Productive Engagement and Social Interaction as Predictors of Body Mass Index for Older Adults

Background and Purpose: The purpose of this longitudinal study was to investigate impact of productive engagement and social interaction as predictors of body mass index for older adults. Previous research has found that productive and social engagement predict lower depressive symptomatology, better physical functioning, and lower risk of mortality (Glass et al., 1999; Hinterlong, 2007), but has not examined body mass index as an outcome.

Methods: Longitudinal data (five waves, 1998 to 2006) from the Health and Retirement Study were analyzed. Individuals (N=16,789; 2,512 Blacks and 14,277 Whites) were selected for this study if they were non-Hispanic, Black or White race, interviewed in 1998 and at least 50 years old at baseline. This sample was 59% female and 65% married, with a mean age was 66.5 years (SD=10.17, range=50-109) at baseline. Body mass index (BMI) is the ratio of height and weight. Responses were recoded into a three-category outcome for underweight (<18.5), normal weight (18.5 to 24.9), and overweight/obese (>= 25). Physical activity was measured in terms of engaging in vigorous physical activity at least three times per week (yes/no). Productive engagement was measured as the number of productive roles occupied (0-4: paid employment, formal volunteering, informal volunteering, and parental caregiving). For social engagement, indicate the mean number of days per week of social contact with non-family members. As the BMI outcome distribution was multinomial, these longitudinal data were analyzed using hierarchical generalized linear modeling with full maximum likelihood estimation.

Results: At baseline, the majority of respondents were overweight/obese (BMI M=26.9, SD=5.2, range 11.7-74.5), engaged in one or more productive roles (M=1.26, SD=1.05), and had between one and two days of social contact per week (M=1.86, SD=2.43). In multilevel analyses, each additional productive role reduced the probability of being underweight as opposed to being normal weight (OR=.62, p<.01), while each additional day of social contact increased the odds of being underweight by 1.08 times (OR=1.08, p<.01) relative to normal weight, controlling for age, race, gender, and education. In contrast, each additional productive engagement role increased the odds of being overweight/obese by 1.18 times (p<.01), but frequency of social contact was not a significant predictor of overweight/obesity relative to those of normal weight. Conclusions and Implications: While previous research has found that productive engagement is a protective factor for health, this study found that engaging in multiple productive roles predicted increased odds of being underweight or overweight, relative to normal weight. It is possible that engagement in multiple productive roles reduces time for meal preparation and consumption, resulting in skipping meals or eating food of poor nutritional quality. The finding that frequent social contact increased the odds of being underweight but not overweight/obese was surprising since social gatherings often involve food, but perhaps individuals engaging in socialization were missing meals to do so. These findings can inform efforts by practitioners and clinicians working to promote healthy body mass index for older adults.