Human epidemiological evidence about the associations between exposure to organochlorine chemicals and endometriosis: Systematic review and meta-analysis

Background: Endometriosis is a gynaecological disease characterized by the presence of ectopic endometrial tissue that affects women during their reproductive years, having a strong impact on their lives, fertility and healthcare costs. The aetiology remains largely unknown, but current evidence suggests that it is multi-causal and oestrogen-dependent. Many epidemiologic studies have explored associations between organochlorine chemicals (OCCs) and endometriosis, but the findings are inconsistent. Objectives: A systematic review (SR) and meta-analysis were conducted to gather and synthesize all the available evidence from human epidemiological studies about the associations between OCCs and endometriosis. Data sources: The searches were conducted in PubMed and Web of Science in June 2016 with a final follow-up in August 2018. Study eligibility criteria: Only human epidemiological studies were considered, independent of participant age, body mass index or life-stage. Studies reporting individual measures of exposure to OCCs were included, considering but not limited to polychlorinated dibenzodioxins and dibenzofurans (PCDD/Fs), polychlorinated bi-phenyls (PCBs), or organochlorine pesticides (OCPs). The primary health outcome was presence of endometriosis, including all sub-types. Eligibility criteria excluded articles not written in English, conference papers, reviews and studies with overlapping information. Study appraisal and synthesis methods: A SR protocol pre-registered at PROSPERO was applied in duplicate to gather and extract all eligible original papers from PUBMED and Web of Science databases. Odds ratios were pooled using the inverse variance method for random effects meta-analysis for each group of OCCs. Risk of bias was assessed using the National Toxicology Program/Office of Health Assessment and Translation (NTP/ OHAT) Risk of Bias Rating Tool for Human and Animal Studies adapted to the review question. The confidence in the body of evidence and related level of evidence was measured by using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) based NTP/ OHAT framework. The results were structured and presented in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Results: Of the 51 studies retained for the full-text screening, 17 provided effect sizes and metrics sufficient for pooling estimates through meta-analysis. The overall odds ratios and 95% confidence intervals were 1.65 (1.14; 2.39) for dioxins (n=10), 1.70 (1.20; 2.39) for PCBs (n=9), and 1.23 (1.13; 1.36) for OCPs (n=5). Despite being statistically significant, these estimates should be considered with caution given the notable heterogeneity and small estimated effect size. Misclassification of exposure, due to varying laboratory detection rate capabilities, and disease status, due to varying definitions of endometriosis, were identified as major sources of uncertainty. Limitations, conclusions, and implications of key findings: The level of evidence was considered to be moderate with serious risk of bias according the NTP/OHAT criteria, supporting the need for further well-designed epidemiological research to fill lingering data gaps. Given the complexity of endometriosis and lack of known biomarkers suitable for population-based research, carefully designed observational studies play an important role in better understanding the aetiology of endometriosis, as will evolving mixture modeling approaches capable of handling various environmental chemical exposures. Attention to critical windows of exposure will shed further light on the possible developmental origin of endometriosis. Considering the high economic and societal cost associated with endometriosis, further research on this field is urged.