Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder that commonly occurs in childhood, and the symptoms generally persist until adulthood. Its global prevalence is around 5% [1], and it is characterized by two basic symptoms, namely inattention and hyperactivity/impulsivity [2]. It involves gene-by-environment interactions and is a complex genetic disorder with a high heritability rate (74%) [1].

Endocrine disrupting chemicals (EDCs) are natural or manufactured chemicals that interfere with the normal functioning of the endocrine system. They are widely used in various industries, including agriculture and pharmaceuticals. EDCs include pesticides like dichloro-diphenyl-trichloroethane and dioxin, synthetic chemicals designed for specific actions like plasticizers, solvents like polychlorinated biphenyls, phthalates, bisphenols, phenols, and parabens; and heavy metals like lead (Pb), mercury (Hg), and cadmium (Cd) [3,4]. They block the pathway between natural hormones and receptors, act directly on the gland to cause excess or deficiency of hormones, and interfere with normal bodily functions by mimicking hormones [5]. The seriousness of the problems associated with EDCs has recently surfaced, as it has been found in many everyday products, including plastic bottles, metal food cans, detergents, flame retardants, food additives, toys, cosmetics, and pesticides [4,5].

Phthalates, one of the most commonly used EDC, are used in plasticizers, solvents, cosmetics, and bleaches. They are metabolized into urine; metabolites, such as mono-[2-ethyl-5-oxohexyl] phthalate (MEOHP), mono-[2-ethyl-5-hydroxyhexyl] phthalate (MEHHP), and mono-n-butyl phthalate have been detected [6]. Previous studies investigated the association between phthalates and ADHD. In Korean schoolage children, ADHD symptoms reported by teachers showed a strong positive correlation with urinary concentrations of phthalate metabolite [6]. A study of children aged 6–15 years in the U.S. showed a strong association between phthalate exposure, attention deficit disorder, and learning disabilities [7]. Childhood phthalate exposure is associated with temperamental traits in children with ADHD [8]. Additionally, phthalate exposure and ADHD symptoms have various quantitative associations in children [9]. Studies showed that the concentrations of some phthalate metabolites are negatively correlated with the cortical thickness in the right middle and superior temporal gyri [10]. Lead, phthalate, and bisphenol A (BAP) are associated with ADHD ranging from moderate to high levels [11].

BPA is also a widely used EDC. It is commonly used in the manufacturing of polycarbonate plastics and epoxy resins. It is found in food and beverage containers, toys, medical devices, and receipt paper [12] and has estrogen-mimicking properties [12,13]. Research has shown that more than 90% of the population excretes detectable levels of BPA in urine [13]. As the effects of BPA on neurodevelopment and behavioral changes have become known in several studies, in the 2020s, the use of BPA was regulated in some countries and in products used by children or vulnerable groups. Accordingly, similar substances, such as bisphenol F (BPF) and bisphenol S (BPS), are increasingly being used to replace BPA [14]. However, they also have BPA-like molecular structures, similar to those of estrogens and androgens. Furthermore, all three types of bisphenols exhibit anti-androgenic activities [14,15]; BPA and BPF are known to affect thyroid hormone transport [12,14].

There have also been several previous studies on behavioral changes related to bisphenols. In various animal experiments, an exposure to bisphenols induced hyperactivity, increased depressive behavior, increased anxiety, and modified sociosexual behavior [16,17]. Studies involving humans showed that prenatal and childhood exposure to BPA alters normal neurodevelopment and may be associated with adverse behavioral outcomes [18]. However, the relationship between BPA and ADHD remains inconclusive [19]. Some studies on school-aged children found an association between childhood BPA exposure and the symptoms of inattention or hyperactivity reported by parents or teachers [20-22]. However, other studies found no such relationship [23,24]. Studies showed that exposure to BPA, BPF, and BPS is associated with ADHD symptoms in children aged six who have not been diagnosed with ADHD [25]. However, the effects of BPF and BPS on attention and behavioral problems in a clinically diagnosed group have not been investigated.

Therefore, we examined the relationship between phthalate, BPA, BPF, and BPS exposure and the severity of ADHD symptoms using neuropsychological test indicators in a clinically diagnosed ADHD group.

Data collection

This study used the data of children with ADHD aged 6–15 years who were recruited from the Research Center of Neurodevelopment Disorders in the Child Psychiatric Clinic of Seoul National University Hospital from April 2020 to July 2022. The Korean Kiddie-Schedule for Affective Disorders and Schizophrenia - Present and Lifetime Version (K-SADSPL) [26] was used to classify the children diagnosed with ADHD according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [27] criteria. The diagnosis was ascertained through semi-structured interviews conducted by a child psychiatrist. The study group consisted entirely of medication-naïve children.

