Polymorphism of Estrogen Receptor Genes and Its Interactions With Neurodevelopmental Genes in Attention Deficit Hyperactivity Disorder Among Chinese Han Descent

Yiwei Lin; Haimei Li; Jing Zhang; Ziqi Yang; Yi Zhou; Lu Liu; Qiujin Qian

doi:10.30773/pi.2023.0113

Psychiatry Investig > Volume 20(8); 2023 > Article

Lin, Li, Zhang, Yang, Zhou, Liu, and Qian: Polymorphism of Estrogen Receptor Genes and Its Interactions With Neurodevelopmental Genes in Attention Deficit Hyperactivity Disorder Among Chinese Han Descent

Original Article

Psychiatry Investigation 2023;20(8):775-785.

Published online: August 21, 2023

DOI: https://doi.org/10.30773/pi.2023.0113

Polymorphism of Estrogen Receptor Genes and Its Interactions With Neurodevelopmental Genes in Attention Deficit Hyperactivity Disorder Among Chinese Han Descent

Yiwei Lin^1,^2,³

, Haimei Li^1,²

, Jing Zhang^1,²

, Ziqi Yang^1,²

, Yi Zhou^1,²

, Lu Liu^1,²

, Qiujin Qian^1,²

¹Peking University Sixth Hospital, Peking University Institute of Mental Health, Beijing, China

²National Clinical Research Center for Mental Disorders & The Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China

³Peking University Health Science Center, Beijing, China

Correspondence: Lu Liu, PhD Peking University Sixth Hospital, Peking University Institute of Mental Health, 51 Huayuan North Road, Haidian District, Beijing 100191, China Tel: +86 010 82802907, Fax: +86 010 62070258, E-mail: liulupku@bjmu.edu.cn

Correspondence: Qiujin Qian, PhD Peking University Sixth Hospital, Peking University Institute of Mental Health, 51 Huayuan North Road, Haidian District, Beijing 100191, China Tel: +86 010 82802907, Fax: +86 010 62070258 E-mail: qianqiujin@bjmu.edu.cn

Received April 21, 2023 Revised May 29, 2023 Accepted June 7, 2023

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Objective

Attention deficit hyperactivity disorder (ADHD) is a polygenic neurodevelopmental disorder with significant gender differences. The sexual dimorphism of ADHD may be associated with estrogen acting through estrogen receptors (ESR). This study investigates the impact of ESR gene polymorphism and its interactions with neurodevelopmental genes on ADHD susceptibility.

Methods

The study compared genotyping data of single nucleotide polymorphisms in ESR1 and ESR2 in 1,035 ADHD cases and 962 controls. The gene-gene interactions between ESR genes and three neurodevelopmental genes (brain-derived neurotrophic factor [BDNF], synaptosomal-associated protein of 25 kDa gene [SNAP25], and cadherin-13 [CDH13]) in ADHD were investigated using generalized multifactor dimensionality reduction and verified by logistic regression analysis.

Results

The G allele of rs960070/ESR2 (empirical p=0.0076) and the A allele of rs8017441/ESR2 (empirical p=0.0426) were found significantly higher in ADHD cases than in the controls but not in male or female subgroups. Though no difference was found in all subjects or females, the A allele of rs9340817/ESR1 (empirical p=0.0344) was found significantly higher in ADHD cases than controls in males. We also found genetic interaction models between ESR2 gene, neurodevelopmental genes and ADHD susceptibility in males (ESR2 rs960070/BDNF rs6265/BDNF rs2049046/SNAP25 rs362987/CDH13 rs6565113) and females (ESR2 rs960070/BDNF rs6265/BDNF rs2049046) separately, though it was negative in overall subjects.

Conclusion

The ESR gene polymorphism associates with ADHD among Chinese Han children, with interactions between ESR genes and neurodevelopmental genes potentially influencing the susceptibility of ADHD.

Keywords: Attention deficit hyperactivity disorder, Estrogen receptor, Brain-derived neurotrophic factor, Single nucleotide polymorphism, Gene-gene interactions

INTRODUCTION

Attention deficit hyperactivity disorder (ADHD) is a common childhood neurodevelopmental disorder, affecting 5%- 11% of children [1,2]. Heritability estimates for the disorder are Print ISSN 1738-3684 / On-line ISSN 1976-3026 OPEN ACCESS approximately 74% [3], indicating a strong genetic component. Previous studies have further demonstrated that ADHD is polygenic, influenced by the combined effects of multiple genes, each with a minor influence [4]. Simultaneously, ADHD significantly varies across sex. Epidemiological studies have observed an obvious gender dimorphism in ADHD, with a male to female ratio of approximately 3-4:1 [5]. Significant sex differences in clinical symptoms, brain structure and function have also been reported in numerous studies of ADHD patients [6]. Researchers have previously explained the gender differences in ADHD from three aspects: biological factors, sociocultural factors, and gender bias in diagnostic. Research on the genetic mechanism of sex dimorphism in ADHD has mainly focused on X or Y chromosome genes, with comparatively less attention given to genes related to sex hormones.

Estrogen, which plays an important role in establishing neural and behavioral sex differences, has been suggested to be related to the sexual dimorphism of ADHD by some scholars [7,8]. It has been reported that estrogen can stimulate the growth of spinous cells and synapse, while also increase the expression of brain-derived neurotrophic factor (BDNF) [9,10]. The biological effects of estrogen are primarily mediated by the estrogen receptors (ESRs), including ESR1 and ESR2. ESRs possess neuroprotective and anti-inflammatory properties which may underlie their observed psychotropic effects [11]. Previous studies have indicated that ESRs can activate the PI3K/Akt neuroprotective signaling pathway in the brain and reduce neuronal apoptosis [12,13]. Polymorphism of ESR genes has been reported to be associated with a variety of psychiatric disorders. The single nucleotide polymorphisms (SNPs) rs2234693 and rs9340799 in ESR1 were found to be associated with the pathogenesis of schizophrenia [14,15]. ESR1 rs22346939 and rs9340799 polymorphisms were also found to be related to the risk of major depressive disorder [16]. In autism, rs1155819 in ESR1 and rs1152582 in ESR2 were reported to be associated with severities in the impairment of social interaction and emotional regulation [17]. Since ADHD shares similar phenotypes and underlying genetic basis with these disorders, polymorphisms of ESR genes may also play a role in the pathology of ADHD. However, few studies have explored the impact of ESR gene polymorphisms on ADHD.

