Is Loneliness Adaptive? A Dynamic Panel Model Study of Older U.S. Adults

Abstract Objectives Recent evolutionary psychological theory proposes that loneliness is an adaptive mechanism, designed to trigger maintenance and repair of social ties. No population representative analyses have probed loneliness effects on sociality. The present study addressed this gap. Method Data were from the 2006, 2010, 2014, and 2018 waves of the Health and Retirement Study, nationally representative of U.S. adults over age 50. Recently developed cross-lagged models with fixed effects were used to test prospective within-person associations of loneliness with specific dimensions of sociality, taking into account reverse causality as well as all time-invariant confounders with stable effects. Both gender-combined and -specific analyses were conducted. Results Loneliness did not consistently predict overall sociality: sparse linkages were found only among women. The same null pattern held with family ties. Non-family ties, in contrast, were associated with prior loneliness, but in a gender-specific way. Loneliness positively predicted women’s social interactions with friends, but seemed linked to withdrawal from these relationships among men. There were indications that lonely men instead used religious attendance as a social outlet. Discussion Loneliness seems to induce domain- and gender-specific sociality responses. Findings suggest implications for evolutionary models of sociality as well as for psychosocial and physical health. Pending replication in independent samples, inferences remain tentative.