Psychological Medicine, Page 1 of 12. doi:10.1017/S0033291714002694
OR I G I N A L A R T I C L E
© Cambridge University Press 2014
Association between phthalates and externalizing behaviors and cortical thickness in children with attention deficit hyperactivity disorder S. Park1, J.-M. Lee2, J.-W. Kim3,4, J. H. Cheong5, H. J. Yun2, Y.-C. Hong6, Y. Kim1, D. H. Han7, H.J. Yoo4,8, M.-S. Shin3,4, S.-C. Cho3,4 and B.-N. Kim3,4* 1
Department of Child and Adolescent Psychiatry, Seoul National Hospital, Seoul, Republic of Korea Department of Biomedical Engineering, Hanyang University, Seoul, Republic of Korea 3 Division of Child and Adolescent Psychiatry, Department of Psychiatry, Seoul National University Hospital, Seoul, Republic of Korea 4 College of Medicine and Behavioral Medicine Institute, Seoul National University, Seoul, Republic of Korea 5 Uimyung Research Institute for Neuroscience, Sahmyook University, Seoul, Republic of Korea 6 Department of Preventive Medicine, Seoul National University College of Medicine and Institute of Environmental Medicine, Seoul, Republic of Korea 7 Department of Psychiatry, Chung Ang University, College of Medicine, Seoul, Republic of Korea 8 Department of Psychiatry, Seoul National University Bung dang Hospital, Seongnam, Republic of Korea 2
Background. Previous studies have implicated the relationship between environmental phthalate exposure and attention deficit hyperactivity disorder (ADHD) symptoms of childhood, but no studies have been conducted in children who have a confirmed diagnosis of ADHD obtained through meticulous diagnostic testing. We aimed to determine whether phthalate metabolites in urine would be higher in children with ADHD than in those without ADHD and would correlate with symptom severity and cortical thickness in ADHD children. Method. A cross-sectional examination of urine phthalate metabolite concentrations was performed; scores for ADHD symptoms, externalizing problems, and continuous performance tests were obtained from 180 children with ADHD, and brain-imaging data were obtained from 115 participants. For the control group, children without ADHD (N = 438) were recruited. Correlations between phthalate metabolite concentrations and clinical measures and brain cortical thickness were investigated. Results. Concentrations of phthalate metabolites, particularly the di(2-ethylhexyl) phthalate (DEHP) metabolite, were significantly higher in boys with ADHD than in boys without ADHD. Concentrations of the di-n-butyl phthalate (DBP) metabolite were significantly higher in the combined or hyperactive-impulsive subtypes compared to the inattentive subtype, and the metabolite was positively correlated with the severity of externalizing symptoms. Concentrations of the DEHP metabolite were negatively correlated with cortical thickness in the right middle and superior temporal gyri. Conclusions. The results of this study suggest an association between phthalate concentrations and both the diagnosis and symptom severity of ADHD. Imaging findings suggest a negative impact of phthalates on regional cortical maturation in children with ADHD. Received 9 July 2014; Revised 18 October 2014; Accepted 20 October 2014 Key words: Attention-deficit hyperactivity disorder, cortical thickness, neuropsychology, phthalate.
