Research Studies
Study Tracker
Does fluoride exposure affect thyroid function? A systematic review and dose-response meta-analysis.Does fluoride exposure affect thyroid function? A systematic review and dose-response meta-analysis.
Abstract
Introduction
Fluoride exposure may have various adverse health effects, including affecting thyroid function and disease risk, but the pattern of such relation is still uncertain.
Methods
We systematically searched human studies assessing the relation between fluoride exposure and thyroid function and disease. We compared the highest versus the lowest fluoride category across these studies, and we performed a one-stage dose-response meta-analysis for aggregated data to explore the shape of the association.
Results
Most retrieved studies (27 of which with a cross-sectional design) were conducted in Asia and in children, assessing fluoride exposure through its concentrations in drinking water, urine, serum, or dietary intake. Twenty-four studies reported data on thyroid function by measuring thyroid-related hormones in blood (mainly thyroid-stimulating-hormone – TSH), 9 reported data on thyroid disease, and 4 on thyroid volume. By comparing the highest versus the lowest fluoride categories, overall mean TSH difference was 1.05 ?IU/mL. Dose-response curve showed no change in TSH concentrations in the lowest water fluoride exposure range, while the hormone levels started to linearly increase around 2.5 mg/L, also dependending on the risk of bias of the included studies. The association between biomarkers of fluoride exposure and TSH was also positive, with little evidence of a threshold. Evidence for an association between fluoride exposure and blood concentrations of thyroid hormones was less evident, though there was an indication of inverse association with triiodothyronine. For thyroid disease, the few available studies suggested a positive association with goiter and with hypothyroidism in both children and adults.
Conclusions
Overall, exposure to high-fluoride drinking water appears to non-linearly affect thyroid function and increase TSH release in children, starting above a threshold of exposure, and to increase the risk of some thyroid diseases.
Full-text study online at https://www.sciencedirect.com/science/article/pii/S001393512302563X?via%3Dihub
Excerpts:
4. Discussion
In this systematic review and dose-response meta-analysis on thyroid function as related to fluoride exposure, we found a clear pattern of association between fluoride content in drinking water consumed by the study participants and their circulating TSH concentrations. However, this occurred only above 2 mg/L of water fluoride (2.5 mg/L when the studies with the best quality were considered), thus confirming the hypothesis of a non-linear, dose-dependent pattern of association, something suggested to occur also for the adverse effects of drinking water fluoride on children’s intelligence quotient (Veneri et al., 2023b). However, in the current pooled analysis on thyroid function, the removal of studies with the lowest methodological quality had a limited impact on the results, suggesting that the effect of biases on the thyroid effect estimates was not relevant. This positive but non-linear relation between exposure to fluoride through drinking water and circulating TSH concentrations was remarkably consistent across the many studies carried out about this association and included in this dose-response analysis (Andezhath et al., 2005; Hosur et al., 2012; Khandare et al., 2017, 2018; Kumar et al., 2018; Shaik et al., 2019; Singh et al., 2014; Wang et al., 2020; Yang et al., 2008; Yasmin et al., 2013; Zhang et al., 2015; Zulfiqar et al., 2019, 2020), further strengthening the possibility of a causal link.