Neurodevelopmental genes have been widely investigated in ADHD [18,19]. Among these, BDNF, synaptosomal-associated protein of 25 kDa gene (SNAP25), and cadherin-13 (CDH13) show sexual dimorphism and are closely related to ADHD [20,21]. BDNF promotes neuronal growth and differentiation, participating in neuronal plasticity [22]. It contains an estrogen-like response element that can be modulated by estrogen through activating the ESR, regulating its expression and thereby regulating neural activity [23]. Our previous study in 325 Han Chinese samples revealed a sex-specific association between BDNF Val66Met (rs6265) and ADHD [24]. In mice experiments, an ESR2 agonist significantly increased BDNF-TrkB signaling pathways and downstream cascades related to synaptic plasticity [25]. SNAP25, implicated in neurotransmission and neuroplasticity, was also found as an ADHD susceptibility gene. Feng et al. [26] reported that SNAP25 rs362987 polymorphisms significantly increase ADHD risk. Sex differences in altered stress responses were reported in SNAP25 mutant mice [27]. Similarly, some CDH13 SNPs, which negative regulator of axon growth during neural differentiation, were revealed to contribute ADHD pathophysiology, particularly rs6565113 [28]. CDH13 mutant mice exposed to early-life stress showed sex differences, with increased locomotor activity in females but unchanged impulsivity and activity in males [29]. It remains unclear whether the observed sexual dimorphism in the above analyses also involves the ESR genes.

The current study aims to explore whether ESR gene polymorphisms are associated with ADHD and investigate whether genetic interactions among ESR genes and the neurodevelopmental genes (BDNF, SNAP25, and CDH13) contribute to ADHD susceptibility.

METHODS

Subjects

A total of 1,035 children with ADHD were recruited from the child psychiatry clinic at Peking University Sixth Hospital, including 162 females and 873 males. ADHD diagnosis was made according to Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria using the Clinical Diagnostic Interview Scale based on a semi-structured interview conducted by experienced child psychiatrists. The inclusion criteria were: 1) diagnosed as ADHD according to DSMIV, 2) age between 6 and 16 years, 3) full-scale intelligence quotient ≥70, 4) both biological parents were of Chinese Han descent, and 5) never used any psychiatric medication. Children with previous or present childhood schizophrenia, bipolar disorder, autism, intellectual disability, epilepsy, or other mental illness were excluded. Patients with obvious somatic or neurologic abnormalities, consciousness disorders, and current dependence or abuse of substance were also excluded. A total of 962 control samples of Chinese Han descent were recruited, including 355 females and 607 males. Controls included both children and adults and were recruited from three sources: local elementary school students, healthy blood donors attending the Blood Center of Peking University First Hospital, and healthy volunteers from our institute. The parents or adults completed the ADHD Rating Scale IV to exclude ADHD. Major psychiatric disorders, family history of psychosis, severe physical diseases, and substance abuse were also excluded [30].

This study adhered to the guidelines of the Declaration of Helsinki and was approved by the Ethics Review Committee of Peking University Sixth Hospital. All subjects or parents/guardians of children signed an informed consent form to participate. The Institutional Review Board (IRB) number is PKUH6 (2007)-(39).

Genotyping and SNPs selection

Genomic DNA was extracted from peripheral blood samples of cases and controls using standard protocols. SNP genotyping was performed using the Affymetrix6.0 array at CapitalBio Ltd. (Beijing, China) and following the Affymetrix standard protocol. A total of 131 SNPs in ESR1 and 26 SNPs in ESR2 were extracted. Haploview software version 4.2 (http://www.broad.mit.edu/mpg/haploview/) was used to select SNPs with minor allele frequency (MAF) >5% and meeting the Hardy-Weinberg equilibrium (HWE), yielding 97 ESR1 SNPs and 21 ESR2 SNPs. Linkage disequilibrium was identified using the confidence interval (CI) method in Haploview, and tag SNPs with a threshold of r²>0.80 were selected for subsequent analysis. Finally, 20 SNPs (ESR1: 15 SNPs; ESR2: 5 SNPs) were chosen as tag SNPs (Table 1).

SNPs for neurodevelopmental genes were selected based primarily on the previous studies. The BDNF SNPs, rs6265 [24,31], rs2030324 [32], and rs7934165 [33] have been studied in ADHD. The SNAP25 SNP rs362987 [26] and CDH13 SNP rs6565113 [34] were also associated with ADHD. Table 1 demonstrated the HWE test results and MAF information for the SNPs of ESR genes, BDNF gene, SNAP25 gene, and CDH13 gene that included in the final analysis. All the SNPs were with a HWE p-value and MAF over 0.05.

Statistical analysis

Statistical analysis of clinical data

Descriptive statistics were performed on demographic and clinical variables. Continuous variables with a normal distribution were expressed as mean±standard deviation, skewed continuous variables as median (interquartile range). Student’s t-test was used for two-group comparison, and the Mann-Whitney U test was used when the data did not conform to normal distribution. Categorical variables were expressed as numbers (percentages) and compared with the chi-square test. A p<0.05 was considered statistically significant. This part of the analysis was done using IBM SPSS 21 (IBM Corp., Armonk, NY, USA).

Case/control association test for SNPs

We used Haploview version 4.2 to perform HWE test and MAF of SNPs to ensure data quality. HWE was calculated using the exact test developed by Wigginton et al. [35] that implemented in Haploview. Haploview was also used to identify linkage disequilibrium. The linkage disequilibrium (LD) patterns of ESR genes are shown in Figure 1. For ESR1 gene, two blocks of 5 kb and 10 kb were identified with high LD. For ESR2 gene, only one 40 kb block was identified. For the case/control association tests, allelic and haplotypic analyses were conducted with chi-square test by Haploview. An empirical p<0.05, corrected after 5,000 permutations, was considered to be statistically significant. Once a significant association appeared for allele analysis, further genotypes analyses based on dominant and recessive models were conducted using logistics regression in IBM SPSS 21 with age, intellectual quotient (IQ), and gender as covariates.

