Abstract
Background and Aim: Sarcopenic obesity (SO) is a phenotype characterized by increased body fat combined with reduced muscle mass and strength. SO is prevalent among people living with HIV, especially in men (MLWH); however, the link between this phenotype and diet is still unclear in this population. For this reason, in this study, we aim to examine potential associations between self-reported macronutrient intake and SO in MLWH, and, eventually, to evaluate the diagnostic accuracy of a simple nutritional marker for screening SO. Methods: A total of 216 MLWH were selected from a large cohort who completed a total body composition measurement by dual-energy X-ray absorptiometry (DXA), muscle strength assessment by handgrip test, and nutritional recording by 24 h recall interview. The sample was categorized into SO (n = 45), non-SO (NSO) (n = 33), and non-sarcopenic non-obesity (NSNO) (n = 138). Logistic regression analysis was performed to determine the associations between different macronutrients and SO after adjusting for confounders. Receiver operating characteristic (ROC) curve analysis was used to identify discriminating cut-off points of the determined macronutrient intake to screen for SO. Results: The MLWH with SO while compared to NSO and NSNO, were of an older age and had a higher BMI, but with a lower total caloric and protein intake. However, adjusted logistic regression showed that only protein intake (g/kg/day) (OR = 0.017; 95%CI: 0.003-0.094, p < 0.05) and age (OR = 1.051; 95%CI: 1.011-1.093, p < 0.05) were significantly associated with SO. The age-adjusted ROC analysis identified the 0.98 g/kg/day of protein intake (AUC = 0.8149; p < 0.0001; sensitivity = 71%; specificity = 70%) as a cut-off point to screen for SO in the MLWH. Conclusions: We identified a new cut-off point of daily protein intake able to screen for SO in MLWH, and its use can be implemented in clinical settings.