Objective: Prenatal exposure to fluoride has been associated with adverse neurodevelopmental outcomes. However, the neuropsychological profile of fluoride's developmental neurotoxicity at low levels and the stability of this relationship across childhood has not been characterized. We investigated the longitudinal and domain specific effect of prenatal fluoride exposure on IQ among children ages 4, 5, and 6-12 years in the Early Life Exposures in Mexico to Environmental Toxicants (ELEMENT) cohort.

Methods: We measured the average of maternal urinary fluoride at each trimester of pregnancy adjusted for creatinine (MUF_CRE). Children were administered the McCarthy Scales of Children's Abilities at ages 4 (N = 386) and 5 (N = 308), and the Wechsler Abbreviated Scale of Intelligence at age 6-12 (N = 278). We used generalized estimating equation (GEE) models to estimate the population averaged effect of MUF_CRE concentration on longitudinal General Cognitive Index (GCI)/Full-Scale IQ (FSIQ), Verbal IQ (VIQ), and Performance IQ (PIQ) scores (N = 348). We tested for possible interactions between MUF_CRE and child sex as well as for MUF_CRE and time point on children's IQ. All models controlled for relevant available covariates.

Results: The mean/median MUF_CRE concentration was 0.90/0.83 mg/L (SD = 0.39; IQR, 0.64-1.11 mg/L). A 0.5 mg/L increase in MUF_CRE predicted an average 2.12-point decrease in GCI/FSIQ (95% CI: -3.49, -0.75) and 2.63-point decrease in PIQ (95% CI: -3.87, -1.40). MUF_CRE was marginally associated with VIQ across time (B = -1.29, 95% CI: -2.60, 0.01). No interactions between MUF_CRE and child sex or MUF_CRE and time were observed.

Conclusion: The negative association between prenatal fluoride exposure and longitudinal IQ was driven by decrements in non-verbal intelligence (i.e. PIQ), suggesting that visual-spatial and perceptual reasoning abilities may be more impacted by prenatal fluoride exposure as compared to verbal abilities.