摘要：Introduction: A commonly used approach to adjust for urine dilution in analyses of biomarkers is to adjust for urinary creatinine. However, creatinine is a product of muscle mass and is therefore associated with body mass. In studies of urinary analytes and obesity or obesity-related outcomes, controlling for creatinine could induce collider stratification bias. We illustrate this phenomenon with an analysis of urinary arsenic. Objective: We aimed to evaluate various approaches of adjustment for urinary dilution on the associations between urinary arsenic concentration and measures of obesity. Methods: Using data from the National Health and Nutrition Examination Survey, we regressed body mass index (BMI) and waist-to-height ratios on urinary arsenic concentrations. We compared eight approaches to account for urine dilution, including standardization by urinary creatinine, osmolality, and flow rates, and inclusion of these metrics as independent covariates. We also used a recently proposed method known as covariate-adjusted standardization. Results: Inverse associations between urinary arsenic concentration with BMI and waist-to-height ratio were observed when either creatinine or osmolality were used to standardize or as covariates. Not adjusting for dilution, standardizing or adjusting for urinary flow rate, and using covariate-adjusted standardization resulted in null associations observed between arsenic concentration in relation to BMI and waist-to-height ratio. Conclusions: Our findings suggest that arsenic exposure is not associated with obesity, and that urinary creatinine and osmolality may be colliders on the causal pathway from arsenic exposure to obesity, as common descendants of hydration and body composition. In studies of urinary biomarkers and obesity or obesity-related outcomes, alternative metrics such as urinary flow rate or analytic strategies such as covariate-adjusted standardization should be considered. https://doi.org/10.1289/EHP1202