Analysis of gene-gene interaction

Gene-gene interactions were examined by generalized multifactor dimensionality reduction (GMDR). A logistic regression model was used for the gene-gene interaction in ADHD presence, adjusted for age, gender, and IQ. The top-ranked interaction model for a given order interaction was chosen based on: 1) cross validation consistency (CVC) of ≥8/10, 2) prediction accuracy of >50%, and 3) empirical p<0.05 using 1,000 permutation tests. The samples were then divided into highrisk and low-risk combination groups in statistic, ignoring the plausible biological bases. After that, the significant interaction model from GMDR was verified by logistic regression and analyses of covariance using IBM SPSS 21.

RESULTS

Demographic data

A total of 1,997 participants were included: 1,035 in the ADHD group and 962 controls. The demographic characteristics of the subjects are shown in Table 2. The median age of the ADHD group was 109 (93-134) months, and 132 (105-264) months for the control group. A significantly smaller month age was observed in the ADHD group than the control group (p<0.001). Similarly, IQ was significantly lower in the ADHD group (103.91±14.74 vs. 112.73±14.08, p<0.001). There was a statistically significant difference in the gender distribution between the two groups: more males were in the ADHD group than the control group (84.3% vs. 63.1%, p<0.001).

Case/control study

Allelic and genotypic analyses

Table 3 summarized the association test results for SNPs with ADHD presence. For all subjects, no significant associations were found between two SNPs of ESR1 and ADHD. The A allele of rs9340817 and A allele of rs2228480 had nominal p< 0.05 but were not significant after 5,000 permutations (empirical p>0.05). For ESR2, the A alleles of rs8017441 (empirical p=0.0426) and rs960070 (empirical p=0.0076) were significantly associated with ADHD presence after permutation correction. The other two SNPs, rs10459452 (nominal p=0.0137, empirical p=0.0554) and rs2987983 (nominal p=0.0408, empirical p=0.1422) were significantly associated with the occurrence of ADHD prior to correction.

In addition, the association between SNPs of ESR genes and ADHD presence was also analyzed by gender (Table 3). In males, only ESR1 gene rs9340817 A allele was significantly associated with ADHD presence (empirical p=0.0344). The pvalue of ESR1 gene rs2234693 C allele and rs712221 T allele, ESR2 gene rs8017441 A allele were significant before permutations but did not remain statistically significant after correction. In females, no significant SNPs were found to be associated with ADHD after correction.

Genotypic analyses also support association between above SNPs with ADHD. We found that the AA genotype of rs8017441 (p=0.026, OR=1.480 [1.049-2.088]), GG genotype of rs960070 (p=0.008, OR=0.753 [0.61-0.93]) were more frequent in ADHD cases than in controls in all subjects. In males, CC genotype of rs9340817 (p=0.024, OR=1.336 [1.04-1.72]) was less frequent in ADHD cases, suggesting it may be a protective factor.

Haplotypic analyses

The results of haplotypic analyses in all subjects showed that the frequency of the TGGC haplotype (rs944461-rs10459452-rs8017441-rs960070) of ESR2 gene in ADHD cases was lower than that in controls (empirical p=0.0268) (Table 4). In male, the frequency of the AT haplotype (rs9340817-rs712221) of ESR1 in ADHD cases was higher than that in controls (empirical p=0.0058) (Table 4). No significant difference of haplotype between ADHD cases and controls was found in females after permutations.

Gene-gene interaction and the susceptibility of ADHD

All SNPs of the three neurodevelopmental genes and SNPs of ESR genes with nominal p<0.05 in case/control study were included for further gene-gene interaction analysis. When SNPs were in linkage disequilibrium, the SNP with the lowest p-value was included. The included SNPs were: rs2234693, rs9340817, rs6930355, and rs2228480 of ESR1; rs960070 and rs2987983 of ESR2; rs6265, rs2049046, and rs2030324 of BDNF; rs362987 of SANP25; rs6565113 of CDH13.

In the whole sample, no statistically significant interaction model was found between ESR genes and neurodevelopmental genes (p>0.05). In males, a five-locus SNP-SNP interaction model, which composed of ESR2/rs960070, BDNF/rs6265 and BDNF/rs2049046, SNAP25/rs362987, and CDH13/rs6565113, was significantly associated with the presence of ADHD (testing accuracy=54.57%, CVC=8/10, adjusted p<0.001). Details were shown in Table 5 and Figure 2. In females, a three-locus SNPSNP interaction model, which composed of ESR2/rs960070, BDNF/rs6265, and BDNF/rs2049046, was significantly associated with the presence of ADHD (testing accuracy=63.82%, CVC=8/10, adjusted p<0.001). Details were shown in Table 5 and Figure 3.

To verify the above models, subjects were then divided into low-risk and high-risk group according to the GMDR results. As shown in Figures 2 and 3, the light grey cells indicate genotype combinations of the low-risk group and dark grey cells indicate genotype combinations of the high-risk group. After adjusted for month, age, and IQ, the results of logistic regression analyses confirmed the significant difference between low-risk and high-risk group for both models (Table 6).

DISCUSSION

In this case/control study, we investigated the association between the polymorphism of ESR genes and the susceptibility of ADHD among Chinese Han descent. We identified several ESRSNPs related to ADHD. To our knowledge, this is the first study to discover ESRgene SNPs associated with ADHD presence. We also explored the gene-gene interaction effects on ADHD using six screened ESRgene SNPs and five selected neurodevelopmental gene SNPs. Valid interaction models were found for both males and females.

The most important finding of the present study lay in the significant association between polymorphism of ESRgenes and ADHD susceptibility. In all subjects, the polymorphism of ESR2 gene rs960070 and rs8017441 were found significantly related to ADHD presence. Moreover, ESR1 gene rs9340817 was observed associated with male ADHD. These finds were further validated in the case/control tests of genotype and haplotype. However, the impact of ESRgene polymorphism on ADHD is less well-studied presently. No study linked these three SNPs to ADHD. We also retrieved DisGeNET and GWAS Central for these three SNPs in ADHD and other mental disorders, but found no evidence [36,37]. Our data also suggested a trend of association between several SNPs and ADHD susceptibility: ESR1 rs2234693, rs712221, rs6930355, rs2228480, and ESR2 rs10459452, rs2987983. However, these SNPs were not found related to ADHD previously. While rs2234693 was not reported in ADHD, it has been linked to a variety of behavioral phenotypes and clinical outcomes, including anger expression in girl, children and adolescent anxiety, female depression, and etc [15,38-41]. A previous study indicated that rs2144025 is related to the comorbid psychological symptoms of ADHD in both sex. In contrast, the present study revealed no connection between rs2144025 and ADHD susceptibility in either gender [42]. In a study predicting ADHD severity from psychosocial stress and stress-response genes which utilized random forest, rs985191, rs77714417, rs74325817, rs35365822, rs9340910, and rs6930114 of ESR1 gene were ranked in the 25 highest-ranked predictors [43]. All these aforementioned SNPs warrant further investigation in ADHD.