Children with autism spectrum disorder (ASD), pervasive developmental disorder, mental retardation, bipolar disorder, psychotic disorder, obsessive-compulsive disorder, or Tourette’s syndrome were excluded from the study. Patients with a history of organic brain disease, seizure disorders, or other neurological disorders were also excluded. Additionally, individuals with intellectual disabilities (IQ below 70), learning disabilities, and speech disorders were excluded. Children with major depressive disorders, anxiety disorders, or tic disorders requiring medication were also excluded from the study.

The level of EDC exposure was measured using the concentration of EDCs or their metabolites in the urine. Urine samples were collected from the children after fasting for at least eight hours. The severity of ADHD symptoms was evaluated using neuropsychological tests and a clinical symptom scale.

The neuropsychological tests included Advanced Test of Attention (ATA; Visual/Auditory), Wisconsin Sentence Completion Test, STROOP test, and Children’s Color Trails Test. Clinical Global Impression Severity (CGI-S), Children’s Global Assessment Scale (C-GAS) scores were used to measure clinical symptoms. Additionally, brain imaging was performed and whole-brain structural magnetic resonance imaging (MRI) images were obtained using a T1-weighted protocol.

Ethical approval was obtained from the Institutional Review Board of the Medical Research Ethics Committee of Seoul National University Hospital (IRB No.: 1507-118-690, 2008-116-1150). A written informed consent was obtained from both the participants and their parents after study information was explained.

Urine sample analysis

The measured urinary concentrations of phthalate metabolites and total concentrations of bisphenols were quantified using a high-performance liquid chromatography-tandem mass spectrometer as the method used in previously studies [6,25]. Bias in urine concentrations due to age was corrected using creatinine levels.

Statistical analysis

First, we used the Jonckheere–Terpstra test to determine whether there is a trend in ADHD symptom severity and urinary concentrations of phthalate metabolites and bisphenols. To determine the effect of EDC exposure on neuropsychological tests, the significance was verified using Pearson and Spearman correlation analyses. The effect size was analyzed using a linear regression model (LRM) and multiple regression model (MRM). Age, sex, and Full-Scale Intelligence Quotient (FSIQ) were considered covariates according to prior studies [28,29].

During the recruitment period for the experimental group, 75 children with ADHD were enrolled, 4 of whom voluntarily withdrew from the study during the FSIQ test. Subsequently, 4 more children dropped out during the neuropsychological test, and data from 67 children diagnosed with ADHD were used for the analysis. Among them, there were 59 boys and 8 girls, with ages ranging from 6–16 years; the mean age was approximately 8 years (mean: 8.55 years, standard deviation: 2.44 years). We recruited research participants up to the age of 15 years, but due to the time it took to undergo psychological testing after enrollment, 2 children were 16 at the time of the test (Fig. 1). Table 1 lists the participants’ characteristics.

Fig. 1. Diagram of the study population. ADHD, attention-deficit/hyperactivity disorder; F, female; FSIQ, Full-Scale Intelligence Quotient; M, male; SD, standard deviation.

Table 1 <caption>

The characteristics of participants

Characteristics	6-7 yrs (n=29)	8-10 yrs (n=27)	11-13 yrs (n=6)	14-17 yrs (n=5)	Total (n=67)
Sex
Female	2 (6.90)	4 (14.81)	2 (33.33)	0 (0)	8 (11.94)
Male	27 (93.10)	23 (85.19)	4 (66.67)	5 (100)	59 (88.06)
Age (yr)	6.66±0.48	8.78±0.85	11.33±0.82	15.00±1.00	8.55±2.44

<fn id="t1fn1">

Values are presented as mean±standard deviation or n (%).

When examining the average EDC concentration for each group according to the CGI-S score, the Jonckheere–Terpstra test showed no significant trend in the average values (Tables 2, 3 and Fig. 2). Thus, the effects of phthalates, BPA, BPF, and BPS on the severity of ADHD symptoms are unclear.

Fig. 2. Graph presenting the average of urinary EDC concentrations by CGI-S group. BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; CGI-S, Clinical Global Impression Severity; EDC, endocrine disrupting chemical; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate; MnBP, mono-n-butyl phthalate.

