Background: The increasing worldwide incidence of overweight and obesity poses a significant health concern. The global application of neonicotinoids (NEOs) continues to rise. However, the relationship between NEOs and obesity remains unclear in middle-aged and elderly Chinese individuals. Objective: The purpose of this cross-sectional study was to assess the association between urinary concentrations of NEOs and obesity among individuals aged 35– 74 years in Guangxi, China. Methods: In this cross-sectional study, urinary concentrations of 10 NEOs were analyzed in 862 participants. Overweight and obesity (OWO) was defined as body mass index (BMI) ≥ 24 kg/m 2 , and abdominal obesity (ABO) was assessed by waist circumference (WC; male ≥ 90 cm, female ≥ 85 cm). The association between NEOs and obesity was evaluated through multinomial logistic regression, generalised linear models (GLM), quantile g-estimation (Qgcomp), and Bayesian kernel machine regression (BKMR). Machine learning models with Shapley Additive Explanations (SHAP) were used to explore the predictive contribution of NEOs and traditional risk factors to obesity. Results: Multivariate logistic regression showed that clothianidin (CLO), N-desmethyl-acetamiprid (NACE), and imidacloprid (IMI) were positively associated with Group 4 (with both OWO and ABO). The GLM model revealed a significant positive association between NACE and ABO (OR = 1.177, 95% CI: 1.046, 1.328, p < 0.05). In females, CLO was associated with OWO, while IMI was associated with both OWO and ABO. In males, NACE was associated with OWO and ABO. Both the Qgcomp and BKMR models indicated that mixed NEOs exposure was significantly correlated with obesity, showing a positive relationship for OWO in females and ABO in males. Machine learning identified CLO and NACE as factors significantly associated with obesity. Conclusion: Research findings indicated that CLO, NACE, IMI, and NEOs mixture were positively associated with obesity, with CLO and NACE serving as significant factors. Keywords: neonicotinoids, overweight and obesity, abdominal obesity, machine learning
Ou et al. (Sun,) studied this question.