Introduction Phthalates are a group of colorless, odorless liquids which are used as plasticizers or softeners. Their principal use is to soften polyvinyl chloride (PVC). Plasticized PVC is used in a wide range of commercial products, including food packaging, PVC tubing, medical tubing
* Address for correspondence: Dr B.-N. Kim, Division of Child and Adolescent Psychiatry, Department of Psychiatry, Seoul National University, College of Medicine, College of Medicine, 101 Daehakro, Chongro-Gu, Seoul, South Korea. (Email: [email protected])
and blood bags, toys, and cosmetics (Wormuth et al. 2006). Phthalates can leach into foods heated in plastic containers. Mouthing of toys containing phthalates can also result in phthalate exposure. Individuals receiving feedings or transfusions through medical tubing containing phthalates are likely to be exposed to phthalates (U.S. EPA, 2007). Di(2-ethylhexyl) phthalate (DEHP), di-n-butyl phthalate (DBP) and benzyl-butyl phthalate (BBP) are now classified as reprotoxic substances [i.e. category 1B according to Regulation (EC), category 2 according to Directive 67/548/EEC], and the use of these phthalates is prohibited in the production of toys and childcare articles [Registration, Evaluation,
2 S. Park et al. Authorization and Restriction of Chemicals (REACH) Annex XVII, 51]. Since 2004, the use of DEHP, DBP, and BBP has been prohibited in cosmetics (Directive 2004/93/EC) and in materials to come into contact with food since 2007 (Directive 2007/19/EC). Di-iso-nonyl phthalate (DINP), di-iso-decyl phthalate (DIDP), and di(n-octyl) phthalate (DNOP) are restricted only in toys that can be placed into the mouth. Phase 1 bio-transformations rapidly metabolize phthalates to their monoesters. Phthalate monoesters can be further metabolized by oxidation of the lipophilic aliphatic side-chain. Phthalate metabolites are excreted unchanged in urine and feces or as glucuronide conjugates in urine. Secondary oxidized DEHP metabolites like mono-2-ethylhexyl phthalate (MEHP), mono(2-ethyl-5-oxohexyl)phthalate (MEOP), and mono2-ethyl-5-hydroxyhexyl phthalate (MEHHP) are most valuable biomarkers of DEHP exposure. For DBP, the simple monoester, mono-n-butyl phthalate (MBP), is the major metabolite excreted in urine and used as a biomarker of DBP exposure (Latini, 2005). Elimination half-lives of the oxidized metabolites are longer than those of the monoesters. Koch and colleagues estimated an excretion half-life of 5 h for MEHP and 10–24 h for their oxidized metabolites (Koch et al. 2006) and 3.3 h for MBP and 4.5–10.3 h for their oxidized metabolites (Koch et al. 2012). Previous animal studies have reported that phthalate compounds might cause hyperactivity and impulsivity in rats (Masuo et al. 2004). In humans, Engel et al. (2010) reported that prenatal phthalate exposure was associated with problems with aggression, conduct, and attention, as well as with poor executive functioning and emotional control. Previously, we have reported an association between urine DEHP or DBP metabolite levels and attention deficit hyperactivity disorder (ADHD) symptoms (Kim et al. 2009) and low intelligence (Cho et al. 2010) in a cross-sectional study of community-dwelling Korean children. However, no studies have been conducted in children who have a confirmed diagnosis of ADHD obtained through meticulous diagnostic testing. At this point, additional research is needed to investigate the contribution of phthalates to the pathophysiological mechanisms leading to ADHD, related externalizing behaviors, and cognitive problems. First, case-control studies are needed to examine whether phthalate metabolite concentrations are higher in children with ADHD than in healthy controls. Second, studies of clinical sample of ADHD should be conducted to investigate which phenotypes of ADHD were best attributed to the phthalates. Third, neurobiological studies are required to analyze how phthalates affect the neural system, leading to cognitive and/or behavioral changes. Recent animal
toxicity studies have indicated that exposure to phthalates has a negative impact on the development of the hippocampus in male rats (Smith et al. 2011); however, the effects of phthalates on human brain development have not been established. In this study, three levels of analyses were conducted. First, we conducted a case-control study comparing DEHP or DBP metabolite levels between children with and without ADHD. Second, we investigated urine DEHP or DBP metabolite concentrations in relation to the clinical subtype of ADHD diagnosed or each of the individual behavioral measures conducted in a sample of children with ADHD. Third, we investigated the correlation between urine phthalate metabolite concentration and cortical thickness (CT). Based on the results of previous studies (Kim et al. 2009; Engel et al. 2010), we hypothesized that DEHP or DBP metabolites in urine would be higher in children with ADHD than in those without ADHD and that these metabolites would be associated with symptom severity. We also hypothesized that phthalate metabolite levels would be negatively correlated with CT in the right frontal and temporal lobes, which are the cortical regions most closely related to aggression and externalizing behaviors (Wahlund & Kristiansson, 2009; Fahim et al. 2011) and where CT has been found to be reduced in ADHD patients (Seidman et al. 2005; Makris et al. 2007; Almeida Montes et al. 2013).