Table 2 <caption>

Urinary EDC concentrations by CGI-S group

CGI-S		MEHHP (μg/L)	MEOHP (μg/L)	MnBP (μg/L)	BPS (μg/L)	BPF (μg/L)	BPA (μg/L)
2	Average	25.768	10.505	18.532	2.635	0.178	0
	N	1	1	1	1	1	1
	SD	0	0	0	0	0	0
3	Average	18.583	9.786	33.960	2.072	1.647	1.652
	N	11	11	11	11	11	11
	SD	14.468	7.167	39.616	3.594	3.769	1.723
4	Average	18.897	12.992	47.023	0.353	1.030	0.867
	N	34	34	33	36	36	36
	SD	14.850	11.028	47.014	0.294	2.575	0.756
5	Average	11.828	7.840	33.728	0.932	0.288	1.078
	N	18	18	18	18	18	18
	SD	7.810	5.554	23.823	1.224	0.411	0.844
Total	Average	16.962	10.953	40.492	0.832	0.918	1.042
	N	64	64	63	66	66	66
	SD	13.283	9.243	40.011	1.693	2.453	1.023

<fn id="t2fn1">

BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; CGI-S, Clinical Global Impression Severity; EDC, endocrine disrupting chemical; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate; MnBP, mono-n-butyl phthalate; N, number; SD, standard deviation

Table 3 <caption>

Jonckheere−Terpstra test to find out trend

	MEHHP	MEOHP	MnBP	BPS	BPF	BPA
Observed J-T statistic	1.000	2.000	4.000	1.000	3.000	4.000
Mean J-T statistic	3.000	3.000	3.000	3.000	3.000	3.000
SD of J-T statistic	1.472	1.472	1.472	1.472	1.472	1.472
Std. J-T statistic	-1.359	-0.679	0.679	-1.359	0.000	0.679
Asymp. Sig. (2-tailed)	0.174	0.497	0.497	0.174	1.000	0.497

<fn id="t3fn1">

Asymp. Sig., asymptotic significance; BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; J-T, Jonckheere–Terpstra; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate; MnBP, mono-n-butyl phthalate; SD, standard deviation; Std., standardized

In the Pearson correlation analysis used to determine the significance between urine EDC concentrations and neuropsychological tests, there were no significant correlations with EDC concentrations in most of indicators. Although, among the phthalate metabolites, there was a significant correlation between urinary MEHHP, MEOHP, and the T-scores of commission errors in the ATA visual test (Table 4).

Table 4 <caption>

Pearson correlation analysis between urine EDC concentrations and neuropsychological indicators

EDC	CGI-S	C-GAS	ATA(V)_OE	ATA(V)_CE	ATA(V)_RT	ATA(A)_OE	ATA(A)_CE	ATA(A)_RT	STROOP_W	STROOP_C	STROOP_I	CCTT-2	WCST_CLR
MEHHP
Pearson coefficient	-0.107	-0.185	0.184	0.381	0.007	0.084	0.247	-0.219	-0.119	-0.055	0.109	0	-0.005
p	0.451	0.189	0.192	0.005**	0.963	0.556	0.078	0.118	0.401	0.696	0.441	0.998	0.973
MEOHP
Pearson coefficient	-0.011	-0.248	0.184	0.323	0.063	0.118	0.116	-0.149	-0.071	-0.021	0.092	0.059	-0.037
p	0.938	0.076	0.191	0.020*	0.660	0.404	0.415	0.293	0.619	0.885	0.516	0.677	0.796
MnBP
Pearson coefficient	0.030	-0.205	-0.194	-0.185	0.163	0.162	-0.136	0.111	-0.091	-0.112	0.091	0.247	-0.082
p	0.833	0.144	0.168	0.190	0.248	0.252	0.338	0.433	0.523	0.427	0.522	0.077	0.566
BPA
Pearson coefficient	-0.116	0.080	0.035	-0.191	0.239	0.064	0.020	0.048	-0.069	-0.196	0.201	-0.086	-0.105
p	0.403	0.563	0.801	0.167	0.082	0.644	0.885	0.730	0.620	0.157	0.145	0.536	0.449
BPF
Pearson coefficient	-0.162	0.061	-0.126	-0.047	-0.003	0.084	-0.057	0.144	-0.163	-0.065	-0.034	-0.037	0.044
p	0.241	0.661	0.362	0.735	0.985	0.548	0.681	0.299	0.238	0.638	0.806	0.788	0.753
BPS
Pearson coefficient	-0.196	0.195	-0.009	-0.111	0.008	-0.116	0.099	-0.058	-0.043	-0.031	0.067	-0.123	-0.239
p	0.156	0.158	0.950	0.425	0.956	0.404	0.475	0.679	0.759	0.822	0.629	0.375	0.082

