Excessive exposure to fluoride can reduce intelligence. Methylenetetrahydrofolate dehydrogenase, cyclohydrolase, and formyltetrahydrofolate synthetase 1 (MTHFD1) polymorphisms have important roles in neurodevelopment. However, the association of MTHFD1 polymorphisms with children’s intelligence changes in endemic fluorosis areas has been rarely explored.
A cross-sectional study was conducted in four randomly selected primary schools in Tongxu County, Henan Province, from April to May in 2017. A total of 694 children aged 8 to 12 years were included in the study with the recruitment by the cluster sampling method. Urinary fluoride (UF) and urinary creatinine were separately determined using the fluoride ion-selective electrode and creatinine assay kit. Children were classified as the high fluoride group and control group according to the median of urinary creatinine-adjusted urinary fluoride (UFCr) level. Four loci of MTHFD1 were genotyped, and the Combined Raven’s Test was used to evaluate children’s intelligence quotient (IQ). Generalized linear model and multinomial logistic regression model were performed to analyze the associations between children’s UFCr level, MTHFD1 polymorphisms, and intelligence. The general linear model was used to explore the effects of gene-environment and gene-gene interaction on intelligence.
In the high fluoride group, children’s IQ scores decreased by 2.502 when the UFCr level increased by 1.0 mg/L (B = -2.502, 95% confidence interval [CI]:—4.411, —0.593), and the possibility for having “excellent” intelligence decreased by 46.3% (odds ratio = 0.537, 95% CI: 0.290, 0.994). Children with the GG genotype showed increased IQ scores than those with the AA genotype of rs11627387 locus in the high fluoride group (P< 0.05). Interactions between fluoride exposure and MTHFD1 polymorphisms on intelligence were observed (Pinteraction < 0.05).
Our findings suggest that excessive fluoride exposure may have adverse effects on children’s intelligence, and changes in children’s intelligence may be associated with the interaction between fluoride and MTHFD1 polymorphisms.
Low-dose fluoride can prevent dental caries and is beneficial for bone growth.[1,2] However, the health damage caused by excessive fluoride should not be ignored. Dental and skeletal fluorosis are the most specific diseases caused by chronic intake of excessive fluoride.[4,5] Besides, fluoride exposure is related to the dysfunction of the reproductive system, as well as the liver and brain damage.[6-8]
Increasing studies have suggested that excessive fluoride is associated with disorders in cognition, learning, and memory ability.[9-11] Among them, a birth cohort study conducted in Canada suggested that increased fluoride concentrations in drinking water were associated with intellectual impairment in children. Another study reported a negative correlation between fluoride exposure and intelligence quotient (IQ) in children, which was similar to our previous study. In addition, an animal study demonstrated that fluoride exposure during development could induce cognitive deficits in mice. However, there was no evidence that reductions in IQ scores were caused by excessive fluoride in a prospective study conducted in a New Zealand community. Similarly, a cross-sectional study conducted in China mentioned that there was no statistical difference in IQ scores between children in different fluorosis areas and children in the control group (CG). Behind the different levels of fluoride exposure and assessment methods, different ethnic and genetic backgrounds may also explain the inconsistent results.
Individual susceptibility (eg, gene polymorphisms) is an important factor that affects the sensitivity of body health to environmental factors.[18,19] Several studies have reported that genetic polymorphisms can modify the harmful effects of fluoride, such as dental and skeletal fluorosis.[20,21] Studies have also shown that gene polymorphisms may modulate sensitivity to the effects of fluoride exposure on intelligence. For example, polymorphisms of the catechole-o-methyltransferase gene and dopamine receptor-2 gene have been reported to modify the adverse effects of fluoride on IQ scores.[22,23] Methylenetetrahydrofolate dehydrogenase, cyclohydrolase, and formyltetrahydrofolate synthetase 1 (MTHFD1) gene encode a nicotinamide adenine dinucleotide phosphate-dependent trifunctional enzyme, which provides the one-carbon derivatives of tetrahydrofolate in three sequential reactions and is involved in the folate pathway. Comprehension of the influential factors of neural development is incomplete, but folate deficiency has been implicated consistently in neuropathological lesions. Previous studies have pointed out that mouse with MTHFD1 mutations seemingly showed impaired functions of one-carbon metabolism and higher plasma homocysteine levels,[26,27] which have been further linked to cognitive impairment and Alzheimer’s disease.[28,29] In addition, certain loci polymorphisms of MTHFD1 were associated with neurological diseases in an epidemiological study. All these evidences suggest that MTHFD1 mutations may relate to neurodevelopment. However, few studies have focused on the effects of MTHFD1 polymorphisms on children’s intelligence.