Conversely, studies on TSH as related to biomarkers of fluoride exposure, i.e., urinary and serum fluoride, did not yield a pattern of association similar to that for water fluoride. The shape of the curve indicates a no-threshold positive association starting from the lowest exposure levels, and flattening at the highest exposure range, with the exception of studies of the highest methodological quality based on serum fluoride, for which the pattern of association was almost linear. Such discrepancies between fluoride exposure through drinking water and biomarkers could be due to different reasons, including the lower number of studies available for serum fluoride, possible differences among the analytical methods used in the included studies, or more importantly the lower reliability of studies based on biomarkers in reflecting real internal exposure (particularly for urinary fluoride), due to individual changes in fluoride absorption, metabolism and excretion. In addition, assessing fluoride exposure through a long-term, stable, and major determinant of intake, such as fluoride in drinking water, could be more reliable than considering more unstable and short-term indicators like urinary and serum fluoride levels (Hall et al., 2023), unless for very high exposures, when biomarkers tend to be considerably associated with the environmental sources of exposure. In fact, while residential drinking water fluoride level and water consumption tend to be substantially stable over time, fluoride intake with foods, beverages, or dental products can considerably vary. This may induce non-differential misclassification of antecedent exposure when using short-term biomarkers, biasing the risk and effect estimates towards the null. Demographic factors such as age and sex, lifestyle factors such as smoking, alcohol consumption and regular tooth-brushing, and dietary and metabolic factors influencing fluoride absorption and metabolism may also modify the association between fluoride exposure (as that occurring through drinking water or other sources) and biomarkers such as serum and urinary fluoride concentrations (Kishor et al., 2023; Maguire and Zohoori, 2013; Riddell et al., 2021). This is also reflected by the null or low correlation between fluoride exposure and fluoride biomarkers found in some of the studies (Carwile et al., 2020; Lavalle-Carrasco et al., 2023; Saad et al., 2022; Zohoori et al., 2019). Overall, the limitations of biomarkers (unless based on repetitive samples) in adequately reflecting long-term fluoride exposure could explain the different patterns of the dose-response associations for individual studies based on urinary or serum fluoride concentrations, as compared to the substantially homogeneous dose-response curves across studies on drinking water fluoride levels, and the different results for the various exposure indicators within single epidemiologic studies (Hall et al., 2023; Malin et al., 2018; Riddell et al., 2019; Yu et al., 2018b) as does the current meta-analysis. Interestingly, a dose-response pattern comparable to that found in our pooled analysis emerged from a Chinese cross-sectional study (Yu et al., 2018b) assessing the relation of water and urinary fluoride with dental fluorosis, where a threshold was identified of slightly less than 1 mg/L of water fluoride before detecting a positive association, while the association started already at the lowest urinary fluoride concentrations (from 0.01 mg/L). Consistently, in a recently published cross-sectional analysis of a cohort study in Canada (Hall et al., 2023), no linear association emerged between urinary fluoride and risk of hypothyroidism, while such association was positive with fluoride intake through beverages (OR: 1.25; 95% CI: 0.99–1.57), and particularly with drinking water fluoride concentrations (OR: 1.65; 95% CI: 1.04–2.60).
Despite the strong association between fluoride exposure and TSH, we were unable to confirm such a clear association (expected on the opposite direction) for circulating thyroid hormones, either total or free. This may have been due to the limited number of studies that precluded a reliable analysis through a spline-based non-linear model, or to the real absence of such effects, despite the forest plot indicated some slight changes, mostly a decrease, in the circulating concentrations of these hormones. In addition, such possible effects on circulating thyroid hormones should be better appreciated at the individual level, particularly in subjects characterized by subtle thyroid hormone deficiency or located at the lowest range of such levels, these likely being the individuals most susceptible to fluoride-induced adverse effects on thyroid function. Finally, it cannot be excluded that different analytical methods for thyroid hormones might have affected comparisons across studies.
With reference to thyroid disease risk, we could retrieve too few studies for a dose-response meta-analysis. Results of such studies (Barberio et al., 2017; Day and Powell-Jackson, 1972; Eltom et al., 1984; Hong et al., 2008; Jooste et al., 1999; Lathman and Grech, 1967; Peckham et al., 2015; Siddiqui, 1960; Yang et al., 2008), either in children or in adults or in mixed populations, generally indicated a higher (in some cases much higher) risk of goiter, hypothyroidism, or thyroid diseases overall considered, with the exception of the few observations in the mixed population, showing a very inconsistent overall picture. Overall, the evidence available for thyroid disease risk as associated with fluoride exposure is too limited to draw conclusions, though there is a general indication of detrimental effects of exposure to the highest levels of fluoride occurring in the included studies. This appears also supported by biological plausibility, since laboratory and animal studies have suggested that excess fluoride exposure may decrease thyroid hormone synthesis and increase circulating TSH levels (Skórka-Majewicz et al., 2020). Fluoride may damage the proper functioning of the thyroid, interfering with Na/K-ATPase and with iodothyronine deiodinase, disrupting sensitive G-proteins of hormone receptors, mimicking TSH, and impairing T4 conversion into to T3, through effects on the enzymes catalyzing deiodination i.e., iodothyronine deiodinases (Bürgi et al., 1984; Hong-liang et al., 2014; Peeters and Visser, 2000; Shashi and Singla, 2013; Skórka-Majewicz et al., 2020; Spittle, 2016; Strunecka and Strunecky, 2020; Susheela and Toteja, 2018; Waugh, 2019). Fluoride could also damage thyroid cells by determining mitochondrial swelling and disintegration, and by inducing DNA-RNA damage (Peeters and Visser, 2000; Wei et al., 2022). It may also have adverse effects on thyroid cells and thyroid follicular morphology, altering thyroid structure (Liu et al., 2016; Sarkar and Pal, 2014; Yu et al., 2018a).
