Association of endemic arsenic poisoning with liver injury: Meta-analysis

Objective To systematically review the correlation between endemic arsenism and liver injury. Methods Pertinent studies were identified by searching Pubmed, Embase, China National Knowledge Infrastructure (CNKI) and Wanfang Data databases through November 2018. Studies that reported endemic arsenism and liver injury at home and abroad were collected. All of the Meta-analysis were performed by using Review Manager 5.3 software. The odds ratios (OR) and standardized mean differences (SMD) were used to compare continuous and dichotomous variables. The I2 was used to assess heterogeneity among studies, and the fixed-effects model or the random-effects model was chosen for the quantitative analysis. Subgroup analysis were performed to discuss sources of heterogeneity. The publication bias was evaluated by inverted funnel plot. Results Totally 16 documents were included, including 7 in English and 9 in Chinese. With liver injury (including the results of hepatomegaly epidemic, abnormal Bultrasound or pathological examination of liver, cirrhosis with definite diagnosis, etc.), different serum liver function and liver fibrosis indicators as the outcome, 7 274 cases, 410-820 cases and 255 cases were included in the arsenic exposure group, and 5 078 cases, 134-327 cases and 164 cases in the control group, respectively. The Meta-analysis showed that arsenic exposure was associated with increased risk of liver injury [OR value and 95% confidence interval (CI) was 2.40 (1.43-4.05), Z=3.30, P < 0.01]. Arsenic exposure had promoted the levels of serum aspartate aminotransferase (AST), alkaline phosphatase (ALP), glutamyltransferase (γ-GT) and total bile acid [TBA, SMD and 95%CI were 1.71 (0.18-3.23), 1.73 (0.42-3.03), 0.25 (0.09-0.41) and 0.55 (0.35-0.75), respectively, P < 0.05 or < 0.01], while had suppressed the level of serum albumin [ALB, SMD and 95%CI were-2.37 (-4.01 -- 0.73), P < 0.01]. Among the serum indicators for liver fibrosis, the levels of serum procollagen type Ⅲ (PC-Ⅲ) and collagen type Ⅳ(Ⅳ-C) were significantly increased in arsenic exposed patients [SMD and 95%CI were 0.98 (0.05-1.91), 1.60 (1.38-1.83), respectively, P < 0.05 or < 0.01]. Subgroup analysis showed that arsenic exposure had been stopped or not at the time of study, study design (including cross-sectional study and cohort study) had a significant effect on the association between arsenic exposure and liver injury (χ2=17.17, 11.85, P < 0.01). Conclusions Both drinking water-borne arsenicosis and coal burning-borne arsenicosis can cause high incidence of liver injury. ALP, ALB, γ-GT and TBA can be used as serum biomarkers for arsenic exposure induced liver injury. After blocking of arsenic exposure by adopting comprehensive prevention and control measures such as changing water sources and installation of ventilated stoves, the liver injury of patients could be effectively improved. Key words: Arsenic poisoning; Exposure; Liver injury; Meta-analysis