Prediction of body fat content and distribution by anthropometry and biomarkers.

*Model (2) included all biomarkers, age, ethnicity, and key covariates, including smoking status (never vs. former, pack-years of cigarette smoking), education, use of medications (estrogen, statins, aspirin) and dietary supplements, and number of children. Model (3) shows the top 5 predictors from Model (2).Abbreviations: IGFBP1 (insulin-like growth factor binding protein 1); LAR (leptin to high-molecular-weight adiponectin ratio); PAI1 (plasminogen activator inhibitor-1); SHBG (sex hormone binding globulin); sLEPR (soluble leptin receptor).*模型(2)纳入了全部生物标志物、年龄、种族及关键协变量，涵盖吸烟状态（从不吸烟vs. 既往吸烟、卷烟吸烟包年数）、受教育程度、药物使用情况（雌激素、他汀类药物、阿司匹林）与膳食补充剂使用情况，以及生育子女数量。模型(3)展示了模型(2)中排名前5的预测因子。缩写说明：IGFBP1（胰岛素样生长因子结合蛋白1，insulin-like growth factor binding protein 1）；LAR（瘦素与高分子量脂联素比值，leptin to high-molecular-weight adiponectin ratio）；PAI1（纤溶酶原激活物抑制剂-1，plasminogen activator inhibitor-1）；SHBG（性激素结合球蛋白，sex hormone binding globulin）；sLEPR（可溶性瘦素受体，soluble leptin receptor）。