Relative fat mass and an obstructive sleep apnea-related symptom phenotype among U.S. adults: a cross-sectional NHANES study

Background Relative fat mass (RFM) is calculated from height, waist circumference, and sex. Its fixed female term and within-sex monotone relationship with waist-to-height ratio (WHtR) require comparison with conventional anthropometry. Methods We analyzed adults aged ≥ 20 years in four National Health and Nutrition Examination Survey cycles (2005–2008 and 2015–2018). The phenotype required frequent snoring plus occasional or more frequent snorting, gasping, or stopped breathing; it was neither diagnosed obstructive sleep apnea (OSA) nor a validated screening score. Sex-stratified survey-weighted logistic models were primary. Sensitivities addressed thresholds, missing data, era, and restrictions. Apparent areas under the receiver operating characteristic curve (AUCs) and paired differences used 122 model-refitted stratified jackknife replicates and compared RFM with body mass index (BMI), waist circumference, WHtR, and unrestricted waist-plus-height models. Results Among 15,932 adults, weighted prevalence was 8.61%. In BMI-adjusted models, the odds ratio (OR) per RFM unit was 1.063 (95% confidence interval [CI] 1.006 to 1.124; p = 0.032) in women and 1.072 (1.008 to 1.141; p = 0.028) in men; per-sex standard-deviation ORs were 1.442 and 1.453. The secondary pooled OR was 1.067 (1.030 to 1.105; p < 0.001). Apparent AUCs were 0.527 for RFM, 0.670 for BMI, 0.693 for waist, 0.664 for WHtR, 0.627 for age + sex, 0.703 for age + sex+RFM, 0.704 for age + sex+BMI, and 0.708 for age + sex+waist. Adding RFM to age + sex+BMI changed AUC by 0.0024 (95% CI − 0.0008 to 0.0057; p = 0.143); adding it to age + sex+waist changed AUC by − 0.0004 (− 0.0015 to 0.0007; p = 0.447). The unrestricted age + sex+waist+height model exceeded age + sex+RFM by 0.0050 (0.0002 to 0.0098; p = 0.041). Conclusions Higher RFM was associated with the symptom phenotype, but pooled standalone discrimination was near chance and no incremental discrimination beyond BMI or waist circumference was detected. Within sex, RFM is a one-to-one WHtR transformation. These findings do not support replacing conventional anthropometry or validated OSA screening and diagnostic methods with RFM. Clinical trial registration Not applicable.