Materials and method Participants and procedures We recruited 180 medication-naïve children with ADHD from the child psychiatric clinic of the Seoul National University Hospital in South Korea. The recruited children were aged between 6 and 15 years and had been diagnosed with ADHD according to DSM-IV criteria, as ascertained by a child psychiatrist and a semi-structured interview. To diagnose ADHD and any co-morbid disorder, we used the Korean Kiddie-Schedule for Affective Disorders and Schizophrenia – Present and Lifetime Version (K-SADS-PL; Kim et al. 2004). Exclusion criteria included the following: (1) a history of pervasive developmental disorder, mental retardation, bipolar disorder, psychotic disorder, obsessive compulsive disorder, or Tourette’s syndrome; (2) a history of organic brain disease, seizure disorder, or other neurological disorder; (3) an IQ < 70; (4) the presence of learning disabilities or language disorders; (5) the presence of major depressive disorder, anxiety disorder, or tic disorder requiring drug therapy; and (6) any previous course of methylphenidate treatment lasting more than 1 year or occurring within the last 4 weeks.
Phthalate exposure and brain change in ADHD children 3 The comparison group included participants of a research project entitled ‘The effects of pollution on neurobehavioral development and future policies to protect our children’, who were aged 8–11 years and recruited from ordinary primary schools in Korea from 2008 to 2009 (Cho et al. 2010). Phthalate metabolites have been known to be influenced by sex and age (CDC, U.S. Department of Health and Human Services, 2005). Because the patient and control groups differed with respect to age ranges and gender ratios, we used a selection method to match the age and gender distributions. First, among the 180 children with ADHD, 87 who were aged 8–11 years were selected. Among these 87 children with ADHD, only nine were female. Consequently, we excluded female patients entirely and compared the remaining 78 male patients with male controls. Among 473 boys from the control group who had participated in the diagnostic interview, 35 boys met DSM-IV criteria for ADHD (21 inattentive type, nine hyperactiveimpulsive type, five combined type). These boys were excluded from the control group and were used as a non-referred ADHD group. The remaining 438 boys comprised the non-ADHD group. The parents completed the Korean version of the ADHD Rating Scale-IV (ADHD-RS; So et al. 2002), the Disruptive Behavioral Disorder Rating Scale according to DSM-IV (DBDS; Silva et al. 2005), and the Children’s Behavior Checklist (CBCL; Oh et al. 1997). Certified child and adolescent psychiatrists who had completed inter-rater reliability training courses conducted the clinical assessments of children with ADHD using the Clinical Global Severity (CGI) scale (National Institute of Mental Health, 1970). The children with ADHD were administered the Korean version of the computer-based Continuous Performance Test (CPT; Greenberg & Waldman, 1993), named ADHD Diagnostic System (Shin et al. 2000). Raw data on the CPT variables were converted to aged-adjusted T scores. Higher T scores indicated poorer test performance. Among 180 children with ADHD, a subset of 115 participants completed all assessments, including magnetic resonance imaging (MRI). This subset did not differ from the original participant pool in age, sex, or clinical characteristics (Table 1). We provided detailed information about the study to the parents and children and then obtained written informed consent before any child entered the study. The study protocol was approved by the institutional review board of Seoul National University Hospital. Phthalate level measurements The spot urine samples were collected in 50 ml sterile specimen containers at the outpatient clinics in the