Present study focused on four loci of MTHFD1 which are related to neurodevelopment: rs11627387, rs1076991, rs2236224, and rs2236225.[31-33] Studies have suggested that variations of these four loci can modify the impact of certain factors on health.[34-36] However, whether MTHFD1 polymorphisms are involved in the effect of fluoride exposure on children’s intelligence remains unclear.
Based on the above analyses, we conducted a cross-sectional study in an endemic drinking water-borne fluorosis area in Tongxu County, Kaifeng (Henan Province, China). We aimed at evaluating the effects of polymorphisms of MTHFD1 loci (rs11627387, rs1076991, rs2236224, and rs2236225) and excessive exposure to fluoride on children’s intelligence and explore the role of MTHFD1 polymorphisms in relationship between fluoride exposure and changes in children’s intelligence to provide a novel clue for the study of neurotoxicological mechanisms of fluoride.
The study protocol (No. ZZUIRB2017-018) was approved by the Ethics Review Board of Zhengzhou University (Zhengzhou, China). Children and their guardians were fully aware of the aim and process of our research and provided written informed consent.
Study design and population
As described thoroughly in our previous study, we conducted a cross-sectional study in Tongxu County, Henan Province, from April to May in 2017. Four primary schools were randomly selected. We excluded children who were non-local residents, on calcium supplements, had diseases based on calcium or phosphorus metabolism, had digestive diseases, and had thyroid diseases. Subsequently, 694 school-age children aged 8 to 12 years from grades 2 to 6 in four selected schools were recruited by cluster sampling. All children lived on campus and they had similar living conditions, living habits, and dietary structure.
A questionnaire was designed in advance and included information on sociodemographic data, medical history, maternal pregnancy, information on birth, and other information (eg, exercise). The height and weight of children were measured twice and the mean value was taken, which were accurate to 0.1 cm and 0.1 kg, respectively. Then, the body mass index (BMI) was calculated. In addition, fluoride-free containers were used to collect mid-flow morning urine and whole blood from the cubital vein after an overnight fasting. Samples of urine and blood were stored, respectively, at —20°C and —80°C for subsequent measurements.
Exposure assessment has been described in detail in our previous study. In accordance with the standard detailed by the health industry of China (WS/T 892015), a method based on a fluoride ion-selective electrode (Shanghai Exactitude Instruments, Shanghai, China) was conducted to determine the urinary fluoride (UF) level of children. A creatinine assay kit (Jiancheng Bioengineering Institute, Nanjing, China) was used to determine the concentration of urinary creatinine (UCr). Each determination of levels of UF and UCr was undertaken twice and averaged for data analyses. We calculated the urinary creatinine-adjusted urinary fluoride (UFCr) level to correct the influence of urine dilution on the UF level using the following equation: (UFCr[mg/L]=UF[mg/L]/UCr[mg/L])×UCr?(mg/L), where UCr-mean denotes the mean UCr concentration of the total population. Then, according to the median value of UFCr, children were separated into the high fluoride group (HFG, UFCr >1.33 mg/L) and CG, UFCr<1.33 mg/L.
The IQ was assessed using the second revision of the Combined Raven’s Test – the Rural in China (CRTRC2). Each student completed the paper independently with the supervision of trained investigators. Answer sheets were scored in accordance with the standard of the Combined Raven’s Test. The intelligence levels were defined by the IQ scores and classified as “retarded” (<69); “marginal” (70–79); “dull normal” (80–89); “normal” (90–109); “high normal” (110–119); “superior” (120–129) and “excellent” (>130). Only 11 children had retarded, marginal, or dull normal intelligence (IQ score <90); so, 683 children were finally included and were further separated into four groups (normal, high normal, superior, and excellent) according to their IQ scores in this study.
Genotyping of gene polymorphisms
Genomic DNA was extracted from whole-blood samples by a genomic DNA miniprep kit (LifeFeng Biotechnology, Shanghai, China). Four single-nucleotide polymorphism (SNP) loci with minor allele frequency >0.1 were retrieved from Haploview (www.broadinstitute.org/haploview/haploview). All of these loci polymorphisms have been reported to be related to neurodevelopmental defects.[31-33] rs11627387, rs1076991, rs2236224, and rs2236225 loci were located in intron 26, 2 KB upstream, intron 21, and exon 20, respectively, in MTHFD1. All polymorphisms of the four loci were genotyped through a custom-by-design 48-Plex SNPscanTM Kit (catalog number, G0104; Genesky Biotechnologies, Shanghai, China). The kit was developed according to the patented SNP genotyping technology of Genesky Biotechnologies, which was based on double ligation and multiplex fluorescence polymerase chain reaction (PCR). Briefly, DNA samples (100–200 ng) were denatured at 98°C for 5 min and then mixed with the premix containing ligase and the probe. The ligation reaction was carried out in a thermal cycler (ABI2720; Applied Biosystems, Foster City, CA, USA). Then, two fluorescent PCRs were undertaken for each ligation product. PCR products were separated and detected by capillary electrophoresis in a sequencer (ABI3730XL; Applied Biosystems). Genotyping was completed according to the obtained information for the labelingdye color and fragment size of allele-specific ligation PCR products. About 4% of the genotyping was done repeatedly, and the consistency rate was >96%.