Another significant finding of this research is that the impact of ESRgene polymorphism on ADHD differs across gender. The present study indicated that ESR1 gene rs9340817 A allele and rs9340817-rs712221 AT haplotype are risk factors for ADHD presence in males but not females. Similarly, there was a trend for ESR1 gene rs6930355 to be associated with ADHD in females but not males. These findings align with previous studies showing that ESRgenes affect the sex heterogeneity seen in psychiatric disorders [44-46]. Sexual dimorphism in ADHD has also been reported in our previous studies [47,48]. Prior work exploring the role of genes in the sexual dimorphism of ADHD has mainly focused on sex chromosomes and sex hormones. ADHD symptoms increase in the week before menstruation and resolve during pregnancy. Additionally, with the fluctuation of estrogen level, central stimulants have different therapeutic effects on female ADHD patients [49]. In males, ESR1 gene polymorphisms are associated with aggressive behavior [50]. In animal studies, after ESR1 gene knockout, both female and male mice showed a decrease in BDNF transcription and protein levels, but only female mice showed upregulation of dopamine D1 receptor transcription and protein levels, while male mice did not [51].

Overall, our results indicate that ESRgenes may play a role in the pathophysiology of ADHD, though the specific mechanism remains unclear. We also found that gene-gene interactions among ESR2 rs960070, BDNF rs6265 and rs2049046, SNAP25 rs362987, and CDH13 rs6565113 may be a risk factor for ADHD in Chinese Han males, and female ADHD was significantly associated with interaction model composed of ESR2 rs960070, BDNF rs6265 and rs2049046. These findings served as a reminder that neurodevelopmental and ESRgenes may jointly contribute to ADHD.

Multiple studies have indicated that estrogen can play a neuroprotective role through ESR[52]. Research has shown that estrogen signaling pathways protect the brain from inflammation and stress by promoting BDNF production [53]. Küppers et al. [51] observed decreased BDNF expression in the midbrain of ESR1 knockout mice. The role of BDNF rs6265 in the susceptibility of ADHD is controversial across studies. Luo et al. [54] reported that BDNF rs6265 does not seem to play a significant role in ADHD children, while Ozturk et al. [55] reached the opposite conclusion. Although BDNF rs2049046 was not reported previously in ADHD, He et al. [56] suggested that this SNPs may interact with job burnout. Both rs6265 and rs2049046 were related to the serum BDNF level, which can be influenced by the estrogen signaling pathways and exert its biological effect. For SNAP25, the gene has been found to be associated with susceptibility to ADHD [26]. Our previous study also found SNAP25 as a part of an interaction model contributed to susceptibility to ADHD in males [57]. The SNAP25 rs362987 polymorphism has been linked to ADHD, possibly by affecting the exocytosis of the neurotransmitter gamma-aminobutyric acid (GABA) [58]. Estrogen can influence brain morphology by changing GABA (A) receptor expression and function, which future influence learning and memory [59,60]. For CDH13, rs6565113 polymorphism tended to be associated with specific phenotypes of ADHD [34]. The interaction of ESRgenes and the above neurodevelopmental genes may be a possible mechanism contributing to the effect of ESRgene on ADHD susceptibility.

This study still has the following limitations. First, the number of SNP locus from ESRgenes included in this study was relatively small, which may not fully reflect the role of their polymorphisms in ADHD, and important areas may be missed. Second, it is unclear whether the sample size and power sufficient for identifying gene-gene interactions, so we verified results of this part by logistic regression analysis. Certainly, expanding the sample size cloud make the results more reliable. Furthermore, the serum levels of ESRand other biomarkers were not assessed in this study, which could have provided additional insights. In future studies, measuring serum level of biomarkers and including a larger number of ESRgene SNPs could help validate and expand upon these initial findings.

In conclusion, the present study investigated the ESRgene polymorphisms in Chinese Han children with ADHD, and found they are closely associated with susceptibility of ADHD. Furthermore, we also reveal that the genetic interaction among ESRgenes and the neurodevelopmental genes (i.e., BDNF, SNAP25, and CDH13) plays an important role in ADHD. These findings can help explain the gender difference of susceptibility of ADHD, but still need future validation.

Notes

Availability of Data and Material

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors have no potential conflicts of interest to disclose.

Author Contributions

Conceptualization: Yiwei Lin, Haimei Li, Lu Liu, Qiujin Qian. Data curation: Yiwei Lin, Haimei Li, Lu Liu. Formal analysis: Yiwei Lin, Haimei Li, Qiujin Qian. Funding acquisition: Haimei Li, Lu Liu, Qiujin Qian. Methodology: Yiwei Lin, Lu Liu. Writing—original draft: Yiwei Lin, Qiujin Qian. Writing—review and editing: all authors.

Funding Statement

This study was partly funded by the National Basic Research Development Program of China (grant number 973 program 2014CB846104), National Natural Sciences Foundation of China (grant numbers 81071109; 81301171).

ACKNOWLEDGEMENTS

We would like to express our sincere gratitude to Dr. Hao Zhao for his invaluable assistance in data analysis and expert utilization of the program.

Figure 1.

The linkage disequilibrium (LD) patterns of ESR1 (A) and ESR2 (B) genes. LD, measured as D’, was calculated using Haploview program (version 4.2; https://www.broadinstitute.org/haploview/downloads). Squares in red means strong LD. Numbers in bracket indicate the length of blocks.

Figure 2.

The identified best interaction model in male subjects showed a significant associated with ADHD presence (p<0.001) by GMDR. The left bar in each cell reflects a positive score, while the right bar represents a negative score. Dark shading indicates high-risk cells, light shading indicates low-risk cells, and no shading indicates empty cells. It is worth noting that the patterns of high-risk and low-risk cells alter across each of the distinct multi-locus dimensions, indicating epistasis. ADHD, attention deficit hyperactivity disorder; GMDR, generalized multifactor dimensionality reduction.