morning and were refrigerated at −20 °C immediately. Refrigerated specimens were transported to the laboratory within 2 h. The metabolites measured in this study included one primary metabolite of DBP, monon-butyl phthalate (MBP), and two secondary metabolites of DEHP, mono-2-ethylhexyl phthalate (MEHP) and mono(2-ethyl-5-oxohexyl)phthalate (MEOP). The urine phthalate metabolite concentrations were measured with high-performance liquid chromatography tandem mass spectrometry (Agilent 6410 triple Quad LCMS; Agilent, USA). Five hundred microliters of urine were buffered with 30 μl of 2.0 mol/l sodium acetate (pH 5.0) and then spiked with a mixture of isotope phthalate monoester standards (100 ng/ml) and 10 μl of β-glucuronidase. The sample was incubated at 37 °C for 3 h to deconjugate the glucuronidated phthalate metabolites. After incubation, 100 μl of 2 nmol/l hydrogen chloride was added to collect phthalate monoester. The extract was dried with nitrogen gas and reconstituted with 1 ml of high-performance liquid chromatography-grade water in a 2-ml glass vial. One blank and one quality control (QC) sample were included in each batch of samples. The QC sample was spiked with pooled urine and a mixture of phthalate monoester standard (100 ng/ml). The supernatants were purified by solid phase extraction with disposable Agilent C18 1.8 μm (2.1 × 50 mm). The mobile phase was 0.1% acetic acid water: 0.01% acetic acid acetonitrile (90:10, v/v) at a flow rate of 0.25 ml/ min, and the eluates were monitored at target masses of 221, 293, and 291 and internal standard masses of 225, 297, and 95. We used the value (μg/l)/creatinine (g/l) for dilution correction in the analyses. Because the concentrations were not normally distributed, we used their natural log-transformed values. Image acquisition and processing Whole-brain structural MRI was acquired with a T1-weighted magnetization-prepared rapid acquisition gradient echo (MPRAGE) scan on a 3 T Siemens scanner (Siemens Magnetom Trio Tim Syngo MR B17, Germany). Images were acquired with the following parameters: TR = 1900 ms, TE = 3.13 ms, inversion time 900 ms, flip angle = 9°, voxel size 0.9 mm3, FOV = 230 mm, slices 176. T1-weighted images were registered in the ICBM 152 average template using a linear transformation and corrected for intensity non-uniformity artifacts. The images were then classified into white matter (WM), gray matter (GM), cerebrospinal fluid (CSF) and background using an advanced neural net classifier. Hemispheric cortical surfaces were automatically extracted from each T1-weighted image using the Constrained Laplacianbased Automated Segmentation with Proximities
4 S. Park et al. Table 1. Demographic and clinical characteristics of total population of ADHD (N = 180) and subgroup of ADHD with brain imaging (N = 115)
Variable Gender, male/female, no. (%) Age, mean (S.D.), years ADHD subtype, no. (%) Combined Predominantly inattentive Predominantly hyperactive-impulsive Not otherwise specified Psychiatric co-morbidity, no. (%) Oppositional defiant disorder Anxiety disorder Tic disorder Enuresis Intelligence quotient, mean (S.D.) Parent ADHD rating scale, total, mean (S.D.) Clinical Global Impression – Severity, mean (S.D.) DBDS, total, mean (S.D.) CBCL, externalizing problems, mean (S.D.)
Total population (N = 180)
Subgroup with brain imaging (N = 115)
148/32 (82.2/17.8) 8.99 (2.40)
96/19 (83.5/16.5) 8.85 (2.32)
83 (46.1) 69 (38.3) 7 (3.9) 21 (11.7)
49 (42.6) 46 (40.0) 6 (5.2) 14 (12.2)
22 (12.2) 4 (2.2) 5 (2.8) 4 (2.2) 106.21 (14.43) 24.85 (10.71) 4.46 (0.72) 10.16 (7.87) 57.41 (8.40)
12 (10.4) 3 (2.6) 2 (1.7) 2 (1.7) 105.25 (14.19) 24.45 (11.00) 4.46 (0.72) 9.41 (7.32) 57.69 (8.59)
p value 0.781 0.621 0.910
0.639 >0.99 0.709 >0.99 0.575 0.757 >0.99 0.750 0.782
ADHD, Attention deficit hyperactivity disorder; DBDS, Disruptive Behavioral Disorder Rating Scale according to DSM-IV; CBCL, Child Behavior Checklist.
(CLASP) algorithm, which reconstructed the inner cortical surface by deforming a spherical mesh onto the WM/ GM boundary and then expanding the deformable model to the GM/CSF boundary (MacDonald et al. 2000). The reconstructed hemispheric cortical surfaces consisted of 40 962 vertices, each forming highresolution meshes. The inner and outer cortical surfaces had the same number of vertices, and there was a close correspondence between the counterpart vertices of the inner and outer cortical surfaces. CT was defined using the t-link method, which captures the Euclidean distance between these linked vertices (MacDonald et al. 2000). For group analysis, each individual thickness map was transformed to a surface group template using a 2-dimensional (2D) surface-based registration that aligns variable sulcal folding patterns through sphere-to-sphere warping (Lerch & Evans, 2005). Statistical analysis We conducted three principal sets of statistical analyses. First, we compared the urine phthalate metabolite (e.g. MEOP, MEHP, MBP) concentrations between clinic-referred boys with ADHD and boys without ADHD using independent-sample t tests and analysis of covariance (ANCOVA). In the ANCOVA, years of paternal and maternal education, both of which were