A database was set up by Epidata 3.0 (Epidata Association, Odense, Denmark) in which two operators independently imported all data. Mean ± standard deviation and number (%) are presented for continuous and categorical variables, respectively.
Differences in continuous data between the two groups were compared using the Student’s t-test or Mann–Whitney U-test. The distribution of categorical variables was compared using the Chi-squared test. Potential confounders (children’s age, gender, BMI, age at which pregnancy occurred, gestational weeks, birth weight, birth modes, paternal and maternal education level) were chosen as adjustment variables based on existing literature and population characteristics of this study. The generalized linear model (GLM) was used to analyze the association between children’sUFCr level, MTHFD1 polymorphisms, and IQ scores. The multinomial logistic regression model was applied to analyze the relationship between children’s UFCr level, MTHFD1 polymorphisms, and intelligence levels. And normal intelligence children were reference in the analyses of the intelligence levels. For exploration of the relationship between the UFCr level and children’s intelligence, the UFCr level was separated into categorical variables according to tertiles, and the median of each segment was regarded as a continuous variable to estimate the linear trend of children’s intelligence. The general linear model was used to explore models of possible gene-environment and genegene interaction on intelligence. Data were processed using SPSS 21.0 (IBM, Armonk, NY, USA). Plots were drawn by GraphPad Prism 8.0.1. P< 0.05 was considered significant.
General characteristics of participants
Since 11 children had an IQ score < 90, a total of 683 eligible children aged 8 to 12 years were included in this study and were further classified as the CG (n= 342) and HFG (n= 341) according to the median of children’s UFCr level (1.33 mg/L). The distribution of children’s age was consistent in the CG and HFG (10.05 ± 1.24 and 10.08 ± 1.23 years, respectively). The concentration of UFCr in the HFG (2.15 ± 0.91 mg/L) was significantly higher than that in the CG (0.83 ± 0.30 mg/L) (P < 0.001), whereas the distribution of other sociodemographic characteristics (except UCr and UF levels) presented no significant differences between the two groups (P > 0.05 for all) [Table 1].
Table 1 – Demographic data of the study population.
Total (n = 683)
CG (n = 342)
HFG (n = 341)
10.07 ± 1.24
10.05 ± 1.24
10.08 ± 1.23
17.50 ± 2.96
17.68 ± 2.96
17.33 ± 2.95
1089 ± 607
1277 ± 606
901 ± 548
1.27 ± 0.79
0.98 ± 0.62
1.56 ± 0.82
1.49 ± 0.95
0.83 ± 0.30
2.15 ± 0.91
Age at which pregnancy occurred (years)
25.82 ± 4.28
25.95 ± 4.52
25.70 ± 4.01
Gestational weeks (weeks)
36.87 ± 4.84
36.53 ± 5.31
37.22 ± 4.30
Birth weight (kg)
3.34 ± 0.52
3.31 ± 0.55
3.36 ± 0.48
-Primary school and below
-High school and above
-Primary school and below
-High school and above
122.05 ± 11.88
121.50 ± 12.14
122.61 ± 11.61
Data are presented as mean ± SD or n (%).
†X2 test.BMI: Body mass index; CG: Control group; HFG: High fluoride group; IQ: Intelligence quotient; SD: Standard deviation; UCr: Urinary creatinine; UF: Urinary fluoride; UFCr: Urinary creatinine-adjusted urinary fluoride.
Association between children’s UFCr level and intelligence
The GLM and multinomial logistic regression model were employed to evaluate if there were associations between the UFCr level and children’s IQ scores or intelligence levels, respectively [Figure 1]. For each increase of 1.0 mg/L in the UFCr level, children’s IQ scores decreased by 2.502 (? = ?2.502, 95% confidence interval [CI]: ?4.411, ?0.593, P= 0.010), and the possibility of developing “ excellent” intelligence decreased by 46.3% with reference to the normal intelligence children in the HFG (odds ratio [OR] = 0.537, 95% CI: 0.290, 0.994, P = 0.048). After stratifying children according to the tertiles of children’s UFCr concentration in different groups, the trend test showed no significance (Ptrend > 0.05 for all).