Figure 3.

The identified best interaction model in female subjects showed a significant associated with ADHD presence (p<0.001) by GMDR. The left bar in each cell reflects a positive score, while the right bar represents a negative score. Dark shading indicates high-risk cells, light shading indicates low-risk cells, and no shading indicates empty cells. ADHD, attention deficit hyperactivity disorder; GMDR, generalized multifactor dimensionality reduction.

Table 1.

Basic information of SNPs tests in this study

	NCBI SNP reference	Position (GRCh37)	Role	Alleles	HWpval	MAF
ESR1_1	rs12663266	6:152029212	Intron 2	C:T	0.6429	0.157
2	rs1856057	6:152067869	Intron 2	A:G	0.7240	0.313
3	rs7759411	6:152148870	Intron 3	C:T	0.2324	0.098
4	rs2234693	6:152163335	Intron 4	T:C	0.5314	0.384
5	rs9340817	6:152175106	Intron 4	C:A	0.2103	0.246
6	rs712221	6:152180241	Intron 4	A:T	0.8649	0.454
7	rs9340844	6:152201624	Intron 4	G:A	0.7279	0.133
8	rs9322341	6:152224736	Intron 5	T:C	0.3222	0.092
9	rs9383951	6:152295613	Intron 6	G:C	0.0729	0.104
10	rs2144025	6:152307706	Intron 6	T:C	0.3829	0.418
11	rs13216134	6:152328484	Intron 6	A:G	0.3807	0.319
12	rs6930355	6:152371391	Intron 7	T:C	0.2296	0.088
13	rs2207396	6:152382382	Intron 8	G:A	0.3567	0.221
14	rs2228480	6:152420095	Extron 10	G:A	0.7930	0.220
15	rs3798577	6:152421130	Extron 10	T:C	0.2224	0.443
ESR2_1	rs944461	14:64705140	Intron 6	T:C	0.3954	0.303
2	rs10459452	14:64711861	Intron 6	A:G	0.7199	0.066
3	rs8017441	14:64715794	Intron 6	A:G	0.9588	0.092
4	rs960070	14:64745179	Intron 3	G:C	0.9708	0.253
5	rs2987983	14:64763653	Intron 1	A:G	0.2946	0.326
BDNF_1	rs6265	11:27679916	Exon 2	C:T	0.4182	0.462
2	rs2049046	11:27723775	Intron 1	A:T	0.9396	0.485
3	rs2030324	11:27726915	Intron 1	A:G	0.8633	0.462
SNAP25	rs362987	20:10277452	Intron 6	A:G	0.7285	0.178
CDH13	rs6565113	16:83107646	Intron 3	T:G	1.0000	0.065

SNPs, single nucleotide polymorphisms; NCBI, National Center for Biotechnology Information; HWpval, Hardy-Weinberg equilibrium p-value; MAF, minor allele frequency

Table 2.

Demographic characteristics of the included subjects by group

	ADHD group (N=1,035)	Control group (N=962)	Z/t/χ² value	p
Age (mo)	109 (93-134)	132 (105-264)	-14.35	<0.001
IQ	103.91±14.74	112.73±14.08	85.29	<0.001
Sex			117.34	<0.001
Male	873 (84.3)	607 (63.1)
Female	162 (15.7)	355 (36.9)

Values are presented as mean±standard deviation, median (IQR1-IQR3), or number (%). ADHD, attention-deficit/hyperactivity disorder; IQ, intellectual quotient; IQR, interquartile range

Table 3.

Case/control association test for SNPs

Gene		SNP^*	Alleles							Genotypes
			Allele		Ratio counts case, control	Frequencies case, control	Assoc. allele	Nominal p	Empirical p	11 vs. 12+22		11+12 vs. 22
			1	2	Ratio counts case, control	Frequencies case, control	Assoc. allele	Nominal p	Empirical p	OR (95% CI)	p	OR (95% CI)	p
All subjects
	ESR1	rs9340817	A	C	532:1522, 443:1467	0.259, 0.232	A	0.0480	0.4728
		rs2228480	A	G	482:1580, 395:1523	0.234, 0.206	A	0.0344	0.3682
	ESR2	rs10459452	A	G	1953:117, 1778:146	0.943, 0.924	A	0.0137	0.0554
		rs8017441	A	C	1899:167, 1711:199	0.919, 0.896	A	0.0109	0.0426	1.480 (1.049-2.088)	0.026	0.947 (0.24-3.65)	0.937
		rs960070	C	G	1588:480, 1394:528	0.768, 0.725	G	0.0020	0.0076	0.656 (0.423-1.018)	0.060	0.753 (0.61-0.93)	0.008
		rs2987983	A	G	1426:644, 1267:657	0.689, 0.659	A	0.0408	0.1422
Male
	ESR1	rs2234693	C	T	687:1053, 426:784	0.395, 0.352	C	0.0184	0.2132
		rs9340817	A	C	451:1279, 257:951	0.261, 0.213	A	0.0028	0.0344	1.415 (0.802-2.496)	0.231	1.336 (1.04-1.72)	0.024
		rs712221	A	T	814:926, 509:697	0.468, 0.422	T	0.0141	0.1666
		rs2228480	A	G	408:1332, 237:973	0.234, 0.196	A	0.0126	0.1470
	ESR2	rs8017441	A	G	1603:141, 1083:125	0.919, 0.897	A	0.0348	0.1326
Female
	ESR1	rs6930355	C	T	125:191, 224:457	0.396, 0.328	C	0.0372	0.3786
	ESR2	rs960070	C	G	253:71, 504:206	0.781, 0.710	G	0.0168	0.0604

^* SNPs with nominal p<0.05 were listed in the Table.

SNPs, single nucleotide polymorphisms; Assoc., association; OR, odd ratio; CI, confidence interval

Table 4.

