Fluoride Exposure and Skeletal Fluorosis: a Systematic Review and Dose-response Meta-analysis.

Abstract

We performed a systematic review and meta-analysis on the relation between fluoride exposure and skeletal fluorosis (SF) using a novel statistical methodology for dose-response modeling. Skeletal fluorosis, a major health issue that is endemic in some regions, affects millions of people worldwide. However, data regarding the dose-response relation between fluoride exposure and SF are limited and outdated. We included twenty-three studies in the meta-analysis. When comparing the highest versus the lowest fluoride category, the summary risk ratio (RR) for SF prevalence was 2.05 (95% CI 1.60; 2.64), with a value of 2.73 (95% CI 1.92; 3.90) for drinking water and 1.40 (95% CI 0.90; 2.17) for urinary fluoride. The RR by the risk of bias (RoB) was 2.37 (95% CI 1.56; 3.58) and 1.78 (95% CI 1.34; 2.36) for moderate and high RoB studies, respectively. The dose-response curve based on a one-stage cubic spline regression model showed an almost linear positive relation between exposure and SF occurrence starting from relatively low concentrations up to 5 mg/L and 2.5 mg/L, respectively, for water and urinary fluoride, with no substantial increase above this threshold. The RR for developing moderate-severe forms increases at 5.00 mg/L and 2.5 mg/L of water and urinary fluoride, respectively. Better-quality studies are needed to confirm these results, but greater attention should be given to water fluoride levels to prevent SF, in addition to the other potential adverse effects of fluoride exposure.