Tryptophan metabolism in neonicotinoids exposure-induced diabetes: Emerging insights and predictive implications

Background: Exposure to individual neonicotinoids (NNIs) suggestively correlates with the type 2 diabetes mellitus (T2DM) risk in animals. However, existing systematic epidemiological investigations remain scarce and lack inclusion of predictive biological markers. Objective: This study aimed to evaluate the associations between exposure to NNI and diabetes, and to identify potential biomarkers by developing predictive models. Methods: We enrolled 1119 older individuals from community health screenings and categorized them into three groups: diabetes (n = 224), prediabetes (n = 184), and controls (n = 706). Serum levels of 15 NNIs and their metabolites were measured. Generalized linear models were used to assess correlations between the exposure to individual NNIs and T2DM and prediabetes (PDM). Quantile gcomputation (qgcomp), Bayesian kernel machine regression (BKMR) models, and weighted quantile sum regression were utilized to investigate the correlations between NNI mixtures and T2DM and PDM risks. Untargeted metabolomics was performed to examine relevant metabolites in the NNI-induced T2DM. Targeted metabolomics were then quantified. Mediation analysis was conducted to examine whether amino acid metabolism mediates the observed associations, and a prediction model was established to assess the risk of diabetes owing to NNI exposure. Result: The detection rates of most NNI pesticides exceeded 90%, with the highest reaching 99.55%. The qgcomp data showed that 6-chloronicotinic acid contributed most significantly to T2DM, whereas thiacloprid-amide (THD-A) showed the strongest contribution to PDM. BKMR models indicated that co-exposure to multiple NNIs increase the risk of developing both T2DM and PDM. Non-targeted metabolomics revealed that the tryptophan (Trp) metabolic pathway mediates the development of diabetes mellitus, with kynurenine (KYN)-mediated THD association with T2DM showing the largest proportion of mediation at 24.93%. A predictive model for NNI-diabetes was developed by incorporating KYN, tryptamine, xanthurenic acid, and the body mass index. An 80.4% consistency rate was determined between predicted and actual probabilities. Conclusions: This study reveals a dose-dependent association between neonicotinoid exposure and diabetes in older adults, accompanied by disturbances in tryptophan metabolite profiles. These findings provide new evidence for global research on the metabolic toxicity of environmental neonicotinoids and international public health risk assessment.