Association between MTHFD1 polymorphisms and intelligence
The genotype distributions of rs11627387, rs1076991, rs2236224, and rs2236225 loci in MTHFD1 were in accordance with the Hardy–Weinberg equilibrium (P > 0.05 for all loci) [Supplementary Table 1, https://links.lww.com/CM9/A987], which indicated that the investigated participants were representative of the population. We also estimated the difference in the distribution of genotypes/allele of rs11627387, rs1076991, rs2236224, and rs2236225 between the CG and HFG, but significant differences were not found (P > 0.05 for all) [Supplementary Table 2, https://links.lww.com/CM9/A987].
Furthermore, the relationship between MTHFD1 polymorphisms and intelligence was assessed by the GLM and multinomial logistic regression model. In the general population and HFG, children with the GG genotype of rs11627387 showed increased IQ score relative to those with the AA genotype of rs11627387 (? = 3.574, 95% CI: 0.274, 6.874, P= 0.034 for the general population, ? = 4.723, 95% CI: 0.277, 9.168, P= 0.037 for the HFG) [Figure 2]. There was an increment in IQ score in children carrying the G allele than in those with the A allele of rs11627387 in the HFG, and the association was borderline significant (P = 0.059) [Figure 2]. In addition, the possibility of developing “high normal” intelligence was lower in the HFG in children with the AG genotype when compared with those carrying the AAgeno type of rs11627387 (OR = 0.212, 95% CI: 0.045,0.997, P= 0.049) [Figure 3]. With respect to rs2236225 locus, participants with the AA genotype seemingly showed a lower possibility of developing “high normal” intelligence when compared with children with the GG genotype in the HFG, and the association was borderline significant (P = 0.056) [Figure 3]. However, a statistical significance was not found in the association between MTHFD1 polymorphism and children’s intelligence levels in the total group and CG (P > 0.050) [Supplementary Tables 3 and 4, https://links.lww.com/CM9/A987].
Effects of gene-gene and gene-environment interaction on children’s intelligence
The general linear model was used to explore gene environment and gene-gene interactions. Loci rs11627387, rs1076991, and rs2236225 may have interactive effects on the IQ scores according to analyses of gene-gene interaction [Figure 4], where the association showed marginal significance (F = 1.726, P= 0.059). In addition, the interaction between rs11627387, rs1076991, rs2236224, and the UFCr level might affect children’s IQ scores (F = 1.669, P = 0.021), as well as the interactive effects of loci rs11627387, rs1076991, rs2236225, and the UFCr level (F = 1.764, P= 0.012). An effect of the interaction between these four loci and the UFCr level on the IQ scores of children was also found (F = 1.614, P= 0.012). However, statistical significances were not found in other models [Supplementary Table 5, https://links.lww.com/CM9/A987].
We evaluated the interaction between fluoride exposure and MTHFD1 polymorphisms on the intelligence of children living in areas of endemic fluorosis. We found that the decline in the IQ scores in children was associated with excessive exposure to fluoride and that changes in children’s intelligence might be modified by rs11627387 locus polymorphisms of MTHFD1 to some extent. Moreover, the interactive effects of the four loci of MTHFD1 and fluoride exposure on children’s IQ scores showed different models in our study.
The health and function of the nervous system can be affected by exposure to various environmental factors.[40,41]
In terms of fluoride, drinking water containing a high concentration of fluoride is the main way to the exposure of superabundant fluoride. Fluoride entering the body can be distributed widely throughout the body after absorption, and most of it is deposited in bone and teeth. The absorption rate of fluoride in children is about 80% to 90%, much higher than that in adults. In addition, fluoride can penetrate the blood-brain barrier into brain tissue and seems to accumulate in the areas of the brain responsible for learning and memory functions.[45,46] But the evidence regarding a link between fluoride exposure and intelligence impairment is not definitive. Most of the scholars conceive that fluoride in brain tissue can damage nerve functions and even lead to intellectual loss.[47-49] For example, animal studies found that, upon exposure to increasing concentrations of sodium fluoride in drinking water, the learning abilities of mice were impaired.[10,50] However, the results of population-based epidemiological studies are not completely consistent. A cohort study from Canada revealed a positive association between fluoride exposure during pregnancy and intelligence decline in offspring. Our current study support this finding, that is, an inverse association between excessive exposure to fluoride and children’s IQ scores in the HFG, which recollects that excessive fluoride exposure may have negative influence on normal development of children’s intelligence. Whereas, a prospective study conducted in New Zealand did not find a positive association between lower intelligence and a higher fluoride level. A cross-sectional study conducted in China did not observe the significant difference in IQ scores between children in fluorosis areas and children in the CG, either. Differences in study design, levels of fluoride exposure, exposure patterns, ethnicity, and assessment methods can explain these inconsistences to a certain extent. On the other hand, genetic susceptibility (eg, genetic polymorphisms) is also one of the important reasons that is worthy of further discussion.