significantly different between groups, were used as covariates. We also compared the urine phthalate metabolite concentrations between non-referred boys with ADHD and boys without ADHD. Second, we conducted a comparison of the urine phthalate metabolite concentrations among the children with ADHD according to subtype. Because only seven children were hyperactive-impulsive subtype, we grouped the hyperactive-impulsive and combined subtypes together for the purpose of statistical analysis; these two subtypes of ADHD are known to have similar phenomenology and pathophysiology (Grizenko et al. 2010).Twenty-one children with ADHD, not otherwise specified were excluded from this analysis. We also conducted linear regression analyses to elucidate the associations of urine phthalate metabolite concentrations with ADHD symptoms, externalizing behaviors, and attentional deficits among children with ADHD. Regression analyses were performed using a set of covariates based on established predictors of children’s cognitive and behavioral functioning: age, gender, years of parental education, yearly income, subtype of ADHD, and IQ. The statistical analyses above were performed using SPSS v. 21.0 (SPSS Inc., USA), with statistical significance defined as alpha level < 0.05. Third, to investigate the correlation between brain CT and phthalate metabolite concentrations, multiple regression analyses were performed with urine
Phthalate exposure and brain change in ADHD children 5 phthalate metabolite concentrations, age, sex, and intracranial volume (ICV) as independent variables and each of the vertices of CT as a dependent variable. All the 81 924 vertices (40 962 per hemisphere) were used in statistical analysis. We utilized the SurfStat (K. Worsley, http://www.math.mcgill.ca/keith/surfstat/), which is a MATLAB toolbox (MathWorks Inc., USA) for the statistical analysis of univariate and multivariate surface data using linear mixed-effects models. The regression equation was defined as follows: Yb0 + b1 urine phthalate metabolite concentrations + b2 Age + b3 Sex + b4 ICV where Y is the CT, b0 is the Y intercept, b1–4 are the regression coefficients, and ε is the residual error. Then, we included the ADHD-RS score and the DBDS score as well as age, sex, and ICV as covariates. The regression equation was defined as follows: Yb0 + b1 urine phthalate metabolite concentrations + b2 Age + b3 Sex + b4 ICV + b5 ADHD
matched control boys without ADHD. The geometric mean concentrations of MEHP (44.46 ± 2.14 v. 25.49 ± 2.41), MEOP (43.65 ± 1.89 v. 20.71 ± 1.94), and MBP (66.00 ± 1.93 v. 50.31 ± 1.79) were significantly higher in boys with ADHD than in boys without ADHD (p < 0.001 each). ANCOVA revealed that all significant group differences that were found in the t tests persisted even after adjusting for years of paternal and maternal education, both of which were significantly different between groups (Table 2). To cross-validate these findings, we compared the urine phthalate metabolite concentrations between nonreferred boys with ADHD and boys without ADHD. The geometric mean concentrations of MEHP (29.68 ± 1.72 v. 25.49 ± 2.41, p = 0.158) and MEOP (23.79 ± 1.83 v. 20.71 ± 1.94, p = 0.230) were not significantly different between two groups. The geometric mean concentrations of MBP (60.13 ± 1.98 v. 50.31 ± 1.79, p = 0.085) were higher, although not statistically significant, in non-referred boys with ADHD than in boys without ADHD.
− RS + b6 DBDS + ε The results were thresholded at an uncorrected p value of 0.001 and at a false discovery rate (FDR)-corrected p value of 0.05. Results Characteristics and urine phthalate metabolite concentrations among children with ADHD Table 1 shows the demographic and clinical characteristics of 180 clinic-referred children with ADHD (148 males and 32 females, age 6–15 years, mean age 8.99 ± 2.40 years). The combined type (46.1%) was the most common in our sample, followed by the inattentive type (38.3%). The ADHD, hyperactive-impulsive type and the ADHD, not otherwise specified (NOS) type were identified in 3.9% and 11.7% of the sample, respectively. The most common co-morbidity was oppositional defiant disorder (ODD) (12.2%), followed by tic disorder (2.8%). Anxiety disorder and tic disorder were found in 2.2% each of the sample. The geometric mean (ln) concentrations of creatinine-corrected MEHP, MEOP, and MBP were 45.60 μg/g [geometric S.D. (GSD) = 2.41, range 2.95– 892.65], 43.82 μg/g (GSD = 2.18, range 3.70–785.44), and 68.03 μg/g (GSD = 2.10, range 91.90–109.76), respectively. Phthalate metabolite concentration differences between boys with and without ADHD Table 2 shows comparisons between the clinic-referred boys with ADHD (age 8–11 years) and the age-