Case/control association test for Haplotypes

Haplotypes			Block	Freq.	Case, control frequencies	Nominal p	Empirical p
All subjects
	ESR1
		rs9340817-rs712221	CA	0.540	0.527, 0.554	0.0800	0.3166
			AT	0.239	0.253, 0.224	0.0305^*	0.1294
			CT	0.215	0.215, 0.214	0.9570	>0.9999
		rs6930355-rs2207396	TG	0.434	0.438, 0.430	0.5991	0.9888
			CG	0.346	0.352, 0.339	0.3767	0.8978
			TA	0.220	0.210, 0.231	0.1003	0.3890
	ESR2
		rs944461-rs10459452-rs8017441-rs960070	TGGC	0.065	0.055, 0.075	0.0093	0.0268^*
			TAAC	0.162	0.153, 0.171	0.1290	0.3924
			TAAG	0.445	0.456, 0.433	0.1313	0.3958
			CAAG	0.301	0.310, 0.292	0.2194	0.6274
			TAGC	0.025	0.024, 0.028	0.4184	0.9020
Male
	ESR1
		rs9340817-rs712221	CA	0.544	0.527, 0.568	0.0244^*	0.0890
			AT	0.233	0.254, 0.203	0.0012	0.0058^*
			CT	0.216	0.214, 0.219	0.7265	0.9974
		rs6930355-rs2207396	TG	0.430	0.442, 0.413	0.1082	0.4002
			CG	0.345	0.345, 0.345	0.9862	>0.9999
			TA	0.225	0.213, 0.243	0.0596	0.2240
	ESR2
		rs944461-rs10459452-rs8017441-rs960070	TGGC	0.062	0.055, 0.073	0.0535	0.1598
			TAAG	0.445	0.455, 0.431	0.1901	0.5596
			TAGC	0.025	0.023, 0.028	0.3840	0.8446
			TAAC	0.158	0.155, 0.163	0.5993	0.9812
			CAAG	0.306	0.309, 0.302	0.6825	0.9938
Female
	ESR1
		rs9340817-rs712221	CA	0.530	0.527, 0.531	0.9128	>0.9999
			AT	0.256	0.246, 0.260	0.6396	0.9916
			CT	0.211	0.223, 0.206	0.5287	0.9718
		rs6930355-rs2207396	TG	0.445	0.415, 0.459	0.1926	0.5770
			CG	0.348	0.393, 0.327	0.0395^*	0.1806
			TA	0.206	0.415, 0.459	0.1926	0.5770
	ESR2
		rs944461-rs10459452-rs8017441-rs960070	TAAC	0.170	0.139, 0.184	0.0723	0.2382
			TGGC	0.071	0.052, 0.079	0.1242	0.3828
			CAAG	0.288	0.318, 0.274	0.1553	0.4378
			TAAG	0.443	0.460, 0.435	0.4589	0.9270
			TAGC	0.027	0.027, 0.027	0.9495	>0.9999

^* p<0.05. Freq., frequency

Table 5.

Interaction models between ESR genes and neurodevelopmental genes

Phenotype	Best interaction model	Testing accuracy (%)	CVC	Adjusted p
ADHD-male	ESR2 rs960070/BDNF rs6265/BDNF rs2049046/SNAP25 rs362987/CDH13 rs6565113	54.57	8/10	0.0004^***
ADHD-female	ESR2 rs960070/BDNF rs6265/BDNF rs2049046	63.82	8/10	<0.0002^***

^*** p<0.001.

ESR, estrogen receptor; ADHD, attention-deficit/hyperactivity disorder; CVC, cross validation consistency

Table 6.

Logistic regression analyses results for interaction models

Phenotype	Top-ranked interaction model	Genotype groups	OR (95% CI)	p
ADHD-male	ESR2 rs960070/BDNF rs6265/BDNF rs2049046/SNAP25 rs362987/CDH13 rs6565113	High-risk group (N=883)	4.07 (3.15-5.26)	<0.001
ADHD-male		Low-risk group (N=550)
ADHD-female	ESR2 rs960070/BDNF rs6265/BDNF rs2049046	High-risk group (N=252)	4.61 (2.99-7.10)	<0.001
ADHD-female	ESR2 rs960070/BDNF rs6265/BDNF rs2049046	Low-risk group (N=248)

ADHD, attention deficit hyperactivity disorder; OR, odd ratio; CI, confidence interval

REFERENCES

1. Francés L, Quintero J, Fernández A, Ruiz A, Caules J, Fillon G, et al. Current state of knowledge on the prevalence of neurodevelopmental disorders in childhood according to the DSM-5: a systematic review in accordance with the PRISMA criteria. Child Adolesc Psychiatry Ment Health 2022;16:27

2. Li F, Cui Y, Li Y, Guo L, Ke X, Liu J, et al. Prevalence of mental disorders in school children and adolescents in China: diagnostic data from detailed clinical assessments of 17,524 individuals. J Child Psychol Psychiatry 2022;63:34-46.

3. Faraone SV, Larsson H. Genetics of attention deficit hyperactivity disorder. Mol Psychiatry 2019;24:562-575.

4. Thapar A. Discoveries on the genetics of ADHD in the 21st century: new findings and their implications. Am J Psychiatry 2018;175:943-950.

5. Thapar A, Cooper M. Attention deficit hyperactivity disorder. Lancet 2016;387:1240-1250.

6. Carucci S, Narducci C, Bazzoni M, Balia C, Donno F, Gagliano A, et al. Clinical characteristics, neuroimaging findings, and neuropsychological functioning in attention-deficit hyperactivity disorder: sex differences. J Neurosci Res 2023;101:704-717.

7. Galea LAM, Frick KM, Hampson E, Sohrabji F, Choleris E. Why estrogens matter for behavior and brain health. Neurosci Biobehav Rev 2017;76:363-379.

8. Waddell J, McCarthy MM. Sexual differentiation of the brain and ADHD: what is a sex difference in prevalence telling us? Curr Top Behav Neurosci 2012;9:341-360.

9. Hill RA, Wu YW, Gogos A, van den Buuse M. Sex-dependent alterations in BDNF-TrkB signaling in the hippocampus of reelin heterozygous mice: a role for sex steroid hormones. J Neurochem 2013;126:389-399.

10. Hara Y, Waters EM, McEwen BS, Morrison JH. Estrogen effects on cognitive and synaptic health over the lifecourse. Physiol Rev 2015;95:785-807.

