen in those spending at least 20 min/day in moderate walking (at a pace of around 1.4 m/s [5 km/h]) and a further >60 min of light activity per day. In contrast, better mental health is associated with much smaller amounts of deliberate physical activity.
The daily step count and the daily durations of activity of <3 and >3 METs are all influenced by meteorological factors, particularly precipitation and mean ambient temperature. Activity decreases exponentially to about 4000 steps/day as precipitation increases. Excluding the influence of rainfall, the daily step count peaks at a mean outdoor temperature of around 17°C; above and especially below such readings, physical activity decreases as a quadratic function of temperature. Seasonal changes in microclimate should thus be considered when designing interventions intended to increase the habitual physical activity of elderly people.
The observed associations between physical activity and health outcomes point to a need for longitudinal analyses; these should examine potential causal interpretations of the current findings and elucidate possible additional mediating variables.