<fn id="t4fn1">

*p<0.05; **p<0.01. ATA(A), Advanced Test of Attention (Auditory attention); ATA(V), Advanced Test of Attention (Visual attention); BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; CCTT, Children's Color Trails Test; CE, commission errors (T score); C-GAS, Children's Global Assessment Scale; CGI-S, Clinical Global Impression-Severity; EDC, endocrine disrupting chemical; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate; MnBP, mono-n-butyl phthalate; OE, omission errors (T score); RT, response time (T score); STROOP_C, STROOP_Color (T score); STROOP_I, STROOP_Interference (T score); STROOP_W, STROOP_Word (T score); WCST_CLR, Wisconsin Card Sorting Test_Conceptual Level Responses

The CGI-S score obtained by the clinician’s evaluation was classified into five levels and classified as a continuous variable; therefore, there was a limitation in the validity of analyzing the Pearson correlation. For that reason, considering the nonparametric tendency of the CGI-S score, Spearman’s nonparametric correlation analysis was performed on the EDC concentration and CGI-S score. However, no significant correlation were observed. Additionally, we repeated the t-test by dividing the bisphenol concentrations into two groups, above and below the detection limit; the results showed no statistical significance (Table 5).

Table 5 <caption>

Spearman correlation analysis between urine EDC concentrations and CGI-S

	MEHHP	MEOHP	MnBP	BPA	BPF	BPS
CGI-S
Spearman coefficient	-0.285	-0.202	0.030	0.017	0.016	-0.024
p (2-tailed)	0.023*	0.112	0.815	0.896	0.904	0.855

<fn id="t5fn1">

*p<0.05. BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; CGI-S, Clinical Global Impression-Severity; EDC, endocrine disrupting chemical; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate; MnBP, mono-n-butyl phthalate

Subsequently, we attempted to define the correlation by performing a LRM on the phthalate metabolite concentrations and visual ATA commission error T-score data, which revealed a significant correlation. When the Log2 scale was applied to the metabolite concentrations, the p-values decreased, and a more accurate regression equation was obtained (Table 6).

Table 6 <caption>

Linear regression on phthalate metabolite concentrations and ATA(V)_CE T-score

	R2	p	β	Constant
MEHHP	0.139	0.003**	0.582	61.781
MEOHP	0.100	0.012*	0.708	63.931
log2MEHHP	0.170	0.001**	8.190	41.206
log2MEOHP	0.147	0.002**	8.190	46.283

<fn id="t6fn1">

*p<0.05; **p<0.01. ATA(V), Advanced Test of Attention (Visual attention); CE, commission errors; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate

In addition to the LRM, a MRM was applied to the two independent variables by correcting for age, sex, and FSIQ variables, which have been identified as co-variants in previous studies. In the MRM, MEOHP was excluded from the variable, and the same equation was derived as in the LRM. This is covered under Discussion.

In this study, we found that phthalate exposure was correlated with commission errors on the visual ATA test. According to the MRM, when the metabolite concentration in the urine doubled, the commission error T-score increased eightfold. No associations were found with other neuropsychological test indicators, including the auditory ATA test. Bisphenol exposure was not significantly associated with any item on the neuropsychological test index.

The ATA commission error indicator was calculated based on the number of incorrect responses, responding impulsively to stimuli that should not have been responded. Therefore, the results of the present study indicate that phthalate exposure increases hyperactivity. This is consistent with the results of a previous study that used the parent-reported ADHD Rating Scale-IV [10]. However, in this study, no association was observed between EDCs and indicators of inattention, such as ATA omission errors.

When analyzed using various statistical methods, values representing comprehensive symptom severity, such as the CGI-S and C-GAS, showed no significant correlation with phthalate or bisphenol exposure. Thus, the effect of phthalate exposure on hyperactivity does not seem to be sufficiently significant to affect the severity score, which comprehensively reflects the main symptoms of ADHD.

Considering the results of the multiple regression analysis among the statistical analyses presented above (Table 7), as MEOHP was excluded from the variable, the same equation as the linear regression analysis using MEHHP as a variable was obtained (Table 6). Based on the rationale, most children are mainly exposed to di-[2-ethylhexyl] phthalate, which goes through mono-[2-ethylhexyl] phthalate metabolism to become MEHHP. Then, some MEHHP is excreted in urine, while the rest is oxidized to produce MEOHP (Fig. 3) [30]. When MEHHP and MEOHP were simultaneously included as independent variables and analyzed by multiple regression, MEHHP, which had a lower significance probability, was analyzed as a leading variable, as the association between the two metabolites was reflected, and the analysis using MEOHP as a variable may have lost its effect.

