Using longitudinal data to understand children's activity patterns in an exposure context: Data from the Kanawha county health study

An important component of assessing the levels, the sources, and the health effects of children's exposure to air pollution is understanding how and where members of this sensitive population spend their time. There are, however, few data bases that allow the documentation of the day-to-day nature of children's activities. Of particular concern is whether the one-day snapshots provided by time/activity diaries typically used in exposure studies represent the actual temporal and spatial extent of children's activities. As part of a community health study, longitudinal data on children's time/activity patterns were recently collected. A respiratory health status and gender stratified sample of 90 children kept daily diaries over two-week periods during both the summer and the fall. This paper first presents baseline information of children's activity patterns: the sample distribution of time spent in each of five microenvironments (travel, outdoor, at school, at home, and inside other locations) and the daily temporal pattern of activities. The consistent patterns of children on school days suggest that for most days we can accurately predict children's locations by time of day. The second part of the analysis shows that there is both high child-to-child variation in the average time spent in each microenvironment, even after controlling for gender and respiratory health status, and strong temporal variability in activity patterns within a child over time, even after controlling for school days versus nonschool days.