Loneliness in the United States: A 2018 National Panel Survey of Demographic, Structural, Cognitive, and Behavioral Characteristics

Purpose: To inform health behavior intervention design, we sought to quantify loneliness and its correlates, including social media use, among adults in the United States. Design: Cross-sectional research panel questionnaire. Setting: Responses were gathered from individuals in all 50 states surveyed via Internet from February 2018 to March 2018. Participants: A total of 20 096 US panel respondents aged 18+. Measures: The University of California at Los Angeles (UCLA) Loneliness Scale (theoretical score range = 20-80) was administered along with demographic, structural, cognitive, and behavioral items. Analysis: After calibrating the sample to population norms, we conducted multivariable linear regression analysis. Results: The overall mean survey-weighted loneliness score was 44.03 (standard error = 0.09). Social support (standardized beta [s beta] = -0.19) and meaningful daily interactions (s beta = -0.14) had the strongest associations with lower loneliness, along with reporting good relationships, family life, physical and mental health, friendships, greater age, being in a couple, and balancing one's daily time. Social anxiety was most strongly associated with greater loneliness (s beta = +0.20), followed by self-reported social media overuse (s beta = +0.05) and daily use of text-based social media (s beta = +0.03). Conclusion: Our findings confirm that loneliness decreases with age, and that being in a relationship as well as everyday behavioral factors in people's control are most strongly related to loneliness. Population health promotion efforts to reduce loneliness should focus on improving social support, decreasing social anxiety, and promoting healthy daily behaviors.