We must outline a few limitations of our assessment. First, some the epidemiologic studies did not provide the data to assess the dose-response relation between fluoride exposure and thyroid function and disease (for instance reporting only linear regression coefficients for the overall association). Secondly, it is known that both excess and deficient iodine intake may affect thyroid function and morphology, as well as could interact with fluoride in modifying the impact of both elements (Jiang et al., 2016; Malin et al., 2018). However, such potential confounding effect and interactions have not been generally addressed in the included studies, with some exceptions. Some studies investigated subjects with homogeneous iodine exposure while others determined iodine exposure in the same categories of fluoride exposure, most of which finding changes in the same direction of iodine and fluoride levels. Only one study (Du et al., 2021) adjusted for urinary iodine in the multivariable analysis, finding no association between urinary fluoride and TSH. In addition, there is some evidence of susceptibility of pregnant women to fluoride-induced hypothyroidism but based on one study only (Hall et al., 2023). More data are clearly required to better investigate the relation between fluoride exposure and thyroid function and disease risk, considering the overall nutritional status in relation to iodine and other trace elements as well as environmental pollutants possibly associated with thyroid function and disease (Chamot et al., 2023; Ge et al., 2023; Kim et al., 2022; Loomba et al., 2020; Urbano et al., 2022a, 2022b). In addition, fluoride toxicity on thyroid function, disease risk, and morphology could theoretically be enhanced in individuals carrying some functional alterations and impairment of the gland, or at an early stage of thyroid disease, and such a possible higher susceptibility should be investigated in-depth in epidemiologic and clinical studies. Genetic factors should also be taken into consideration when assessing the risk of any environmental exposure, including fluoride (Chakraborty et al., 2022; Cui et al., 2018; González-Casamada et al., 2022), and the availability of data with reference to fluoride overexposure and its effects on the thyroid gland is unfortunately very poor, being limited to few studies with inconclusive results (Wang et al., 2022; Xu et al., 2022; Zhang et al., 2015). Finally, since all the included studies are observational, the potential impact of unmeasured confounding cannot be ruled out, adding caution to the interpretation of the individual studies and of our pooled analysis. However, most of the studies appear to have considered major factors affecting TSH levels, such as age and sex (Xing et al., 2021).
Our findings are of considerable public health relevance given the importance of thyroid function for human health (including cognitive neurodevelopment), and the rather frequent occurrence of high fluoride concentrations in groundwaters in several regions of the world, including Africa (Rift Valley like Tanzania, Kenya and Ethiopia), Asia, and America. The World Health Organization (WHO) has estimated that around 260 million people inhabit locations having excessive fluoride levels in drinking water (>1.5 mg/L) globally (Jha and Tripathi, 2021). Our findings are also relevant to the public health issue of water fluoridation in order to prevent dental disease and caries in particular, given the uncertainties and the controversies surrounding this topic and the need of a comprehensive risk assessment of such type of intervention (Till and Green, 2021).
In conclusion, our systematic review and dose-response meta-analysis showed that at the highest levels of naturally occurring fluoride exposure there are detrimental effects on thyroid function and possibly thyroid disease risk, whose most evident and consistent finding is a dose-dependent increase in TSH concentrations associated with consumption of drinking water above 2.5 mg/L of fluoride.