Mean difference in the levels of diabetes-related quantitative traits by genetic African ancestry after excluding participants receiving diabetes treatment.

BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); SES, socioeconomic status (including education, income and occupation).aTertiles 1, 2 and 3 of African ancestry are 87.3%, respectively, for hemoglobin A1c, and 87.3%, respectively, for glucose, insulin, and HOMA-IR.bEffects explained is defined as (β1−β2)/β1 where β1 is the regression coefficient of traits in Model 1; β2 is the regression coefficient after adjustment for covariates in each model.cModel 1: Mean difference (95% confidence interval) is adjusted for age, sex, and study.dModel 2: Mean difference (95% confidence interval) is adjusted for covariates in Model 1 and BMI.eModel 3: Mean difference (95% confidence interval) is adjusted for covariates in Model 1 and SES.fModel 4: Mean difference (95% confidence interval) is adjusted for covariates in Model 1, BMI, and SES.gPBMI，即身体质量指数（body mass index，计算公式为体重（千克）除以身高（米）的平方）；SES，即社会经济地位（socioeconomic status，涵盖教育、收入与职业）。a 非洲血统分组的第1、2、3三分位中，糖化血红蛋白（hemoglobin A1c）的对应占比均为87.3%，葡萄糖、胰岛素及稳态模型胰岛素抵抗指数（HOMA-IR）的对应占比亦均为87.3%。b 解释效应定义为(β₁−β₂)/β₁，其中β₁为模型1中性状的回归系数；β₂为各模型校正协变量后的回归系数。c 模型1：均值差（95%置信区间）已校正年龄、性别与研究因素。d 模型2：均值差（95%置信区间）已在模型1校正因素的基础上，额外校正身体质量指数（BMI）。e 模型3：均值差（95%置信区间）已在模型1校正因素的基础上，额外校正社会经济地位（SES）。f 模型4：均值差（95%置信区间）已在模型1校正因素的基础上，额外校正身体质量指数（BMI）与社会经济地位（SES）。g P值