We further discussed the association between MTHFD1 polymorphisms and children’s intelligence according the study design. We found that, compared with children carrying the AA genotype of the rs11627387 locus, children carrying the GG genotype might have increased IQ score in the HFG, whereas the possibility for having “high normal” intelligence was lower for children carrying the AG genotype. It can be suggested that polymorphisms of the rs11627387 locus of MTHFD1 may have effects on changes in children’s intelligence. Specifically, the GG genotype of rs11627387 may retard the intelligence decline caused by excessive fluoride exposure in schoolage children, as carrying the GG genotype of rs11627387 may exert a positive effect on IQ scores compared to carrying the AA genotype when UFCr >1.33 mg/L in children. Although few studies have focused on the relationship between MTHFD1 polymorphisms and intelligence, others are still evaluating the effects of MTHFD1 polymorphisms on neural development based on animals and humans. An animal study reported that mice with loss of one allele of MTHFD1 via a gene-trap mutation showed impaired learning ability. Another study revealed that reduced gene expression of the alpha seven nicotinic cholinergic receptor was observed in a mouse model which simulated polymorphisms at the rs2236225 locus of MTHFD1 in humans. Also, agonists of the alpha seven nicotinic cholinergic receptor gene could be used to treat neurocognitive dysfunction in schizophrenia. In addition, population-based studies also pointed out that polymorphisms of rs11627387, rs1076991, rs2236224, and rs2236225 loci are involved in neural tube defects.[31,33] These studies demonstrate that MTHFD1 polymorphisms can have effects on normal neurocognitive functions and may modify the susceptibility of neurological diseases. Overall, our results support the correlation between MTHFD1 polymorphisms and intellectual changes in school-age children, which is worthy of confirmation in further studies.
Intelligence is not only modified by genetic factors such as gene polymorphisms but also related to gene-environment interaction. For example, Zhao et al conducted a crosssectional study in Tianjin and reported that ankyrin repeat and kinase domain 1 (ANKK1), catechol-O-methyltransferase (COMT), monoamine oxidase A (MAOA) gene polymorphisms may have interactive effects with UF on children’s intelligence. In our study, the effects of gene-environment and gene-gene interaction on children’s IQ were explored. We did not observe an interaction between a single locus of MTHFD1 and fluoride exposure on children’s intelligence but found the different models of interaction between multiple loci of MTHFD1 and fluoride exposure. These results suggest that phenotypes can be affected by the interaction of environmental factors with multiple loci of one gene and even multiple genes; the modification effect of one locus polymorphism is minor. All these findings suggest that MTHFD1 polymorphisms may be involved in the effects of fluoride exposure on intelligence in school-age children, and we further provide a novel clue for the study of neurotoxicological mechanisms of fluoride.
There are several advantages in this study. First, the study was a population-based epidemiological study which explored the impact of fluoride and gene polymorphisms on children’s intelligence, where the results suggest that people living in fluorosis areas should pay attention to the harmful effects of excessive fluoride intake on school-age children’s intelligence. Second, this study was conducted in the middle of the semester which avoided the impact of students’ psychological stress caused by the beginning of the semester or the final exam on intelligence tests. Third, as described in the study design in our previous publication, all the children lived on campus and they had similar living conditions, living habits, and dietary structure, which minimized the bias. Finally, the investigated areas in Tongxu County, Henan Province, are relatively underdeveloped, with no industrial fluorine and other pollutants such as lead and mercury, etc, that may affect intelligence.
Our study also had some limitations. First, we did not adjust the children’s diet, although children included in the study were all boarding school students, and the dietary structure was relatively consistent; the differences in dietary habits and physical fitness of children may lead to differences in nutrients intake, which may further affect mental development. Therefore, children’s diet will be considered in our subsequent studies. Second, the study is a cross-sectional study with only one sampling; so, the causal relationship is weak. So, long-term large-scale epidemiological or cohort studies should be conducted to provide more evidence. Third, although we adjusted for confounding factors such as children’s age, gender, BMI, maternal age at which pregnancy occurred, gestational weeks, birth weight, birth modes, and paternal and maternal education level, there may still be some other confounding factors (such as folic acid). However, bounded to the eugenics policy, almost all the mothers of the investigated students had taken folic acid in the early stage of pregnancy ; so, the bias may be reduced. Studies involving larger populations and more questionnaire information are in progress.
