Abstract
BACKGROUND: The role of body mass index (BMI) as an effect modifier for readmission risk in people with HIV (PWH) remains underexplored. OBJECTIVE: To compare factors associated with two-year all-cause readmission in PWH, stratified by baseline BMI. METHODS: A retrospective matched (1:1 by age/sex) cohort study was conducted at a major public infectious diseases hospital in Shenzhen, China, serving a diverse urban population including migrants and low-income individuals. Adults admitted (Jan-Jun 2020) were stratified into underweight (BMI < 18.5 kg/m², n = 80) and normal/overweight (BMI 18.5-<28 kg/m², n = 231) groups. Multivariable logistic regression and ROC analysis were performed within each stratum. RESULTS: The readmission rate was higher in the underweight group (61.3% vs. 46.8%, p = 0.023). For underweight PWH, unemployment (adjusted Odds Ratio [aOR] = 8.11, 95% CI 1.82-36.16) and lower nadir CD4 + count (aOR = 0.994 per cell/µL decrease, 95% CI 0.989-0.999) were independent risk factors. For normal/overweight PWH, unemployment (aOR = 2.21, 95% CI 1.12-4.38) and longer time since diagnosis (> 5 vs. 0-1 years, aOR = 4.28, 95% CI 2.02-9.07) were risk factors, while being widowed/divorced was protective (aOR = 0.30, 95% CI 0.10-0.86). Predictive models showed good to excellent accuracy (AUCs: 0.820 and 0.748, respectively). CONCLUSIONS: Risk factors for readmission differ markedly by baseline BMI, confirming its role as an effect modifier. Interventions should be tailored: integrated nutritional and immunological support for underweight PWH, and chronic disease management with attention to social determinants for others.