Receiver-operating characteristic of adiposity for metabolic syndrome: HANDLS Study
Beydoun MA, Fanelli Kuczmarski M, Wang Y, Mason M, Evans MK, Zonderman AB. Receiver-operating characteristic of adiposity for metabolic syndrome: HANDLS Study. Paper presented at the 42nd Annual Meeting of the Society for Epidemiologic Research; 23-26 June 2009; Anaheim, CA. American Journal of Epidemiology. 2009;169(suppl):S114.
BACKGROUND. Body fat percent (TtFM) or regional body fat may better predict the metabolic syndrome (MetS) and its component metabolic risk factors compared to common anthropometric indices. METHODS. Body mass index (BMI), waist circumference (WC), body composition by DXA and metabolic risk factors such as triglycerides (TA), HDL-cholesterol (HDL-C), systolic and diastolic blood pressures (SBP and DBP), fasting glucose (GLU), insulin resistance by homeostasis model assessment (HOMA-IR), uric acid and C-reactive protein (CRP) were measured in up to 933 adults from the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study. Receiver-operating characteristic (ROC) curves and logistic regression analyses were conducted. RESULTS. In predicting risk of MetS using obesity-independent NCEP criteria, percent body fat (TtFM) assessed using DXA measuring overall adiposity had no added value over WC. This was particularly true among women (Areas under curve; AUC1⁄40.586 vs. 0.712 for TtFM and WC, respectively, p< 0.05). Rib fat mass (RbFM) was superior to TtFM only in women for MetS (AUC1⁄40.771 and 0.586 for RbFM and TtFM, respectively, p< 0.05), and was superior to WC among African-American women in predicting MetS. Elevated Leg fat mass (LgFM) was protective against MetS. Among White women and African-American men, BMI was inferior to WC in predicting MetS. Optimal WC cut points varied across ethnic-gender groups and differed from those recommended by the NIH/NAASO. CONCLUSION. The present study provides evidence that WC is among the most powerful tools to predict MetS, and that optimal cut-points for various indices including WC may differ by sex and race.
Beydoun MA, Fanelli Kuczmarski M, Wang Y, Mason M, Evans MK, Zonderman AB. Receiver-operating characteristic of adiposity for metabolic syndrome: HANDLS Study. Paper presented at the 42nd Annual Meeting of the Society for Epidemiologic Research; 23-26 June 2009; Anaheim, CA. American Journal of Epidemiology. 2009;169(suppl):S114.
BACKGROUND. Body fat percent (TtFM) or regional body fat may better predict the metabolic syndrome (MetS) and its component metabolic risk factors compared to common anthropometric indices. METHODS. Body mass index (BMI), waist circumference (WC), body composition by DXA and metabolic risk factors such as triglycerides (TA), HDL-cholesterol (HDL-C), systolic and diastolic blood pressures (SBP and DBP), fasting glucose (GLU), insulin resistance by homeostasis model assessment (HOMA-IR), uric acid and C-reactive protein (CRP) were measured in up to 933 adults from the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study. Receiver-operating characteristic (ROC) curves and logistic regression analyses were conducted. RESULTS. In predicting risk of MetS using obesity-independent NCEP criteria, percent body fat (TtFM) assessed using DXA measuring overall adiposity had no added value over WC. This was particularly true among women (Areas under curve; AUC1⁄40.586 vs. 0.712 for TtFM and WC, respectively, p< 0.05). Rib fat mass (RbFM) was superior to TtFM only in women for MetS (AUC1⁄40.771 and 0.586 for RbFM and TtFM, respectively, p< 0.05), and was superior to WC among African-American women in predicting MetS. Elevated Leg fat mass (LgFM) was protective against MetS. Among White women and African-American men, BMI was inferior to WC in predicting MetS. Optimal WC cut points varied across ethnic-gender groups and differed from those recommended by the NIH/NAASO. CONCLUSION. The present study provides evidence that WC is among the most powerful tools to predict MetS, and that optimal cut-points for various indices including WC may differ by sex and race.
