Loneliness as a Benchmark Social-connection Signal for Psychological Distress: Cross-national, Longitudinal and Genetically Informed Triangulation Evidence

Background: Loneliness is associated with psychological distress, but whether broader social measures add information is unclear. Aims: To assess incremental information from broader social measures beyond loneliness. Method: This study combined cross-sectional data from European Social Survey Round 6 (ESS6; 48,377 respondents, 29 countries), prospective Health and Retirement Study (HRS) intervals, and exploratory Mendelian randomisation. Results: In ESS6, loneliness had the largest survey-weighted standardised coefficient ( b = 0.221 per SD, 95% CI [0.195, 0.246]), 4.3 times the next-largest coefficient. Adding non-loneliness social variables to covariates plus loneliness yielded little cross-validation gain (delta AUC = 0.0058, corrected [0.0004, 0.0111]), whereas adding loneliness to covariates improved AUC by 0.0739 (0.0653, 0.0825). In HRS, baseline loneliness predicted next-wave depressive symptoms after adjustment for baseline symptoms and covariates ( b = 0.108, [0.096, 0.120]); relationship strain was the clearest adjacent signal. Exploratory MR suggested that the loneliness-depressive-symptom association may not be unidirectional, but pleiotropy, relaxed reverse instruments, sample overlap and mixed marker findings precluded directional causal interpretation. Conclusions: Loneliness may serve as a benchmark indicator for population research, while broader measures primarily contextualise relational strain and social safety.