Causal effects of maternal exposure to PM2.5 during pregnancy on depression symptoms in adolescence: Identifying vulnerable windows and subpopulations in a national cohort study

Few studies have examined the causal relationship between chronic exposure to air pollutants during pregnancy and depression in adolescent offspring. In addition, it has not been investigated whether exposure is most harmful to adolescents in certain populations and at certain stages of pregnancy. A total of 1975 adolescents from 1632 families from the China Family Panel Study, a representative national longitudinal cohort, were included in this study. We used high-resolution satellite retrieval data to assess the PM2.5 exposure of mothers during pregnancy. Specifically, we employed a two-stage instrumental variable model (IV-2SLS) within the counterfactual causal inference framework, and selected and validated appropriate instruments, thereby mitigating potentially biased results arising from bi-direction between dependent and independent variables. This approach allowed us to explore the causal relationship between maternal PM2.5 exposure during pregnancy and adolescent depression symptoms. The endogeneity of air pollution during pregnancy and the need for a causal model were suggested by the results of the model comparisons. Using the IV-2SLS model, we found that maternal exposure to PM2.5 during pregnancy exacerbates depressive symptoms in the offspring during adolescence (β = 0.2, 95% CI: 0.05–0.34). We also found that exposure during the first trimester may cause greater harm. Adolescents with low household income, being male, irregular exercise habits, living in rural areas, and having mothers with poorer mental status may be more vulnerable. The findings suggest that maternal exposure to PM2.5 during pregnancy may have a negative impact on the depression symptoms of offspring in adolescence and that more attention should be paid to vulnerable populations and the window of vulnerability.