Activity Profiles and Health in Later Life: Does Nature of Engagement Matter?

Background and Purpose. Activity is central to many aging paradigms of healthy aging, and established evidence has suggested involvement in activities as a modifiable strategy to promote healthy aging in later life. Previous studies have showed that activity is positively associated with physical, mental and cognitive health, while the underlying mechanism has not been fully investigated. Recent evidence has suggested that the nature of engagement may be important. The purpose of this study is to advance understanding about the role of the nature of engagement (i.e. physical, cognitive, and social demands of activity) in the relationship between activity and later-life health.

Methods. We analyzed data of 6,044 older adults from the 2010 and 2012 Health and Retirement Study (HRS) in conjunction with 2011 Consumption and Activity Mail Survey (CAMS). Confirmatory factor analysis was used to identify the factor structures of 33 activity items in the CAMS, and a 9-factor solution was obtained (CFI = 0.93; TLI = 0.91). These 9 factors were further analyzed through latent class analysis (LCA) to explore the latent activity profiles. Four measures capturing physical, cognitive, and social demands of activity in the CAMS were also analyzed via LCA to understand the nature of engagement. Health measures—self-rated health, depression, and cognitive function—were obtained from 2012 HRS. We used structural equation modeling approach to conceptualize a path model to examine the indirect effect of nature of engagement between activity profiles and health when the covariates and baseline health outcomes were controlled in the analyses. We used WLSMV estimator in conjunction with 20 multiple imputation data sets to handle categorical measures and missing values, respectively. The whole analyses were weighted to account for the complex survey design in the HRS.

Results. The LCA results showed that 5 distinct activity profiles (working, high activity, moderate activity, low activity, and passive leisure) could be found among older Americans. Further, three clusters of engagement—full engagement (high physical, cognitive and social demand), partial engagement (high physical and cognitive but low in social demand), and minimal engagement (low in every demand)—were revealed in LCA results. Using low activity and partial engagement as reference group, the path model results showed that all the activity profiles, except moderate activity, were associated with nature of engagement, and nature of engagement was associated with three health outcomes. Particularly, nature of engagement perfectly mediated the relationships between activity profiles and health.

Conclusions and Implications. The nature of engagement should be considered when examining the effects of activity on health. Further, this study finds that certain types of activity (i.e., working and high activity) could produce higher engagement, which in turn leads to better health outcomes in later life. Program or policy efforts that facilitate the involvement of older adults in types of activities with high physical, cognitive and social demands might be most effective in producing a healthy aging society. Further study should consider individual capacity as it relates to the demands of activity so that activity programming is maximally inclusive and health producing.