Objective To determine whether social contacts and spousal characteristics predict incident instrumental or basic activities of daily living (I/ADL) limitations and whether effects differ for individuals with high risk of dementia. Design Cohort study. Setting Biennial interviews of Health and Retirement Study participants over up to 12 years. Participants 4,125 participants aged 65 years and older without baseline I/ADL limitations. Measurements Participants' family characteristics (living arrangements, proximity to children, contacts with friends, marital status, and spouse's depression, employment, and education) and dementia probability (high versus low risk of dementia based on direct and proxy cognitive assessments) were characterized at baseline. Family characteristics and their interactions with dementia probability were used to predict incident I/ADL limitations in pooled logistic regressions. Results ADL limitation incidence was higher among the unmarried (odds ratio [OR] versus married: 1.14; 95% CI: 1.01-1.30); those married to a depressed spouse (OR versus nondepressed spouse: 1.56, 95% CI: 1.21-2.00); or whose spouse had less than high school education (OR versus spouse with high school or more: 1.29, 95% CI: 1.06-1.57). Living with someone other than a spouse compared with living with a spouse predicted higher risk of both incident ADL (OR: 1.35; 95% CI: 1.11-1.65), and IADL (OR: 1.30; 95% CI: 1.06-1.61) limitations. Effects were similar for respondents with high and low dementia probability. Conclusions Regardless of dementia risk, older adults may receive important marriage benefits, which help delay disability. The salience of spouse's education and depression status implicate modifiable mechanisms, such as information and instrumental support, which may be amenable to interventions.
- cognitive function
- social ties