Association between relative fat mass and cardiometabolic multimorbidity in middle-aged and older adults and those with sarcopenia in two prospective cohorts

Obesity is a well-established risk factor for cardiometabolic diseases. While the Relative Fat Mass (RFM) index has been proposed as an alternative adiposity measure, its value for predicting cardiometabolic multimorbidity (CMM) in Chinese and American populations remains insufficiently evidenced. Sarcopenia poses a significant health burden to the aging Chinese population, but the predictive utility of RFM for CMM risk specifically in middle-aged and older adults with sarcopenia is currently unexplored. This study aimed to evaluate whether RFM is a superior predictor of CMM compared to Body Mass Index (BMI) and Waist-to-Height Ratio (WHtR), and to assess its specific predictive advantage over BMI within the Chinese sarcopenic population. This study utilized data from two prospective cohorts: China Health and Retirement Longitudinal Study (CHARLS) and Health and Retirement Study (HRS). RFM was calculated using height, waist circumference, and sex. CMM was defined as the co-occurrence of two or more conditions among diabetes, heart disease, and stroke. Associations between RFM and CMM were assessed using logistic regression for prevalence and Cox proportional hazards models for incidence. Model predictive performance was compared using Harrell’s C-index, NRI, and IDI. A total of 13,102 participants from CHARLS (male: 47.5%, median age: 59 years [IQR: 52.0–66.0]) and 8,470 from HRS (male: 42.1%, median age: 64 years [IQR: 56.0–74.0]) were included. The median follow-up periods was 9 years in the CHARLS and 14 years in the HRS. RFM (CHARLS, OR 1.68, 95% CI 1.19–2.40; HRS, OR 1.85, 95% CI 1.28–2.68), BMI (CHARLS, OR 1.17, 95% CI 1.01–1.34; HRS, OR 1.23, 95% CI 1.01–1.47), and WHtR (CHARLS, OR 1.25, 95% CI 1.06–1.48; HRS, OR 1.29, 95% CI 1.07–1.52) were all associated with CMM prevalence. A linear association was observed between RFM and CMM risk (CHARLS: p for nonlinearity = 0.770; HRS: p for nonlinearity = 0.744), whereas associations for BMI (BMI: CHARLS, p for nonlinearity < 0.001; HRS, p for nonlinearity = 0.003) and WHtR (CHARLS, p for nonlinearity = 0.014; HRS, p for nonlinearity = 0.001) were nonlinear. Regarding incidence, higher RFM (CHARLS, HR 2.22, 95% CI 1.84–2.69; HRS, HR 1.83, 95% CI 1.46–2.28), BMI (CHARLS, HR 1.25, 95% CI 1.17–1.32; HRS, HR 1.35, 95% CI 1.20–1.52), and WHtR (CHARLS, HR 1.43, 95% CI 1.31–1.56; HRS, HR 1.35, 95% CI 1.21–1.51) were associated with increased CMM risk. Adding RFM, BMI, or WHtR to a basic model improved its predictive performance (Harrell’s C-index and NRI). Notably, the IDI improvement was greatest for models incorporating RFM (CHARLS: IDI = 0.011, 95% CI 0.004–0.020; HRS: IDI = 0.013, 95% CI 0.002–0.028) compared to those adding BMI (CHARLS: IDI = 0.005, 95% CI 0.001–0.012; HRS: IDI = 0.006, 95% CI 0.001–0.011) or WHtR (CHARLS: IDI = 0.010, 95% CI 0.003–0.021; HRS: IDI = 0.009, 95% CI 0.001–0.016). Crucially, within the Chinese middle-aged and older adults with sarcopenia, RFM (HR 1.84, 95% CI 1.23–2.74) was significantly associated with CMM incidence, while BMI (HR 1.15, 95% CI 0.98–1.35) was not. RFM is a better predictor of CMM risk than BMI and WHtR in middle-aged and older Chinese and American adults. It demonstrates a particular advantage over BMI for CMM risk assessment in the Chinese population with sarcopenia.