Comparing creatinine and osmolality for urine dilution adjustment in a case-cohort study of urine cadmium and risk of incident heart failure in the Danish Diet, Cancer, and Health Cohort

Abstract

Background/Aim: Traditionally, analyses using urinary biomarkers adjusted for urine creatinine to account for differences in urine dilution. Urine creatinine, however, is associated with muscle mass, dietary intake of creatine, and meat intake; therefore consideration of other measures for urine dilution adjustment may be critical. We measured osmolality, in addition to creatinine, and compared epidemiologic associations of urine cadmium with heart failure across several strategies of urine dilution adjustment.Methods: In our case-cohort study of never smokers from the Danish Diet, Cancer, and Health cohort, we identified 958 cases of incident heart failure occurring between baseline (1993-1997) and 2015 using the Danish National Patient Registry and we randomly selected a sub-cohort of 600 males and 600 females for comparison. We quantified cadmium concentrations, creatinine, and osmolality in urine samples collected at baseline. Using an unweighted case-cohort approach we estimated adjusted hazard ratios for heart failure in Cox proportional hazards models with age as the time scale.Results: In our primary analysis using creatinine standardized cadmium levels (μg Cd/g cr), we report HR=1.12 per interquartile range difference (95% CI= 1.02-1.22) in models adjusted for gender, BMI, education, and urine cotinine. Findings were similar when other strategies for urine dilution adjustment were implemented: regression adjustment for creatinine (HR=1.15; 95% CI=1.02-1.30); standardization for covariate-adjusted creatinine (HR=1.17; 95% CI=1.05-1.30); osmolality standardized cadmium (HR=1.11; 95% CI=1.01-1.22); and regression adjustment for osmolality (HR=1.13; 95% CI=1.02-1.26). Results were also similar when we compared upper quartile with lower quartile exposure; additional case populations of stroke and acute myocardial infarction also showed similar results.Conclusion: In our case-cohort study of cadmium and heart failure among never-smokers, selection of urine dilution marker or adjustment strategy did not alter findings.