Multilevel meta-analysis and dose-response analysis for bisphenol A (BPA) exposure on metabolic and neurobehavioral effects.

Evidence suggests that oral exposure to bisphenol A (BPA) may result in adverse metabolic and neurobehavioral effects. The aim of the present meta-analysis is to examine this association based on systematically selected laboratory rodent studies published from 2012 to 2021 and sourced from Scopus, Web of Science, EmBase, and PubMed. Articles satisfying eligibility and inclusion criteria were included for the calculation of the summary standardised mean difference (SMD). Subgroup analysis and subsequent dose-response analysis were conducted if applicable. In total, 32 studies were analysed for 6 metabolic endpoints (cholesterol, triglycerides, insulin, glucose, leptin, and adiponectin) and 6 neurobehavioral endpoints (locomotor activity, exploratory, anxiety, depression, spatial learning and memory, non-spatial learning and memory). Summary SMDs implied that no significant effects were observed in endpoints considered. The dose was not determined as a significant moderator with regards to medium or high heterogeneity; however, there was significant impairment of spatial learning and memory at health-based guidance value ('HBGV') (0.05-9 mg (kg bw)-1 day-1) and 'High' (>9 mg (kg bw)-1 day-1) dose group. As a result, an indicative toxicological reference dose value of 0.034 mg (kg bw)-1 day-1 was proposed due to large variability. Potential harm to spatial learning and memory from BPA exposure requires further investigation. This study has provided some additional information on potential adverse metabolic and neurobehavioral effects of BPA from the perspective of meta-analysis which can inform the public, regulatory authorities, and policymakers.