Friend and Confidant Thresholds: Social Network Size as a Mediator Between Marital Status and Major Depressive Disorder

Le résumé fourni par la source

Background: Marriage rates are declining globally, a trend particularly notable in South Korea, where demographic shifts raise concerns about social well-being. While marital status is consistently linked to mental health, the extent to which social network size mediates the association between marital status and depression, particularly among contemporary Korean women, remains insufficiently understood. Methods: We analyzed 5443 Korean women aged 40 or older from the Korean Mood Disorder Genetic Study-Depression (KOMOGEN-D) study (2394 recurrent major depressive disorder [MDD]; 3049 controls). Marital status was dichotomized into married and unmarried (never married, divorced/separated, widowed). Social network size was operationalized as the count of friends and confidants. We used multivariable logistic regression and parallel mediation analysis to examine associations. Exploratory piecewise logistic regression identified network size thresholds for MDD risk reduction. Results: Unmarried women showed nearly four-fold higher odds of MDD (odds ratio [OR] 3.99), with the highest risk among divorced/separated women (OR 4.90). Increased social network size was associated with reduced MDD odds (per additional friend: OR 0.90; per additional confidant: OR 0.84, in a mutually adjusted model), and network size accounted for ~13% of this association. A protective baseline of approximately two friends and one confidant was observed in the overall sample; however, this threshold shifted upward for unmarried women (3-4 friends/confidants). Conclusions: Living without a spouse is strongly associated with MDD, partially due to smaller social networks. Maintaining a minimum core of supportive relationships is essential, suggesting that interventions to strengthen social networks are vital public health strategies for unmarried Korean women.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.