We evaluated the effects of polymorphisms of MTHFD1 and fluoride exposure on children’s intelligence in endemic fluorosis areas. Excessive fluoride exposure may have adverse effects on children’s intelligence, and changes in children’s intelligence may be associated with the interaction between fluoride and MTHFD1 polymorphisms.
The authors would like to express their sincere thanks to all the staff and volunteers involved in the study. The authors also thank Francis-Kojo Afrim for his English editing help.
This study was supported by the National Natural Science Foundation of China (Nos. 81972981, 82003401, and 81673116) and Key Projects of Colleges and Universities of Henan Education Department (21A330006).
Conflicts of interest
1. Whelton HP, Spencer AJ, Do LG, Rugg-Gunn AJ. Fluoride revolution and dental caries: evolution of policies for global use. J Dent Res 2019;98:837–846. doi: 10.1177/0022034519843495.
2. Riedel C, Zimmermann EA, Zustin J, Niecke M, Amling M, Grynpas M, et al. The incorporation of fluoride and strontium in hydroxyapatite affects the composition, structure, and mechanical properties of human cortical bone. J Biomed Mater Res A 2017;105:433–442. doi: 10.1002/jbm.a.35917.
3. Guth S, Huser S, Roth A, Degen G, Diel P, Edlund K, et al. Toxicity of fluoride: critical evaluation of evidence for human developmental neurotoxicity in epidemiological studies, animal experiments and in vitro analyses. Arch Toxicol 2020;94:1375–1415. doi: 10.1007/s00204-020-02725-2.
3. Guth S, Huser S, Roth A, Degen G, Diel P, Edlund K, et al. Toxicity of fluoride: critical evaluation of evidence for human developmental neurotoxicity in epidemiological studies, animal experiments and in vitro analyses. Arch Toxicol 2020;94:1375–1415. doi: 10.1007/s00204-020-02725-2.
5. Yuan L, Fei W, Jia F, Jun-Ping L, Qi L, Fang-Ru N, et al. Health risk in children to fluoride exposure in a typical endemic fluorosis area on Loess Plateau, North China, in the last decade. Chemosphere 2020;243:125451. doi: 10.1016/j.chemosphere.2019.125451.
8. Ma Q, Huang H, Sun L, Zhou T, Zhu J, Cheng X, et al. Geneenvironment interaction: does fluoride influence the reproductive hormones in male farmers modified by ERalpha gene polymorphisms? Chemosphere 2017;188:525–531. doi: 10.1016/j.chemosphere.2017.08.166.
10. Yuan J, Li Q, Niu R, Wang J. Fluoride exposure decreased learning ability and the expressions of the insulin receptor in male mouse hippocampus and olfactory bulb. Chemosphere 2019;224:71–76. doi: 10.1016/j.chemosphere.2019.02.113.
11. Xin J, Wang H, Sun N, Bughio S, Zeng D, Li L, et al. Probiotic alleviate fluoride-induced memory impairment by reconstructing gut microbiota in mice. Ecotoxicol Environ Saf 2021;215:112108. doi: 10.1016/j.ecoenv.2021.112108.
12. Till C, Green R, Flora D, Hornung R, Martinez-Mier EA, Blazer M, et al. Fluoride exposure from infant formula and child IQ in a Canadian birth cohort. Environ Int 2020;134:105315. doi: 10.1016/j.envint.2019.105315.
13. Wang M, Liu L, Li H, Li Y, Liu H, Hou C, et al. Thyroid function, intelligence, and low-moderate fluoride exposure among Chinese school-age children. Environ Int 2020;134:105229. doi: 10.1016/j.envint.2019.105229.
14. Xu K, An N, Huang H, Duan L, Ma J, Ding J, et al. Fluoride exposure and intelligence in school-age children: evidence from different windows of exposure susceptibility. BMC Public Health 2020;20:1657. doi: 10.1186/s12889-020-09765-4.
15. Ge Y, Chen L, Yin Z, Song X, Ruan T, Hua L, et al. Fluoride-induced alterations of synapse-related proteins in the cerebral cortex of ICR offspring mouse brain. Chemosphere 2018;201:874–883. doi: 10.1016/j.chemosphere.2018.02.167.
16. Broadbent JM, Thomson WM, Ramrakha S, Moffitt TE, Zeng J, Page LAF, et al. Community water fluoridation and intelligence: prospective study in New Zealand. Am J Public Health 2015;105:72–76. doi: 10.2105/AJPH.2013.301857.
