Age-dependent performance of obesity measures in screening for cardiometabolic risk over the lifespan

OBJECTIVES: To evaluate age-specific associations of body mass index (BMI) and waist-to-height ratio (WHtR) with cardiometabolic risk and to explore potential age- and sex-specific screening thresholds. STUDY DESIGN: This cross-sectional study included 58,712 participants from the NHANES in the United States from 1999 to 2020. METHODS: Using data from 58,712 participants (aged 12-85 years) from NHANES 1999-2020, we analyzed age-stratified associations between obesity metrics (BMI, WHtR) and cardiometabolic risk (hypertension, dyslipidemia, diabetes). We used subsample population weights to produce estimates representative of the U.S. POPULATION: Covariates included demographics, lifestyle, socioeconomic factors, and medication use. Logistic regression, receiver operating characteristic (ROC) curves, and population attributable fraction were used. Exploratory age- and sex-specific screening thresholds were estimated using Youden's index. RESULTS: ROC analysis suggested that the discriminative performance of both BMI and WHtR for identifying prevalent cardiometabolic risk decreased with increasing age. The association between adiposity measures and prevalent cardiometabolic risk appeared to attenuate with age, as reflected by lower odds ratios and population attributable fractions. Nonlinear, J-shaped cross-sectional associations of BMI and WHtR with cardiometabolic risk were observed, particularly among older adults, in whom the risk appeared to plateau at higher adiposity levels. The screening performance of conventional obesity thresholds varied across age groups. The exploratory age- and sex-specific thresholds based on the maximum Youden index showed relatively balanced sensitivity and specificity, but require prospective validation. CONCLUSIONS: BMI and WHtR exhibited nonlinear associations with cardiometabolic risk. The strength of these associations and their discriminative performance appeared to decrease with age, suggesting that age- and sex-specific screening thresholds may warrant further investigation.