Fluoride exposure and cognitive neurodevelopment: Systematic review and dose-response meta-analysis

Many uncertainties still surround the possible harmful effect of fluoride exposure on cognitive neurodevelopment in children. The aim of this systematic review and meta-analysis was to characterize this relation through a dose-response approach, by comparing the intelligence quotient (IQ) scores in the highest versus the lowest fluoride exposure category with a random-effects model, within a one-stage dose-response meta-analysis based on a cubic spline random-effects model. Out of 1996 potentially relevant literature records, 33 studies were eligible for this review, 30 of which were also suitable for meta-analysis. The summary mean difference of IQ score, comparing highest versus lowest fluoride categories and considering all types of exposure, was -4.68 (95% confidence interval-CI -6.45; -2.92), with a value of -5.60 (95% CI -7.76; -3.44) for drinking water fluoride and -3.84 (95% CI -7.93; 0.24) for urinary fluoride. Dose-response analysis showed a substantially linear IQ decrease for increasing water fluoride above 1 mg/L, with -3.05 (95% CI -4.06; -2.04) IQ points per 1 mg/L up to 2 mg/L, becoming steeper above such level. A weaker and substantially linear decrease of -2.15 (95% CI -4.48; 0.18) IQ points with increasing urinary fluoride emerged above 0.28 mg/L (approximately reflecting a water fluoride content of 0.7 mg/L). The inverse association between fluoride exposure and IQ was particularly strong in the studies at high risk of bias, while no adverse effect emerged in the only study judged at low risk of bias. Overall, most studies suggested an adverse effect of fluoride exposure on children's IQ, starting at low levels of exposure. However, a major role of residual confounding could not be ruled out, thus indicating the need of additional prospective studies at low risk of bias to conclusively assess the relation between fluoride exposure and cognitive neurodevelopment.