Urinary Concentration Correction Methods for Arsenic, Cadmium, and Mercury: a Systematic Review of Practice-Based Evidence

Background: Urinary biomonitoring is widely used to assess environmental chemical exposure; however, a critical gap exists in whether and how to correct for the physiological variation in water content of spot urine samples. Objective: The aim of this systematic review is to summarize the available evidence comparing the performance of urinary concentration correction methods used to determine urinary levels of arsenic, cadmium, and mercury. Methods: We searched PubMed/MEDLINE, Embase, LILIAC, Web of Science, and TOXNET up to Sept. 5, 2017 for articles evaluating urinary concentration correction methods (e.g., urine creatinine [U-Cre], specific gravity [U-SG], osmolality [U-Osm]) compared to 24-h or timed urine specimens for levels of arsenic, cadmium, and mercury. Data on study design, methods of urine collection, and the performance of selected correction methods were extracted. Results: A total of 10 papers met the inclusion criteria. Two papers evaluated the performance of urinary concentration correction methods for arsenic, four for cadmium, three for mercury, and one for multiple metals. The median sample size for arsenic was 105, for cadmium 107, and for mercury 35. The studies were highly heterogeneous in population selection, urine collection, urine quality control, statistical comparison among selected correction methods, and presentation of the results. The median (range) of correlation coefficients comparing each corrected values with corresponding levels of timed urine specimens are 0.74 (0.17–0.92) for un-correction (n = 13), 0.82 (0.52–0.98) for U-Cre (n = 13), and 0.75 (0.28–0.98) (n = 12) for U-SG. Conclusion: Findings from limited evidence support that urine creatinine and urine-specific gravity corrections remain practical approaches to correct metal concentrations for urine dilution as compared to 24-h or 12-h urine samples. Further studies with larger sample sizes are needed to clarify this fundamental issue of environmental biomonitoring using spot urine samples in both general and priority populations.