Association Between Fluoride Exposure and Asthma Among <scp>US</scp> Children and Adolescents

Asthma is one of the most common chronic non-communicable diseases, affecting 4.7 million United States (US) children. Its prevalence varies greatly across regions and countries, highlighting the potential role of local environmental factors. Approximately 62.8% of the US population receives fluoridated water [1]. Toxicological evidence shows that fluoride can induce lung damage through mechanisms involving oxidative stress and inflammation [2]. Additionally, high fluoride exposure may alter oral and gut microbiomes, which have been linked to immune function and asthma susceptibility [3, 4]. Although occupational exposure to fluoride has been associated with asthmatic symptoms [5], no study has explored the association between fluoride exposure and asthma among children or adolescents. Therefore, we explored whether fluoride exposure was associated with asthma risk among a US-representative sample of children and adolescents, stratified by tobacco smoke exposure—a strong risk factor for asthma. Detailed methodological information is shown in the Supplementary Text. Briefly, in 2013–2016, the National Health and Nutrition Examination Survey (NHANES) measured fluoride concentrations in plasma, urine, and drinking water, collected information on fluoride supplementation, and assessed dental fluorosis among 5451 children and adolescents aged 6–19 years. We excluded those who did not complete the physical examination (n = 5) or had missing cotinine data (n = 39), leaving 5407 participants for the present study (Figure 1A and Table S1). Confounders were selected using a Directed Acyclic Graph (Figure S1). In the primary analysis with adjustment for confounders, we found positive dose–response relationships between quartiles of plasma fluoride concentrations and ever asthma risk among participants who consumed tobacco products or had serum cotinine exceeding 1 ng/mL (p = 0.006; Figure 2A). In a study conducted in Turkey, Sezgin and colleagues examined asthma prevalence among 30 children living in a village with high drinking-water fluoride concentrations and found that asthma frequency was similar to national estimates [6]. However, its sample size is extremely small. Additionally, this study did not account for key covariates (e.g., tobacco exposure). In our present study, we found a high relative excess risk of ever asthma due to the interaction between plasma fluoride and tobacco exposure (P for additive interaction = 0.005), suggesting that fluoride may exacerbate the effects of tobacco exposure on asthma. In a secondary analysis by considering recent asthmatic episodes, we found a suggestive dose–response relationship between plasma fluoride and current asthma risk among participants exposed to tobacco smoke (p = 0.07; Figure 2A). Fluoride supplement intake and dental fluorosis were also associated with an increased current asthma risk, especially among participants exposed to tobacco smoke (p = 0.03 and 0.09, respectively). Fluoride supplement intake reflects long-term exposure, whereas dental fluorosis represents an irreversible biomarker of early childhood exposure. In contrast, drinking-water fluoride reflects recent external exposure levels, which were unrelated to asthma in our study population (Figure 2A). This is not unexpected given the low predictive power of drinking-water fluoride for plasma fluoride (Figure 1C). Both plasma and urinary fluoride reflect internal exposure status, but plasma fluoride represents the circulating fraction most likely to interact with target tissues and pathways relevant to asthma. A lack of association for urinary fluoride in relation to asthma may also partly reflect its high within-person variability. The association of plasma fluoride and asthma was further confirmed using spline models and population-averaged predicted probability estimates (Figure 2B,C). Sensitivity analyses showed similar results when additionally adjusting for examination sessions and sampling seasons, replacing missing covariates using the multiple imputation method, excluding participants with renal impairment, and using an alternative cotinine cutoff (3 ng/mL) (Figure S2). There are several limitations. First, NHANES is a cross-sectional survey, which cannot directly establish causal relationships. Second, uncontrolled confounding (e.g., duration of exposure, recent dental procedures, co-exposure to other chemicals) cannot be fully ruled out. However, the relatively large E-values suggest that our findings are unlikely to be fully explained away by unmeasured or unknown confounders (Table S2). Third, chance findings cannot be fully excluded due to multiple tests. Nevertheless, we are reassured by existing toxicological and epidemiological evidence and the consistency in results across sensitivity analyses. In conclusion, our findings underscore the importance of reevaluating current fluoridation standards in water supply systems and reducing exposure to both known (tobacco smoke) and inadvertent (fluoride) risk factors of asthma among young populations. However, additional longitudinal studies are needed to validate our novel findings. Q.X. and Y.S. analyzed the data. Q.X. and Y.