Disinfection by-products in drinking water and colorectal cancer: a meta-analysis

Methods A literature search, without language or time limits, was performed to identify relevant case-control and cohort studies. Separate risk estimates for colon and rectal cancer were extracted from studies meeting the inclusion criteria. Relative risks (RRs) or odds ratios (ORs) comparing the highest exposure category with the lowest were pooled using random effects methods. Results A total of 13 studies (3 cohort and 10 case-control) were analysed. For colon cancer, the pooled RR estimates were 1.11 [95% confidence interval (CI): 0.73-1.70] for cohort studies, 1.33 (95% CI: 1.12-1.57) for case-control studies and 1.27 (95% CI: 1.08-1.50) combining both study types. For rectal cancer, the corresponding RR estimates were 0.88 (0.57-1.35), 1.40 (1.15-1.70) and 1.30 (1.06-1.59). Sensitivity analysis showed these results were not importantly influenced by any single study. Publication bias was not evident for the colon cancer analysis but may have been a minor issue for the rectal cancer analysis. The results for rectal cancer may have been influenced by the quality of the studies. Conclusions The study findings provide limited evidence of a positive association between colorectal cancer and exposure to DBPs in drinking water. The small number of studies and limitations in study quality prevent causal inference.