Mediating role of sleep in the association between HIV infection and depression: a cross-sectional analysis with NHANES data

Depression is a major public health concern and frequently coexists with sleep disturbances, particularly among individuals with Human Immunodeficiency Virus (HIV) infection. However, it remains unclear to what extent sleep disturbance may account for the observed association between HIV infection and depressive symptoms. This study examined the associations among HIV infection, sleep status, and depressive symptoms and assessed the potential contribution of sleep status. Using data from the National Health and Nutrition Examination Survey (NHANES) 2005-2014 cycles, weighted logistic regression, Firth's penalized regression, Bayesian mediation analysis, and LightGBM with SHAP interpretation were applied. Among 14444 adults, HIV-positive status was associated with higher odds of depressive symptoms, but the association was attenuated and became non-significant after adjustment for demographic, socioeconomic, and behavioral factors. Sleep quality showed a graded association with depression, with ORs of 2.17, 5.83, and 13.60 for fair, poor, and very poor sleep. Bayesian mediation analysis suggested that sleep disturbance may explain the association between HIV infection and depressive symptoms. SHAP analysis identified sleep status as the most important predictor of depression. These findings highlight the importance of sleep assessment and management for improving mental health among individuals with HIV infection.