-X.W. drafted the manuscript. Y.-X.W. and Y.J.Y. led the study conception, study design, analysis plan, and interpretation of findings. C.M., C.Y.D., Y.Z., V.M., C.L., K.Y.S., and J.L. validated the accuracy of data analysis with a technical review. All authors contributed to interpreting results and critically reviewed the manuscript for important intellectual content. We acknowledge the participants in the study. This study was supported by the National Natural Science Foundation of China [No. 82473581]. The authors declare no conflicts of interest. The data that support the findings of this study are available in the National Health and Nutrition Examination Survey (NHANES) at https://wwwn.cdc.gov/nchs/nhanes/Default.aspx. These data were derived from the following resources available in the public domain: NHANES 2013–2014, https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2013: NHANES 2015–2016, https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2015. Figure S1: The final directed acyclic graph (DAG) depicting the assumed causal relationships between fluoride exposure (including plasma, urine, and drinking water samples, fluoride supplement use, and dental fluorosis) and asthma, and their covariates. In the final DAG, each node (i.e., each circle) represents a specific variable, and a directed edge (i.e., a line with an arrow) between a pair of nodes represents a causal relationship between them, starting from the cause and ending at the outcome. In the centre of the DAG figure, the green node of fluoride exposure is the exposure variable, the blue node of asthma is the outcome variable, and the green line between them represents the causal path of interest in this study. Finally, we identified the following confounding factors: age, sex, race/ethnicity, body mass index (BMI) z-score, family income to poverty ratio, survey cycle, Healthy Eating Index-2015 scores, tobacco smoke exposure, examination session, sampling seasons, and family history of asthma. Although family history of asthma was not associated with exposure, given that it had a significant impact on the outcome, we ultimately adjusted for family history of asthma in the analysis. Figure S2: Sensitivity analyses for the associations of fluoride exposure with ever and current asthma according to tobacco smoke exposure (NHANES, 2013–2016). (A) Models were adjusted for age, sex, race/ethnicity, body mass index (BMI) z-score, family income to poverty ratio, family history of asthma, survey cycle, Healthy Eating Index-2015 scores, tobacco smoke exposure, examination sessions, and sampling seasons. (B) Multiple imputation methods were used instead of imputing missing covariates, with adjustment for age, sex, race/ethnicity, body mass index (BMI) z-score, family income to poverty ratio, family history of asthma, survey cycles, Healthy Eating Index-2015 scores, and tobacco smoke exposure. (C) Associations between creatinine-adjusted urinary fluoride concentrations and asthma after excluding participants with renal impairment, which is defined as having a glomerular filtration rate (GFR) of < 60 mL/min/1.73 m2 or an albumin-to-creatinine ratio (ACR) of ≥ 30 mg/g. (D) Associations between plasma fluoride exposure and asthma according to active smoking (active smokers were defined as those who recently consumed tobacco/nicotine products or whose serum cotinine concentrations exceeded 3 ng/mL). (E) The interaction between plasma fluoride exposure and serum cotinine concentrations on asthma risk. CI, confidence interval; OR, odds ratio; Q, quartile; RERI: relative excess risk due to interaction. Two groups were created based on plasma fluoride concentrations (< 25th and ≥ 25th) to assess the additive interaction. Table S1: Characteristics of study participants according to fluoride exposure (NHANES, 2013–2016). Table S2: The E-value to assess the potential effect of unmeasured confounding for the associations between fluoride exposure and asthma according to tobacco smoke exposure among children and adolescents (NHANES, 2013–2016). Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.