Methods: Data came from the Family Life, Activity, Sun, Health, and Eating Study, a cross-sectional internet study conducted on behalf of the National Cancer Institute in 2014. Parents and their teenage children (ages 12-17), recruited through consumer opinion panels, completed surveys about demographics, physical activity, and diet. Current analyses include all dyads in which at least one member completed demographic information and/or the physical activity survey (N = 1,851). Actor Partner Interdependence Models with the MLR estimator in Mplus were used to determine whether parent and teenager social isolation (a scale consisting of 2 Likert items given to each person, α = .91 for both parents and teenagers) were associated with self-reported health (1 Likert item) for each dyad member. Analyses tested whether effects differed by dyad type (mother–daughter, mother–son, father–daughter, father–son). Models included household income, parent ethnicity, parent marital status, and age of parent and teenager as covariates. Missing data were estimated via the robust maximum likelihood estimator.
Results: Most dyads included a mother (38% mother–daughter and 36% mother–son). Most parents were non-Hispanic White (66%), were married/partnered (77%), and reported household income below $100,000 (79%). Mean teenager age was 14.5 (SD = 1.6), and most parents were either 35-44 (44%) or 45-59 (42%). Both social isolation and self-reported health were significantly correlated between parents and their teenagers (intraclass correlation coefficient = .38 for isolation, .21 for health). Because a series of difference tests found no difference in effects between dyad types (mother–daughter, father–son, etc.) or between parent ethnicity (non-Hispanic White versus other ethnicities) (ps > .07), a single-group model was used for the entire sample. Actor effects were found for a negative association between parent isolation and parent health (p < .001), as well as teenager isolation and teenager health (p < .001). Partner effects were found for a negative association between parent isolation and teenager health (p = .022), but not for teenager isolation and parent health (p = .276).
Conclusions and implications: Self-reported social isolation on the part of parents—whether mothers or fathers—was significantly associated with lower self-reported health not only for themselves but also for their teenagers. Interventions addressing parents' social isolation may improve not only their own health but the health of other family members. Longitudinal measurement will be essential to fully model direct and indirect effects of parent social isolation on the social integration and physical health of their teenage children.