Effects of antidepressants in the reproduction of aquatic organisms: a meta-analysis

Pharmaceuticals are widespread contaminants across the aquatic systems and despite the role that these compounds play in society today, little is known about their effects in aquatic organisms once they are released into the environment. This study aims to provide quantitative insight on the effects of antidepressants on the reproduction of fish and crustaceans with a multilevel meta-analysis. A systematic literature search identified 19 studies investigating effects of antidepressant exposure in fish or crustaceans' reproduction, according to specific selection criteria. Heterogeneity analysis was performed and the moderating effect of the range of exposure concentrations, exposure time, organism group and toxicant was tested. Additionally, publication bias was also addressed. The results showed that, overall, there is no significant association between antidepressant exposure and the reproduction of fish and crustaceans, however, moderator analysis revealed that the range of concentrations is a significant moderator for fish and crustacean fecundity, showing contrary results between the two groups. Antidepressant concentration had a small, yet positive effect on fish fecundity, meaning that increased concentrations resulted in increased fish fecundity, whilst a negative effect on crustaceans' fecundity was apparent with increasing concentrations. This difference could be related to data artifacts, or, more likely, evidencing a hormetic dose-response curve, with different ranges of exposure concentrations considered in studies on fish and crustaceans. Antidepressants have shown effects on reproductive outcomes in aquatic organisms, based on individual studies and narrative reviews. However, our results show that other factors can have an important role. Additionally, data available for a quantitative assessment is scarce, focusing mainly on a few freshwater species, low concentration ranges and one SSRI compound, fluoxetine. Thus, more research on the subject is needed since meta-analysis are only as statistically powerful as the number of, good quality, studies they include.