Food insecurity (FI) is a critical social determinant of poor psychosocial health. While data on the specific roles of sex and age in the FI-loneliness link among older adults are limited, the underlying mechanisms are largely unknown. This study examines the age-sex-specific associations of FI with loneliness among older adults in Ghana and quantifies the extent to which psychosomatic factors mediate the association.

We analyzed cross-sectional data from the Aging, Health, Psychological, and Health-seeking Behavior Study in Ghana. The past 30-day FI was assessed using items on hunger and breakfast skipping frequency due to a lack of resources. We assessed loneliness severity with the University of California, Los Angeles 3-item Loneliness Scale. Multivariable OLS regressions and bootstrapping mediation analysis using the Hayes PROCESS macro plug-in were used to evaluate the associations.

We included 1,201 individuals aged ≥50 years (mean = 62.9 [SD = 11.9]; women = 63.3%). The prevalence of loneliness was 17.7%. The prevalence of moderate and severe FI was 44.0% and 8.5%, respectively. In the adjusted model, greater FI was significantly associated with loneliness severity (B = .22, SE = .029, p < .001). We found significant interactive effects of FI × age (B = −.17, SE = .023, p < .01) and FI × sex (B = −.28, SE = .036, p < .001) on loneliness. Thus, the FI-loneliness link was respectively more marked among women (B = .25, SE = .035, p < .001) and ≥65 age groups (B = .34, SE = .041, p < .001) than men (B = .16, SE = .051, p < .01) and those aged 50−64 (B = .22; SE = .040, p < .001). Finally, comorbid depression/anxiety (41.07%), hopelessness (48.6%), worthlessness (42.1%), functional limitations (8.2%), and pain severity (6.4%) mediated the FI-loneliness association.

Age- and sex-specific associations between FI and loneliness exist among older Ghanaians. Addressing FI in concert with psychosomatic problems in older adults may contribute meaningfully to reducing loneliness in later life.