11. Hwang WJ, Lee TY, Kim NS, Kwon JS. The role of estrogen receptors and their signaling across psychiatric disorders. Int J Mol Sci 2020;22:373

12. Ma Y, Guo H, Zhang L, Tao L, Yin A, Liu Z, et al. Estrogen replacement therapy-induced neuroprotection against brain ischemia-reperfusion injury involves the activation of astrocytes via estrogen receptor β. Sci Rep 2016;6:21467

13. Gu Y, Chen X, Fu S, Liu W, Wang Q, Liu KJ, et al. Astragali Radix isoflavones synergistically alleviate cerebral ischemia and reperfusion injury via activating estrogen receptor-PI3K-Akt signaling pathway. Front Pharmacol 2021;12:533028. Retraction in: Front Pharmacol 2023;14:1252192

14. Min JA, Kim JJ, Pae CU, Kim KH, Lee CU, Lee C, et al. Association of estrogen receptor genes and schizophrenia: a preliminary study. Prog Neuropsychopharmacol Biol Psychiatry 2012;36:1-4.

15. Wang S, Li W, Zhao J, Zhang H, Yang Y, Wang X, et al. Association of estrogen receptor alpha gene polymorphism with age at onset, general psychopathology symptoms, and therapeutic effect of schizophrenia. Behav Brain Funct 2013;9:12

16. Ozsoy F, Nursal AF, Karakus N, Demir MO, Yigit S. Estrogen receptor 1 gene rs22346939 and rs9340799 variants are associated with major depressive disorder and its clinical features. Curr Neurovasc Res 2021;18:12-19.

17. Doi H, Fujisawa TX, Iwanaga R, Matsuzaki J, Kawasaki C, Tochigi M, et al. Association between single nucleotide polymorphisms in estrogen receptor 1/2 genes and symptomatic severity of autism spectrum disorder. Res Dev Disabil 2018;82:20-26.

18. Franke B, Faraone SV, Asherson P, Buitelaar J, Bau CH, Ramos-Quiroga JA, et al. The genetics of attention deficit/hyperactivity disorder in adults, a review. Mol Psychiatry 2012;17:960-987.

19. Hess JL, Radonjić NV, Patak J, Glatt SJ, Faraone SV. Autophagy, apoptosis, and neurodevelopmental genes might underlie selective brain region vulnerability in attention-deficit/hyperactivity disorder. Mol Psychiatry 2021;26:6643-6654.

20. Liu YS, Dai X, Wu W, Yuan FF, Gu X, Chen JG, et al. The association of SNAP25 gene polymorphisms in attention deficit/hyperactivity disorder: a systematic review and meta-analysis. Mol Neurobiol 2017;54:2189-2200.

21. Sokolowski M, Wasserman J, Wasserman D. Polygenic associations of neurodevelopmental genes in suicide attempt. Mol Psychiatry 2016;21:1381-1390.

22. Li Y, Li F, Qin D, Chen H, Wang J, Wang J, et al. The role of brain derived neurotrophic factor in central nervous system. Front Aging Neurosci 2022;14:986443

23. Harte-Hargrove LC, Maclusky NJ, Scharfman HE. Brain-derived neurotrophic factor-estrogen interactions in the hippocampal mossy fiber pathway: implications for normal brain function and disease. Neuroscience 2013;239:46-66.

24. Li H, Liu L, Tang Y, Ji N, Yang L, Qian Q, et al. Sex-specific association of brain-derived neurotrophic factor (BDNF) Val66Met polymorphism and plasma BDNF with attention-deficit/hyperactivity disorder in a drug-naïve Han Chinese sample. Psychiatry Res 2014;217:191-197.

25. Chhibber A, Woody SK, Karim Rumi MA, Soares MJ, Zhao L. Estrogen receptor β deficiency impairs BDNF-5-HT2A signaling in the hippocampus of female brain: a possible mechanism for menopausal depression. Psychoneuroendocrinology 2017;82:107-116.

26. Feng Y, Crosbie J, Wigg K, Pathare T, Ickowicz A, Schachar R, et al. The SNAP25 gene as a susceptibility gene contributing to attention-deficit hyperactivity disorder. Mol Psychiatry 2005;10:998-1005.

27. Thompson Gray AD, Simonetti J, Adegboye F, Jones CK, Zurawski Z, Hamm HE. Sexual dimorphism in stress-induced hyperthermia in SNAP25Δ3 mice, a mouse model with disabled Gβγ regulation of the exocytotic fusion apparatus. Eur J Neurosci 2020;52:2815-2826.

28. Rivero O, Sich S, Popp S, Schmitt A, Franke B, Lesch KP. Impact of the ADHD-susceptibility gene CDH13 on development and function of brain networks. Eur Neuropsychopharmacol 2013;23:492-507.

29. Kiser DP, Popp S, Schmitt-Böhrer AG, Strekalova T, van den Hove DL, Lesch KP, et al. Early-life stress impairs developmental programming in Cadherin 13 (CDH13)-deficient mice. Prog Neuropsychopharmacol Biol Psychiatry 2019;89:158-168.

30. Guan L, Wang B, Chen Y, Yang L, Li J, Qian Q, et al. A high-density single-nucleotide polymorphism screen of 23 candidate genes in attention deficit hyperactivity disorder: suggesting multiple susceptibility genes among Chinese Han population. Mol Psychiatry 2009;14:546-554.

31. Aureli A, Del Beato T, Sebastiani P, Marimpietri A, Melillo CV, Sechi E, et al. Attention-deficit hyperactivity disorder and intellectual disability: a study of association with brain-derived neurotrophic factor gene polymorphisms. Int J Immunopathol Pharmacol 2010;23:873-880.

32. Wang N, Wang Z, Yan F, Fu W. [Correlation between single nucleotide polymorphisms of neurotrophic factors and executive function characteristics in children with attention deficit hyperactivity disorder]. Wei Sheng Yan Jiu 2198;48:577-582. Chinese.

33. Lozano M, Murcia M, Soler-Blasco R, González L, Iriarte G, Rebagliato M, et al. Exposure to mercury among 9-year-old children and neurobehavioural function. Environ Int 2021;146:106173

34. Lasky-Su J, Neale BM, Franke B, Anney RJ, Zhou K, Maller JB, et al. Genome-wide association scan of quantitative traits for attention deficit hyperactivity disorder identifies novel associations and confirms candidate gene associations. Am J Med Genet B Neuropsychiatr Genet 2008;147B:1345-1354.

35. Wigginton JE, Cutler DJ, Abecasis GR. A note on exact tests of HardyWeinberg equilibrium. Am J Hum Genet 2005;76:887-893.

