Thyroid function, intelligence, and low-moderate fluoride exposure among Chinese school-age children.

3. Results

3.4. Associations between THs and IQ scores

4. DiscussionIn the present study, we found that fluoride exposure was inversely related to IQ scores and levels of TT₄ and FT₄, while positively associated with levels of TSH. Higher TT₃, FT₃ were related to increased odds of children having high normal intelligence, and TSH may show a modification effect on the association between urinary fluoride and IQ scores.

The median concentration of water fluoride in this study is equivalent to the Standards for Drinking Water Quality in China (1 mg/L) (Jin et al., 2006) and within the WHO recommended limit (1.5 mg/L) (World Health Organization, 2017), suggesting the residents were exposed to fluoride in drinking water at low-moderate levels. The annual surveillance data from the local CDC revealed that fluoride concentrations maintained at stable levels in the villages in which our study participants resided. Consequently, the water fluoride levels could represent long-term external fluoride exposure concentrations. Notably, kidney as a site of active metabolism excretes approximately 60% of absorbed fluoride among healthy adults, and 45% among healthy children (Buzalaf et al., 2011). Our results further showed that urinary fluoride levels presented a strongly positive correlation with water fluoride concentration (r_s = 0.73, P < 0.001), indicating that fluoride from drinking water makes important contribution to urinary fluoride. Therefore, urinary fluoride concentration as an internal exposure index can systematically reflect the burden of fluoride in drinking water. To evaluate the influences of fluoride on children comprehensively, we selected water fluoride concentrations as an external exposure parameter and urinary fluoride levels as an internal measure of exposure.

Epidemiological studies showed that fluoride concentration in the thyroid exceeds that found in all other soft tissues except for the kidney and that there is an association between endemic goiter and either fluoride exposure or enamel fluorosis in human populations (National Research Council, 2006), suggesting a close relationship between fluoride exposure and thyroid function. It has been shown that higher water fluoride levels were associated with increased likelihood of a hypothyroidism diagnosis among adults, manifesting as the higher TSH values with a higher fluoride concentration in the drinking water (Kheradpisheh et al., 2018, Peckham et al., 2015). Water fluoride contents were positively related to serum TSH levels in school going children in randomly selected villages of Nalgonda district, Telangana, India (Khandare et al., 2018). These results are partly consistent with our findings. Moreover, we found negative relationships between urinary fluoride levels and TT₄, FT₄ as well. However, another Indian study conducted in Doda district of Jammu and Kashmir showed that serum TSH levels in school children from high fluoride area were significantly lower than those in control area (Khandare et al., 2017). The discrepancies among these findings may be attributable to the different study areas, different exposure levels, different population and multiple sources of fluoride exposure exist. A cross-sectional study conducted in Canada showed that the combination of higher fluoride and low iodine was associated with higher TSH levels (Malin et al., 2018). Furthermore, another study in China discovered that a low iodine coupled with high fluoride intake was associated with the somatic developmental disturbance of iodine deficiency (Lin et al., 1991). However, the results from our study showed that fluoride exposure was independently associated with alterations in THs concentrations and IQ scores. Consistently, there is evidence suggesting that the impact of fluoride on the thyroid gland can occur independently of iodine (Peckham et al., 2015).

In the current work, our results demonstrated clearly that, across the full range of water and urinary fluoride concentrations and using a measure to focus on children’s IQ scores, higher fluoride levels were associated with lower IQ scores. These results are consistent with a systematic review and meta-analysis of 27 cross-sectional studies of children (mainly from China) exposed to fluoride (Choi et al., 2012), especially with the findings reported by Ding et al., who showed that water fluoride exposure at comparable low levels (range, 0.24–2.84 mg/L; mean value, 1.31 ± 1.05 mg/L) had significant negative associations with children’s intelligence (Ding et al., 2011). Moreover, we found that TT₃, FT₃ were related to increased odds of children having high normal intelligence, consistent with a previous study that has reported associations between decreased maternal or neonatal T3 levels and poorer neurodevelopment in children (Simic et al., 2009). Two case-control studies demonstrated that low maternal serum FT₄ concentrations were associated with impaired neurodevelopment in children (Hollowell et al., 1999, Pop et al., 1999). Conversely, there is also evidence that neonatal T4 levels were not associated with the risk of a heterogeneous group of developmental diagnoses in 5–12 years old children, including attention deficit disorder, cognitive disorder and learning disability (Soldin et al., 2003), which was consistent with our findings.

We further examined the roles of THs in the relationship between fluoride exposure and IQ scores and found that higher urinary fluoride levels were associated with lower IQ scores for children in the first two quartiles of TSH compared to those in the third and fourth quartiles and there was a significant modification effect by TSH on the association between urinary fluoride levels and IQ scores. As for TT₄ and FT₄, associations between fluoride concentrations and IQ scores also existed for the highest quartile group, though the effect modification was not significant. Taken together, these findings suggest that across the full range of water and urinary fluoride concentrations children who have lower TSH levels or higher TT₄ and FT₄ levels may be at an increased risk for lower intelligence. To our knowledge, this is the first population-based study to examine the modification effects of THs on the relationship between fluoride and IQ scores in Chinese school-age children. More large-scale studies are warranted to understand the mechanisms by which fluoride and THs interact within the body to affect children’s intelligence.

Our study has several strengths. Although there have been plenty of literatures demonstrating the association between fluoride exposure and intellectual development, this is the first study to detect the role of THs in this relationship. Compared with most previous studies that focus on the impact of high concentrations of fluoride, our research provided more information on the health effects of low-moderate fluoride exposure, which could further complete the epidemiological evidence on the biological disadvantages of fluoride acroSss different levels. Additionally, not only did we detect associations between fluoride exposure and THs, we also uncovered the modification effects of gender on the relationship.

Our study also has some limitations. We measured fluoride levels in early morning spot urine samples instead of 24-h urine collections. However, others have noted a close relationship between the fluoride concentrations of early morning samples and 24-h specimens (Zohouri et al., 2006). Moreover, given variation in urine dilution, crude urinary fluoride measures may influence the accuracy for exposure assessment, therefore the data remain to be compared with those after adjustment for urine creatinine or specific gravity. In addition, our study was cross-sectional in nature, resulting a weak contributor towards causal inference. Statistical power is also limited by the small size of the study population. However, in this perspective, the associations between fluoride exposure and IQ outcome and changes in THs levels are noteworthy.

5. Conclusions