17. Li FH, Chen X, Huang RJ, Xie YP. Intelligence impact of children with endemic fluorosis caused by fluoride from coal burning (in Chinese). J Environ Health 2009;26:338–340. doi: 10.16241/j.cnki.1001-5914.2009.04.040.
18. Kang X, Guo T, Liu L, Ding SZ, Lei C, Luo H. Association between PTCH1 gene polymorphisms and chronic obstructive pulmonary disease susceptibility in a Chinese Han population: a case-control study. Chin Med J 2020;133:2071–2077. doi: 10.1097/CM9.0000000000000858.
19. Wen D, Zhou XL, Du X, Dong JZ, Ma CS. Association of interleukin-18 gene polymorphisms with Takayasu arteritis in a Chinese Han population. Chin Med J 2020;133:2315–2320. doi: 10.1097/CM9.0000000000001047.
21. Chu Y, Liu Y, Guo N, Lou Q, Wang L, Huang W, et al. Association between ALOX15 gene polymorphism and brick-tea type skeletal fluorosis in Tibetans, Kazaks and Han, China. Int J Environ Health Res 2021;31:421–432. doi: 10.1080/09603123.2019.1666972.
22. Zhang S, Zhang X, Liu H, Qu W, Guan Z, Zeng Q, et al. Modifying effect of COMT gene polymorphism and a predictive role for proteomics analysis in children’s intelligence in endemic fluorosis area in Tianjin, China. Toxicol Sci 2015;144:238–245. doi: 10.1093/toxsci/kfu311.
23. Cui Y, Zhang B, Ma J, Wang Y, Zhao L, Hou C, et al. Dopamine receptor D2 gene polymorphism, urine fluoride, and intelligence impairment of children in China: a school-based cross-sectional study. Ecotoxicol Environ Saf 2018;165:270–277. doi: 10.1016/j. ecoenv.2018.09.018.
26. Field MS, Shields KS, Abarinov EV, Malysheva OV, Allen RH, Stabler SP, et al. Reduced MTHFD1 activity in male mice perturbs folate- and choline-dependent one-carbon metabolism as well as transsulfuration. J Nutr 2013;143:41–45. doi: 10.3945/jn.112.169821.
27. MacFarlane AJ, Perry CA, Girnary HH, Gao D, Allen RH, Stabler SP, et al. Mthfd1 is an essential gene in mice and alters biomarkers of impaired one-carbon metabolism. J Biol Chem 2009;284:1533–1539. doi: 10.1074/jbc.M808281200.
29. Dayon L, Guiraud SP, Corthesy J, Da Silva L, Migliavacca E, Tautvydaite D, et al. One-carbon metabolism, cognitive impairment and CSF measures of Alzheimer pathology: homocysteine and beyond. Alzheimers Res Ther 2017;9:43. doi: 10.1186/s13195-017-0270-x.
30. Wu J, Bao Y, Lu X, Wu L, Zhang T, Guo J, et al. Polymorphisms in MTHFD1 gene and susceptibility to neural tube defects: a casecontrol study in a Chinese Han population with relatively low folate levels. Med Sci Monit 2015;21:2630–2637. doi: 10.12659/MSM.895155.
31. Etheredge AJ, Finnell RH, Carmichael SL, Lammer EJ, Zhu H, Mitchell LE, et al. Maternal and infant gene-folate interactions and the risk of neural tube defects. Am J Med Genet A 2012;158A:2439–2446. doi: 10.1002/ajmg.a.35552.
33. Carroll N, Pangilinan F, Molloy AM, Troendle J, Mills JL, Kirke PN, et al. Analysis of the MTHFD1 promoter and risk of neural tube defects. Hum Genet 2009;125:247–256. doi: 10.1007/s00439-008-0616-3.
34. Liu AY, Scherer D, Poole E, Potter JD, Curtin K, Makar K, et al. Gene-diet-interactions in folate-mediated one-carbon metabolism modify colon cancer risk. Mol Nutr Food Res 2013;57:721–734. doi: 10.1002/mnfr.201200180.
35. Shaw GM, Lu W, Zhu H, Yang W, Briggs FBS, Carmichael SL, et al. 118 SNPs of folate-related genes and risks of spina bifida and conotruncal heart defects. BMC Med Genet 2009;10:49. doi: 10.1186/1471-2350-10-49.
36. Zhu H, Yang W, Lu W, Etheredge AJ, Lammer EJ, Finnell RH, et al. Gene variants in the folate-mediated one-carbon metabolism (FOCM) pathway as risk factors for conotruncal heart defects. Am J Med Genet A 2012;158A:1124–1134. doi: 10.1002/ajmg.a.35313.
