Adjusted Effect Measures for the Association of Participant Characteristics with Year One Blood Creatinine Concentration Using Multiple Linear Regression (N = 1,500).

Of N = 1,627 participants with year one creatinine available for analysis, 127 are not included due to missing data (see Table 1). CI, confidence interval.Characteristics chosen for inclusion by backward elimination using the Akaike Information criterion for model selection, with race, Hispanic ethnicity and randomized vitamin D treatment group forced into the model; characteristics are from baseline, except diuretic use (see 2); results for study centers not shown.1Characteristics are modeled as continuous variables with units as shown.2Diuretic use between baseline and year one blood draws.3Change in creatinine (mg/dL) associated with the exposure (mean, coefficient).Adjusted Effect Measures for the Association of Participant Characteristics with Year One Blood Creatinine Concentration Using Multiple Linear Regression (N = 1,500).本研究纳入可获取1年随访期肌酐检测数据的1627名受试者，其中127名因数据缺失被排除（详见表1）。 CI，即置信区间（confidence interval）。 本研究通过反向消元法结合赤池信息准则（Akaike Information Criterion）进行模型筛选以纳入特征变量，将种族、西班牙裔族裔及随机分配的维生素D干预组强制纳入模型；除利尿剂使用情况外，其余特征变量均取自基线数据（详见注2）；研究中心的分析结果未予展示。 1 特征变量以连续变量形式建模，单位如文中所示。 2 利尿剂使用情况指基线至1年采血时段内的用药情况。 3 暴露因素对应的肌酐变化量（单位：mg/dL，以均值、系数表示）。 基于多元线性回归分析的受试者特征与1年随访期血肌酐浓度关联的校正效应量（N=1500）。