Demographic predictors of self-reported loneliness in adults.

The relative magnitude of individual demographic characteristics, compared with other demographic characteristics, in the prediction or explanation of frequency of loneliness has not been examined or reported in the literature. The relative strengths of a series of demographic variables (gender, age, marital status, household income, educational attainment, race or ethnicity, employment status, and occupation) in explaining frequency of loneliness in a random sample of 8,634 adults residing in a large metropolitan county were examined. Logistic regression analysis indicated several variables significantly affected group membership as lonely vs not lonely: marital status, household income, gender, and educational attainment. The strongest predictor of all was marital status. Age group as a predictor variable approached significance, but employment status, occupational, and race/ethnicity were not significant predictors of group membership as lonely vs not lonely in the logistic regression model.