Excerpts: Our study has several strengths. So far as we know, this is the first epidemiological study to uncover the effect modification of DAT1 and MAOA gene polymorphism in the relationship between fluoride exposure and IQ, and the first study to analyze the high-dimensional interaction among fluoride exposure and the four DA relative genes. Our findings suggest a novel clue for the neuro-toxicological mechanism of fluoride. Our study also has some limitations. We used urinary fluoride to re

Abstract

Highlights

  • Urine fluoride was inversely associated with IQ.
  • IQ of children with high-activity MAOA genotype was lower than IQ of those with low-activity or female heterozygote genotype.
  • DAT1 and MAOA gene polymorphism modify the effects of UF on IQ.
  • UF, ANKK1, COMT and MAOA have a high-dimensional interaction on IQ.

Background

Excessive fluoride exposure is related to adverse health outcomes, but whether dopamine (DA) relative genes are involved in the health effect of low-moderate fluoride exposure on children’s intelligence remain unclear.

Objectives

We conducted a cross-sectional study to explore the role of DA relative genes in the health effect of low-moderate fluoride exposure in drinking water.

Methods

We recruited 567 resident children, aged 6–11 years old, randomly from endemic and non-endemic fluorosis areas in Tianjin, China. Spot urine samples were tested for urinary fluoride concentration, combined Raven`s test was used for intelligence quotient test. Fasting venous blood were collected to analyze ANKK1 Taq1A (rs1800497), COMT Val158Met (rs4680), DAT1 40 bp VNTR and MAOA uVNTR. Multivariable linear regression models were used to assess associations between fluoride exposure and IQ scores. We applied multiplicative and additive models to appraise single gene-environment interaction. Generalized multifactor dimensionality reduction (GMDR) was used to evaluate high-dimensional interactions of gene-gene and gene-environment.

Results

In adjusted model, fluoride exposure was inversely associated with IQ scores (B = -5.957, 95% CI: -9.712, -2.202). The mean IQ scores of children with high-activity MAOA genotype was significantly lower than IQ scores of those with low-activity (P = 0.006) or female heterozygote (P = 0.016) genotype. We detected effect modification by four DA relative genes (ANKK1, COMT, DAT1 and MAOA) on the association between UF and IQ scores. We also found a high-dimensional gene-environment interaction among UF, ANKK1, COMT and MAOA on the effect of IQ (testing balanced accuracy = 0.5302, CV consistency: 10/10, P = 0.0107).

Conclusions

Our study suggests DA relative genes may modify the association between fluoride and intelligence, and a potential interaction among fluoride exposure and DA relative genes on IQ.


*Original abstract online at https://www.sciencedirect.com/science/article/pii/S0147651320316626

Excerpts:

Our study has several strengths. So far as we know, this is the first epidemiological study to uncover the effect modification of DAT1 and MAOA gene polymorphism in the relationship between fluoride exposure and IQ, and the first study to analyze the high-dimensional interaction among fluoride exposure and the four DA relative genes. Our findings suggest a novel clue for the neuro-toxicological mechanism of fluoride.

Our study also has some limitations. We used urinary fluoride to represent the fluoride exposure from drinking water, which ignored the bias of short-term fluoride intake from diet. However, our finding could still reveal the health effects of internal fluoride exposure. Due to the cross-sectional design, our results are not strong for causal demonstration. Small sample size also limited the statistical power. Further prospective studies with larger sample size are essential to validate our findings. Nevertheless, the effect modification of DA relative genes and interaction among UF and these genes are still noteworthy.


Appendix A. Supplementary material

HelpDownload : Download Word document (16KB)


https://ars.els-cdn.com/content/image/1-s2.0-S0147651320316626-gr1_lrg.jpg


Table 1. Basic characteristics of study population.

Characteristics Mean ±?SD or n (%)
Sample size 567
Age (years old) 9.15?±?1.02
Gender
 Boys 284 (50.1)
 Girls 283 (49.9)
Body mass index (kg/m2) 17.76?±?3.55
Paternal education level
 <high school 162 (28.6)
 ?high school 405 (71.4)
Maternal education level
 <?high school 177 (31.2)
 ??high school 390 (68.8)
Household income (RMB/year)
 ??30,000 135 (23.8)
 30,000–100,000 303 (53.4)
 >?100,000 129 (22.8)
Maternal age at delivery (years old)
 ??30 365 (64.4)
 >?30 202 (35.6)
Abnormal birth
 no 461 (81.3)
 yes 106 (18.7)
Log_UF 0.015 (0.252)
IQ scores 112.17 (11.75)