Comparison of the Predictive Values of CKD in Each Surrogate Maker of Obesity.

a Adjusted for age, smoking status, alcohol drinking behavior, regular exercise, hypertension, increased low-density lipoprotein cholesterol, decreased high-density lipoprotein cholesterol, hypertriglyceridemia, hyperuricemia, diabetes mellitus, urine occult blood, hemoglobin and preserved eGFR at baseline. Cox proportional-hazards models were used to calculate hazard ratios and 95% confidence intervals. Definitions of these confounding factors are shown in Table 1.
CI, confidence interval; Q1, lowest quartiles; Q2, second quartiles; Q3, third quartiles; Q4, highest quartiles.

本研究校正了基线时的年龄、吸烟状态、饮酒行为、规律运动情况、高血压、低密度脂蛋白胆固醇（low-density lipoprotein cholesterol）升高、高密度脂蛋白胆固醇（high-density lipoprotein cholesterol）降低、高甘油三酯血症（hypertriglyceridemia）、高尿酸血症（hyperuricemia）、糖尿病（diabetes mellitus）、尿潜血（urine occult blood）、血红蛋白水平及估算肾小球滤过率（estimated glomerular filtration rate, eGFR）正常情况。采用Cox比例风险模型（Cox proportional-hazards models）计算风险比及95%置信区间。上述混杂因素的定义详见表1。
CI：置信区间；Q1：最低四分位数组；Q2：第二四分位数组；Q3：第三四分位数组；Q4：最高四分位数组。