Fig. 3. The metabolic pathway of DEHP in humans. DEHP, di-[2-ethylhexyl] phthalate; MEHP, mono-[2-ethylhexyl] phthalate; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate.

Table 7 <caption>

Multiple regression on phthalate metabolite concentrations and ATA(V)_CE T-score

	R2	p	β	Constant	Excluded variable
MEHHP	0.139	0.003**	0.582	61.781	MEOHP
log2MEHHP	0.170	0.001**	8.190	41.206	log2MEOHP

<fn id="t7fn1">

**p<0.01. ATA(V), Advanced Test of Attention (Visual attention); CE, commission errors; MEHHP, mono-[2-ethyl-5-hydroxyhexyl] phthalate; MEOHP, mono-[2-ethyl-5-oxohexyl] phthalate

Limitations

This study had the limitation of a small sample size of 67 individuals, and the sex ratio of females to males in the group was uneven. Maternal EDC exposure level, parental education level, and child’s body mass index at birth are known to affect ADHD symptoms by previous studies. We did not obtain these data; therefore, we could not correct for the effects of the covariant factors. Additionally, although it was less than one year, there was a time gap between the date of urine sample collection and the date of neuropsychological testing.

Significance

Despite the limitations discussed above, this study is significant in several respects. First, unlike previous studies, this study examined the correlation between ADHD symptoms and neuropsychological test indicators reported by the children themselves and not by others. Second, while many studies used the EDC concentration at birth to examine the association with ADHD symptoms during brain development, this study used the amount of EDC detected in urine samples, reflecting direct exposure in daily life. Another significance of this study is that we investigated the relationship between ADHD symptom severity in a clinically diagnosed group rather than in a general pediatric group.

Previous studies on the effects of BPF and BPS are insufficient, and the results of previous studies on BPA exposure are inconsistent. In this study, no significant correlation was found for bisphenols, which supports previous findings that the relationship between bisphenols and ADHD symptoms is unclear.

Suggestions

As a result of the study, the commission error score in auditory ATA test does not have a clear correlation with EDC exposure. Brain MRI data were acquired during the recruitment of the research group. In future studies, by analyzing brain volumetry using these data or by obtaining and analyzing functional brain imaging data, it will be possible to clarify whether EDCs particularly affect visual-related brain areas. Additionally, if we analyze brain volumetry of the frontostriatal limbic system, it will be possible to identify the structural changes caused by EDCs exposure in the brain, in addition to the externalized symptoms revealed through neuropsychological tests. Furthermore, it would be meaningful to perform the same analysis on children with clinically diagnosed ASD in addition to ADHD among the cohort participants.

In the future, additional data could be collected to supplement the sample size and balance the sex ratio. The reliability of this study can be increased by confirming the reproducibility of the results through reanalysis.

In this study, we confirmed that phthalate exposure increases the visual ATA commission error T-score by eight times in children with clinically diagnosed ADHD. Thus, it is possible to infer that phthalate exposure affects impulsivity in children with ADHD, which is consistent with the results of previous studies obtained through parental surveys. Other neuropsychological test indicators, including inattention indicators or scores that comprehensively show symptom severity, such as the CGI-S and C-GAS, are not clearly associated with phthalate exposure. BPA, BPF, and BPS exposure do not affect the indicators of ADHD symptoms, which is consistent with the results of previous studies that did not show a consistent correlation with bisphenols.

None

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

The authors have no potential conflicts of interest to disclose.

Conceptualization: Bung-Nyun Kim. Data curation: Kang-Eun Yeo, Seungbee Lim, Aelin Kim. Formal analysis: Kang-Eun Yeo. Funding acquisition: Bung-Nyun Kim, Johanna Inhyang Kim. Investigation: Kang-Eun Yeo. Methodology: Bung-Nyun Kim, Kang-Eun Yeo, Johanna Inhyang Kim, You Bin Lim, Chae Rim Lee. Project administration: Bung-Nyun Kim, Johanna Inhyang Kim. Resources: Bung- Nyun Kim. Software: Kang-Eun Yeo. Supervision: Bung-Nyun Kim. Validation: Bung-Nyun Kim. Visualization: Kang-Eun Yeo. Writing—original draft: Kang-Eun Yeo. Writing—review & editing: Bung-Nyun Kim, Kang-Eun Yeo.

This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2020M3E5D9080787, No. NRF-2015M3C7A1028926).