Abstract
High dose arsenic in drinking water (≥100 μg L(-1)) is known to induce lung cancer, but lung cancer risks at low to moderate arsenic levels and its dose-response relationship remains inconclusive. We conducted a systematic review of cohort and case-control studies that quantitatively reported the association between arsenic concentrations in drinking water and lung cancer risks by searching the PubMed database till June 14, 2018. Pooled relative risks (RRs) of lung cancer associated with full range (10 μg L(-1)-1000 μg L(-1)) and low to moderate range (<100 μg L(-1)) of water arsenic concentrations were calculated using random-effects models. A dose-response meta-analysis was performed to estimate the pooled associations between restricted cubic splines of log-transformed water arsenic and the lung cancer risks. Fifteen studies (9 case-control and 6 cohort studies) involving a total of 218 481 participants met the inclusion criteria. Meta-analysis identified significantly increased risks of lung cancer on exposure to both full range (RR = 1.21; 95% confidence interval [CI] = 1.05-1.37; heterogeneity I (2) = 54.3%) and low to moderate range (RR = 1.18; 95%CI = 1.00-1.35; I (2) = 56.3%) of arsenic-containing water. In the dose-response meta-analysis of eight case-control studies, we found no evidence of non-linearity, although statistical power was limited. The corresponding pooled RRs and their 95%CIs for exposure to 10 μg L(-1), 50 μg L(-1), and 100 μg L(-1) water arsenic were 1.02 (1.00-1.03), 1.10 (1.04-1.15), and 1.20 (1.08-1.32), respectively. We provide evidence on the association between increased lung cancer risks and inorganic arsenic in drinking water across low, moderate and high levels. Minimizing arsenic levels in drinking water may be of public health importance.