Abstract
BACKGROUND: Loneliness is a serious public health problem. This became even more visible during the COVID-19 pandemic. Yet, the key social network aspects contributing to loneliness remain unknown. Here, we evaluated social network structure and function and associations with (moderate/severe) social and emotional loneliness in older adults. METHODS: This cross-sectional study includes online questionnaire data (SaNAE cohort, August-November 2020), in independently living Dutch adults aged 40 years and older. For the separate outcomes of social and emotional loneliness, associations with structural social network aspects (e.g., network diversity - having various types of relationships, and density - network members who know each other), and functional social network aspects (informational, emotional, and practical social support) were assessed and risk estimates were adjusted for age, educational level, level or urbanization, comorbidities, and network size. Multivariable logistic regression analyses were stratified by sex. RESULTS: Of 3396 participants (55 % men; mean age 65 years), 18 % were socially lonely which was associated with a less diverse and less dense network, living alone, feeling less connected to friends, not having a club membership, and fewer emotional supporters (men only) or informational supporters (women only). 28 % were emotionally lonely, which was associated with being socially lonely, and more exclusively online (versus in-person) contacts (men only), and fewer emotional supporters (women only). CONCLUSION: Network structure and function beyond the mere number of contacts is key in loneliness. Public health strategies to prevent loneliness in older adults should be sex-tailored and promote network diversity and density, club membership, informational and emotional support, and in-person contact.