Phthalate metabolite concentrations according to the subtype of ADHD Table 3 shows comparison of phthalate metabolite concentrations according to the subtype of ADHD. We found that the children with the combined or hyperactive-impulsive subtype of ADHD exhibited higher MEOP and MBP concentrations compared to the children with the inattentive subtype (GM ± GSD: 49.38 ± 2.03 v. 38.08 ± 2.36 for MEOP, p = 0.033 and 76.67 ± 2.01 v. 55.68 ± 2.10 for MBP, p = 0.006, respectively). Associations between phthalate metabolite concentrations and ADHD symptoms, externalizing problems, and CPT scores Table 4 shows associations between phthalate metabolite concentrations and clinical and neuropsychological variables among 180 children with ADHD. After adjusting for age, gender, years of paternal and maternal education, yearly income, subtype, and IQ, significant associations were found between urine DEHP or DBP metabolite concentrations and the CGI-S [B = 0.50, 95% confidence interval (CI) −0.10 to 0.90 for DEHP and B = 0.53, 95% CI 0.10–0.96 for DBP], ODD (B = 4.12, 95% CI 1.23–7.02 for DEHP and B = 6.11, 95% CI 3.03–9.10 for DBP), and total DBDS scores (B = 5.54, 95% CI 1.52–9.55 for DEHP and B = 7.56, 95% CI 3.24–11.87 for DBP). Urine DBP metabolite concentrations were also significantly associated with the aggressive behavior (B = 6.16, 95% CI 1.47–10.85) and externalizing problem (B = 2.16, 95% CI 0.46–10.22) scores. Urine DEHP metabolite
6 S. Park et al. Table 2. Comparison of boys with ADHD and boys without ADHD (age 8–11 years)
Variable
Clinic-referred boys with ADHD (N = 78)a
Boys without ADHD (N = 438)b
9.10 (1.06) 108.78 (15.43) 14.81 (1.84) 14.70 (2.12) 50 (74.6) 25.42 (10.52)
9.09 (0.70) 112.25 (14.58) 13.82 (2.25) 13.26 (2.14) 259 (62.7) 9.51 (7.74)
8.59 (5.50) 1.39 (2.28) 54.93 (7.77) 57.78 (8.64) 57.25 (8.68)
Age, mean (S.D.), years Intelligence quotient, mean (S.D.) Paternal years of education, mean (S.D.) Maternal years of education, mean (S.D.) Yearly income > $25 000, no. (%) ADHD rating scale, mean (S.D.) DBDS, mean (S.D.) Oppositional defiant disorder score Conduct disorder score CBCL, mean (S.D.) Delinquent behavior Aggressive behavior Externalizing problems ADHD subtype, no. (%) Combined Predominantly inattentive Predominantly hyperactive-impulsive Not otherwise specified
42 (53.8) 24 (30.8) 3 (3.8) 9 (11.5)
Phthalate concentrations MEHP MEOP MEHP+MEOP MBP
Adjusted GM (95% CI)a,c 48.18 (41.14–56.43) 43.99 (37.34–51.88) 93.13 (79.60–108.85) 65.96 (56.83–76.63)
t
p
0.13 −1.88 3.92 5.10 3.57 12.19
0.900 0.061
OR I G I N A L A R T I C L E
© Cambridge University Press 2014
Association between phthalates and externalizing behaviors and cortical thickness in children with attention deficit hyperactivity disorder S. Park1, J.-M. Lee2, J.-W. Kim3,4, J. H. Cheong5, H. J. Yun2, Y.-C. Hong6, Y. Kim1, D. H. Han7, H.J. Yoo4,8, M.-S. Shin3,4, S.-C. Cho3,4 and B.-N. Kim3,4* 1
Department of Child and Adolescent Psychiatry, Seoul National Hospital, Seoul, Republic of Korea Department of Biomedical Engineering, Hanyang University, Seoul, Republic of Korea 3 Division of Child and Adolescent Psychiatry, Department of Psychiatry, Seoul National University Hospital, Seoul, Republic of Korea 4 College of Medicine and Behavioral Medicine Institute, Seoul National University, Seoul, Republic of Korea 5 Uimyung Research Institute for Neuroscience, Sahmyook University, Seoul, Republic of Korea 6 Department of Preventive Medicine, Seoul National University College of Medicine and Institute of Environmental Medicine, Seoul, Republic of Korea 7 Department of Psychiatry, Chung Ang University, College of Medicine, Seoul, Republic of Korea 8 Department of Psychiatry, Seoul National University Bung dang Hospital, Seongnam, Republic of Korea 2
