Retirement, Marital Status, and Activity Engagement Among Older Men and Women

Background and Purpose: Earlier studies have shown positive effect of activity engagement on physical and psychosocial well-being. However, significant life transitions such as retirement or change in marital status may entail great change in activity engagement. Only few studies examined the relationship between life transitions and activity engagement. Among these few existing studies, none has explored specific categories of retirement (e.g. complete retirement, partial retirement) and marital status (e.g., widowhood, divorce/separation, never married), which may have overlooked great variation among retired and unmarried individuals. The present study addresses following research question; “how does engagement in activities vary by different categories of retirement and marital status among older men and women?”

Methods: This study used 2011 Consumption Activities and Mail Survey (CAMS), a supplementary survey of the Health and Retirement Study. From 6,531 CAMS participants, we limited our sample to 5,405 by excluding who are younger than 50 years old, had proxy interview, had missing/irrelevant response for retirement item, or was cognitively impaired. Activity engagement was measured by time spent on each 33 activities and categorized into five domains based on the previous studies; leisure, physical, social, religious, household activities. Each measure was dichotomized (0=no participation, 1=any participation) and total score was derived for each domain. Self-report of retirement and marital status were both dummy coded; 1) working (referent), completely retired, and partly retired, 2) married/partnered (referent), divorced/separated, widowed, and never-married. Multivariate regression analyses were conducted on five domains of activities, separately for older men and women. Age, race, education, income, health, and depression were included as covariates.

Results: Distinguishable results were found across five domains of activities by retirement and marital status, indicating wide variation even among retired and non-married participants. When compared to the working group, partial retirees engaged in more types of activities than complete retirees (men: not only leisure and household but additionally engaging more in physical and social activities, women: not only leisure and social but additionally engaging more in religious activities). Divorced/separated and never-married men and women engaged less in social and religious activities than married persons. However, gender differences were shown for the widowed. Widowers engaged less in social activities while widows engaged more in physical activities compared to married counterparts.

Conclusions and Implications: Future studies specifically comparing completely and partly retired individuals and widowed, divorced/separated, and never-married individuals are needed to elucidate different findings among these groups. Nevertheless, the overall trend of positive relationship with activity engagement for retirees (vs. workers) or the married (vs. unmarried) suggests that retirement may still provide a chance to active lifestyle whereas the loss or absence of a spouse may serve as a barrier to it. Future studies using a longitudinal approach to assess risk and protective factors of activity engagement among these populations is recommended. Lastly, gender differences in findings also suggest further studies examining the moderating effect of gender, which may provide better understanding of life transitions and activity engagement relationship for older men and women in this field of research.