37. Thomas DB, Basu N, Martinez-Mier EA, Sanchez BN, Zhang Z, Liu Y, et al. Urinary and plasma fluoride levels in pregnant women from Mexico City. Environ Res 2016;150:489–495. doi: 10.1016/j. envres.2016.06.046.
38. Liu HL, Lam LT, Zeng Q, Han SQ, Fu G, Hou CC. Effects of drinking water with high iodine concentration on the intelligence of children in Tianjin, China. JPublic Health (Oxf) 2009;31:32–38. doi: 10.1093/pubmed/fdn097.
39. Chen X, Li S, Yang Y, Yang X, Liu Y, Liu Y, et al. Genome-wide association study validation identifies novel loci for atherosclerotic cardiovascular disease. J Thromb Haemost 2012;10:1508–1514. doi: 10.1111/j.1538-7836.2012.04815.x.
40. Liu WT, Yan XX, Cheng DZ, Zhang HZ, Ding N, Xu KM, et al. Oxcarbazepine monotherapy in children with benign epilepsy with centrotemporal spikes improves quality of life. Chin Med J 2020;133:1649–1654. doi: 10.1097/CM9.0000000000000925.
41. Tian DY, Wang J, Sun BL, Wang Z, Xu W, Chen Y, et al. Spicy food consumption is associated with cognition and cerebrospinal fluid biomarkers of Alzheimer disease. Chin Med J 2020;134:173–177. doi: 10.1097/CM9.0000000000001318.
42. Li P, Qian H, Wu J, Chen J, Zhang Y, Zhang H. Occurrence and hydrogeochemistry of fluoride in alluvial aquifer of Weihe River, China. EnvironEarth Sci 2013;71:3133–3145. doi: 10.1007/s12665-013-2691-6.
43. Chakraborti D, Rahman MM, Chatterjee A, Das D, Das B, Nayak B, et al. Fate of over 480 million inhabitants living in arsenic and fluoride endemic Indian districts: magnitude, health, socio-economic effects and mitigation approaches. J Trace Elem Med Biol 2016;38:33–45. doi: 10.1016/j.jtemb.2016.05.001.
46. Pereira M, Dombrowski PA, Losso EM, Chioca LR, Da Cunha C, Andreatini R. Memory impairment induced by sodium fluoride is associated with changes in brain monoamine levels. Neurotox Res 2011;19:55–62. doi: 10.1007/s12640-009-9139-5.
47. Zhou G, Tang S, Yang L, Niu Q, Chen J, Xia T, et al. Effectsoflong-term fluoride exposure on cognitive ability and the underlying mechanisms: role of autophagy and its association with apoptosis. Toxicol Appl Pharmacol 2019;378:114608. doi: 10.1016/j.taap.2019.114608.
48. Ding Y, Gao Y, Sun H, Han H, Wang W, Ji X, et al. The relationships between low levels of urine fluoride on children’s intelligence, dental fluorosis in endemic fluorosis areas in Hulunbuir, Inner Mongolia, China. J Hazard Mater 2011;186:1942–1946. doi: 10.1016/j.jhazmat.2010.12.097.
49. Sun Z, Zhang Y, Xue X, Niu R, Wang J. Maternal fluoride exposure during gestation and lactation decreased learning and memory ability, and glutamate receptor mRNA expressions of mouse pups. Hum Exp Toxicol 2018;37:87–93. doi: 10.1177/0960327117693067.
50. Wang J, Zhang Y, Guo Z, Li R, Xue X, Sun Z, et al. Effects of perinatal fluoride exposure on the expressions of miR-124 and miR-132 in hippocampus of mouse pups. Chemosphere 2018;197:117–122. doi: 10.1016/j.chemosphere.2018.01.029.
51. Green R, Lanphear B, Hornung R, Flora D, Martinez-Mier EA, Neufeld R, et al. Association between maternal fluoride exposure during pregnancy and IQ scores in offspring in Canada. JAMA Pediatr 2019;173:940–948. doi: 10.1001/jamapediatrics.2019.1729.
55. Zhao L, Yu C, Lv J, Cui Y, Wang Y, Hou C, et al. Fluoride exposure, dopamine relative gene polymorphism and intelligence: a cross-sectional study in China. Ecotoxicol Environ Saf 2020;209:111826. doi: 10.1016/j.ecoenv.2020.111826.
57. Wang A, Duan L, Huang H, Ma J, Zhang Y, Ma Q, et al. Association between fluoride exposure and behavioural outcomes of school-age children: a pilot study in China. Int J Environ Health Res 2020;1–10. doi: 10.1080/09603123.2020.1747601.