Youngmin Cho
An estimated 19.3% of U.S. adults were socially isolated and 2.0% severely isolated. Social integration scores declined from 2006 to a nadir in 2010, with partial recovery by 2024. Young adults (18-34) showed the highest bivariate isolation prevalence (22.5%), but lower or comparable adjusted odds once marital status and living arrangement were accounted for. Male gender, unmarried status, lower income, and non-urban residence were independently associated with higher isolation odds, whereas Black and Hispanic respondents showed lower odds than White respondents. Religious participation declined steadily, accounting for approximately 60% of the overall decline in social integration.
BACKGROUND AND OBJECTIVES: Social isolation is a significant public health concern, yet trends across diverse demographic groups-including age-differentiated patterns across the adult life course-remain poorly characterized. This study examines the prevalence, temporal trends, and sociodemographic correlates of social isolation among U.S. adults of all ages.
RESEARCH DESIGN AND METHODS: Drawing on 10 waves of the General Social Survey (2006-2024; N = 28,269), this study constructs an adapted three-domain social isolation typology-living arrangement, social contact frequency, and religious attendance-benchmarked against the Berkman-Syme Social Network Index and Cornwell and Waite's (2009) framework. Weighted multinomial logistic regression and a counterfactual decomposition estimated sociodemographic correlates and the contribution of declining religious participation to the overall trend.
RESULTS: An estimated 19.3% of U.S. adults were socially isolated and 2.0% severely isolated. Social integration scores declined from 2006 to a nadir in 2010, with partial recovery by 2024. Young adults (18-34) showed the highest bivariate isolation prevalence (22.5%), but lower or comparable adjusted odds once marital status and living arrangement were accounted for. Male gender, unmarried status, lower income, and non-urban residence were independently associated with higher isolation odds, whereas Black and Hispanic respondents showed lower odds than White respondents. Religious participation declined steadily, accounting for approximately 60% of the overall decline in social integration.
DISCUSSION AND IMPLICATIONS: Spanning two decades-including the 2008-2009 financial crisis, the opioid epidemic, and the COVID-19 pandemic-this study extends the predominantly older-adult focus of prior research and identifies sociodemographic risk profiles for intervention, with implications for public health policy.