Age at menarche and exposure to non-persistent pesticides in Spanish girls from the INMA (environment and childhood) project

Our objective was to assess the longitudinal association between exposure to non-persistent pesticides and the age of menarche in Spanish girls. The study was conducted in a sample of 506 girls from the INMA Project. Pesticide exposure was assessed by quantifying four insecticide metabolites (TCPy, chlorpyrifos metabolite; IMPy, diazinon metabolite; DETP, non-specific organophosphate metabolite; 3-PBA, pyrethroid metabolite) and the dithiocarbamate fungicides metabolite ethylene thiourea (ETU) in spot urine samples collected at the age of 7-10 years (2013-2016). Information on the age at menarche was collected from questionnaires completed by the girls and/or parents at successive visits from the age of 7 to 16 years. Associations between categorical exposure biomarkers and age at menarche were examined using Cox regression models adjusted for sub-cohort; maternal schooling; age, urinary creatinine, body mass index (BMI) z-score, and height of girls at the age of 7-10 years; and other pesticide biomarkers. Effect modification by BMI was also examined. Menarche was later in girls with detected versus undetected concentrations of urinary TCPy (Hazard ratio [HR] = 0.71; 95%CI = 0.54-0.92); this association was stronger among those with overweight or obesity (HR = 0.69; 95%CI = 0.47-0.99), although the interaction was not statistically significant. By contrast, menarche was earlier in girls with increasing ETU (HR = 1.32; 95%CI = 1.00-1.73 for concentrations between limit of detection and 75th percentile [P75]; HR = 1.39, 95%CI = 1.03-1.89 for concentrations > P75 versus undetected ETU); this association was stronger but not significant in girls with underweight/normal weight. These results are in line with previous findings from the INMA Project on the association of ETU exposure with earlier puberty onset in girls and provide previously unreported evidence of its association with TCPy. Further studies are needed to verify these findings and to elucidate the role of BMI in these associations.