We, the Editors of the British Journal of Psychiatry, have retracted the following article: Gu et al, Reference Gu, Liu, Deng, Wang, Lin, Guo and Wu2019.

In January 2020, the authors alerted the editorial team to an error in the analyses. The authors transformed the effect estimates extracted from all included studies to the same scale (per 10 μg/m³ increase in particulate matter). The error occurred during the data transformation for the 95% confidence intervals of the extracted effect estimates. The authors inappropriately expanded the standard errors of some effect estimates during the log and antilog changes.

Following a detailed investigation by the senior editorial committee, it was identified that the main findings are unreliable as a result of the error. The corrected analyses no longer indicate an association between PM_2.5 and depression and the association with depression at either exposure (PM_2.5 and PM₁₀) are close to identical.

The authors agree to the retraction and the Editors would like to thank the authors for alerting the editorial team to the error.