The association between maternal exposure to ten neonicotinoid insecticides and preterm birth in Guangxi, China

Preterm birth (PTB) is a primary cause of mortality among newborns globally. Prenatal exposure to environmental pollutants has been suggested to increase the PTB risk. Studies have shown NEOs may be linked to adverse birth outcomes. However, the impact of maternal NEOs exposure on PTB remains unclear. Therefore, to examine the association between NEOs exposure and PTB risk, we performed a case-control analysis utilizing data from a birth cohort study in Guangxi, China. A total of 157 preterm infants and 471 full-term infants were included. Concentrations of 10 NEOs and their metabolites in maternal serum were quantified using liquid chromatography-tandem mass spectrometry. We employed logistic regression, quantile g-computation, and restricted cubic spline models to evaluate the effects of individual and mixed NEOs exposures. Subsequently, XGBoost machine learning, combined with SHAP, was employed to predict the implications of serum NEOs on PTB. Finally, for 1-standard deviation increment in ln-transformed concentrations of imidacloprid and dinotefuran, significant correlations with higher odds of PTB were observed, showing odds ratios of 1.17 (95 % CI: 1.02, 1.36) and 1.41 (95 % CI: 1.16, 1.72). Similar patterns and higher risks were observed in late preterm birth. In both mixed exposure and machine learning models, dinotefuran and imidacloprid were identified as major predictors of increased PTB risk. Exposure to n-desmethylacetamiprid, sulfoxaflor, thiacloprid, nitenpyram, and thiamethoxam was negatively associated with PTB. Our findings suggested dinotefuran and imidacloprid exposure during pregnancy were risk factors of PTB, particularly among late preterm births. Subsequent research is necessary to illuminate the underlying mechanisms involved.