36. Beck T, Rowlands T, Shorter T, Brookes AJ. GWAS central: an expanding resource for finding and visualising genotype and phenotype data from genome-wide association studies. Nucleic Acids Res 2023;51(D1):D986-D993.

37. Piñero J, Ramírez-Anguita JM, Saüch-Pitarch J, Ronzano F, Centeno E, Sanz F, et al. The DisGeNET knowledge platform for disease genomics: 2019 update. Nucleic Acids Res 2020;48(D1):D845-D855.

38. Alonso P, Gratacòs M, Segalàs C, Escaramís G, Real E, Bayés M, et al. Variants in estrogen receptor alpha gene are associated with phenotypical expression of obsessive-compulsive disorder. Psychoneuroendocrinology 2011;36:473-483.

39. Vermeersch H, T’sjoen G, Kaufman JM, Van Houtte M. ESR1 polymorphisms, daily hassles, anger expression, and depressive symptoms in adolescent boys and girls. Horm Behav 2013;63:447-453.

40. Prichard Z, Jorm AF, Prior M, Sanson A, Smart D, Zhang Y, et al. Association of polymorphisms of the estrogen receptor gene with anxiety-related traits in children and adolescents: a longitudinal study. Am J Med Genet 2002;114:169-176.

41. Keyes K, Agnew-Blais J, Roberts AL, Hamilton A, De Vivo I, Ranu H, et al. The role of allelic variation in estrogen receptor genes and major depression in the Nurses Health Study. Soc Psychiatry Psychiatr Epidemiol 2015;50:1893-1904.

42. Pinsonneault JK, Frater JT, Kompa B, Mascarenhas R, Wang D, Sadee W. Intronic SNP in ESR1 encoding human estrogen receptor alpha is associated with brain ESR1 mRNA isoform expression and behavioral traits. PLoS One 2017;12:e0179020.

43. van der Meer D, Hoekstra PJ, van Donkelaar M, Bralten J, Oosterlaan J, Heslenfeld D, et al. Predicting attention-deficit/hyperactivity disorder severity from psychosocial stress and stress-response genes: a random forest regression approach. Transl Psychiatry 2017;7:e1145

44. Li C, Xie M, Wang W, Liu Y, Liao D, Yin J, et al. Association between polymorphisms in estrogen receptor genes and depression in women: a meta-analysis. Front Genet 2022;13:936296

45. Mill J, Kiss E, Baji I, Kapornai K, Daróczy G, Vetró A, et al. Association study of the estrogen receptor alpha gene (ESR1) and childhoodonset mood disorders. Am J Med Genet B Neuropsychiatr Genet 2008;147B:1323-1326.

46. Arnold ML, Saijo K. Estrogen receptor β as a candidate regulator of sex differences in the maternal immune activation model of ASD. Front Mol Neurosci 2021;14:717411

47. Liu L, Guan LL, Chen Y, Ji N, Li HM, Li ZH, et al. Association analyses of MAOA in Chinese Han subjects with attention-deficit/hyperactivity disorder: family-based association test, case-control study, and quantitative traits of impulsivity. Am J Med Genet B Neuropsychiatr Genet 2011;156B:737-748.

48. Liu L, Chen Y, Li H, Qian Q, Yang L, Glatt SJ, et al. Association between SYP with attention-deficit/hyperactivity disorder in Chinese Han subjects: differences among subtypes and genders. Psychiatry Res 2013;210:308-314.

49. Camara B, Padoin C, Bolea B. Relationship between sex hormones, reproductive stages and ADHD: a systematic review. Arch Womens Ment Health 2022;25:1-8.

50. Vaillancourt KL, Dinsdale NL, Hurd PL. Estrogen receptor 1 promoter polymorphism and digit ratio in men. Am J Hum Biol 2012;24:682-689.

51. Küppers E, Krust A, Chambon P, Beyer C. Functional alterations of the nigrostriatal dopamine system in estrogen receptor-alpha knockout (ERKO) mice. Psychoneuroendocrinology 2008;33:832-838.

52. Azcoitia I, Barreto GE, Garcia-Segura LM. Molecular mechanisms and cellular events involved in the neuroprotective actions of estradiol. Analysis of sex differences. Front Neuroendocrinol 2019;55:100787

53. Saldanha CJ. Estrogen as a neuroprotectant in both sexes: stories from the bird brain. Front Neurol 2020;11:497

54. Luo L, Jiang X, Cao G, Xiong P, Yang R, Zhang J, et al. Association between BDNF gene polymorphisms and attention deficit hyperactivity disorder in school-aged children in Wuhan, China. J Affect Disord 2020;264:304-309.

55. Ozturk O, Basay BK, Buber A, Basay O, Alacam H, Bacanlı A, et al. Brain-derived neurotrophic factor gene Val66Met polymorphism is a risk factor for attention-deficit hyperactivity disorder in a Turkish sample. Psychiatry Investig 2016;13:518-525.

56. He SC, Wu S, Wang C, Wang DM, Wang J, Xu H, et al. Interaction between job stress, serum BDNF level and the BDNF rs2049046 polymorphism in job burnout. J Affect Disord 2020;266:671-677.

57. Gao Q, Liu L, Chen Y, Li H, Yang L, Wang Y, et al. Synaptosome-related (SNARE) genes and their interactions contribute to the susceptibility and working memory of attention-deficit/hyperactivity disorder in males. Prog Neuropsychopharmacol Biol Psychiatry 2015;57:132-139.

58. Fan HP, Fan FJ, Bao L, Pei G. SNAP-25/syntaxin 1A complex functionally modulates neurotransmitter gamma-aminobutyric acid reuptake. J Biol Chem 2006;281:28174-28184.

59. Locci A, Porcu P, Talani G, Santoru F, Berretti R, Giunti E, et al. Neonatal estradiol exposure to female rats changes GABAA receptor expression and function, and spatial learning during adulthood. Horm Behav 2017;87:35-46.

60. Mukherjee J, Cardarelli RA, Cantaut-Belarif Y, Deeb TZ, Srivastava DP, Tyagarajan SK, et al. Estradiol modulates the efficacy of synaptic inhibition by decreasing the dwell time of GABAA receptors at inhibitory synapses. Proc Natl Acad Sci U S A 2017;114:11763-11768.