Background. Previous studies have implicated the relationship between environmental phthalate exposure and attention deficit hyperactivity disorder (ADHD) symptoms of childhood, but no studies have been conducted in children who have a confirmed diagnosis of ADHD obtained through meticulous diagnostic testing. We aimed to determine whether phthalate metabolites in urine would be higher in children with ADHD than in those without ADHD and would correlate with symptom severity and cortical thickness in ADHD children. Method. A cross-sectional examination of urine phthalate metabolite concentrations was performed; scores for ADHD symptoms, externalizing problems, and continuous performance tests were obtained from 180 children with ADHD, and brain-imaging data were obtained from 115 participants. For the control group, children without ADHD (N = 438) were recruited. Correlations between phthalate metabolite concentrations and clinical measures and brain cortical thickness were investigated. Results. Concentrations of phthalate metabolites, particularly the di(2-ethylhexyl) phthalate (DEHP) metabolite, were significantly higher in boys with ADHD than in boys without ADHD. Concentrations of the di-n-butyl phthalate (DBP) metabolite were significantly higher in the combined or hyperactive-impulsive subtypes compared to the inattentive subtype, and the metabolite was positively correlated with the severity of externalizing symptoms. Concentrations of the DEHP metabolite were negatively correlated with cortical thickness in the right middle and superior temporal gyri. Conclusions. The results of this study suggest an association between phthalate concentrations and both the diagnosis and symptom severity of ADHD. Imaging findings suggest a negative impact of phthalates on regional cortical maturation in children with ADHD. Received 9 July 2014; Revised 18 October 2014; Accepted 20 October 2014 Key words: Attention-deficit hyperactivity disorder, cortical thickness, neuropsychology, phthalate.
Introduction Phthalates are a group of colorless, odorless liquids which are used as plasticizers or softeners. Their principal use is to soften polyvinyl chloride (PVC). Plasticized PVC is used in a wide range of commercial products, including food packaging, PVC tubing, medical tubing
* Address for correspondence: Dr B.-N. Kim, Division of Child and Adolescent Psychiatry, Department of Psychiatry, Seoul National University, College of Medicine, College of Medicine, 101 Daehakro, Chongro-Gu, Seoul, South Korea. (Email: [email protected])
and blood bags, toys, and cosmetics (Wormuth et al. 2006). Phthalates can leach into foods heated in plastic containers. Mouthing of toys containing phthalates can also result in phthalate exposure. Individuals receiving feedings or transfusions through medical tubing containing phthalates are likely to be exposed to phthalates (U.S. EPA, 2007). Di(2-ethylhexyl) phthalate (DEHP), di-n-butyl phthalate (DBP) and benzyl-butyl phthalate (BBP) are now classified as reprotoxic substances [i.e. category 1B according to Regulation (EC), category 2 according to Directive 67/548/EEC], and the use of these phthalates is prohibited in the production of toys and childcare articles [Registration, Evaluation,
2 S. Park et al. Authorization and Restriction of Chemicals (REACH) Annex XVII, 51]. Since 2004, the use of DEHP, DBP, and BBP has been prohibited in cosmetics (Directive 2004/93/EC) and in materials to come into contact with food since 2007 (Directive 2007/19/EC). Di-iso-nonyl phthalate (DINP), di-iso-decyl phthalate (DIDP), and di(n-octyl) phthalate (DNOP) are restricted only in toys that can be placed into the mouth. Phase 1 bio-transformations rapidly metabolize phthalates to their monoesters. Phthalate monoesters can be further metabolized by oxidation of the lipophilic aliphatic side-chain. Phthalate metabolites are excreted unchanged in urine and feces or as glucuronide conjugates in urine. Secondary oxidized DEHP metabolites like mono-2-ethylhexyl phthalate (MEHP), mono(2-ethyl-5-oxohexyl)phthalate (MEOP), and mono2-ethyl-5-hydroxyhexyl phthalate (MEHHP) are most valuable biomarkers of DEHP exposure. For DBP, the simple monoester, mono-n-butyl phthalate (MBP), is the major metabolite excreted in urine and used as a biomarker of DBP exposure (Latini, 2005). Elimination half-lives of the oxidized metabolites are longer than those of the monoesters. Koch and colleagues estimated an excretion half-life of 5 h for MEHP and 10–24 h for their oxidized metabolites (Koch et al. 2006) and 3.3 h for MBP and 4.5–10.3 h for their oxidized metabolites (Koch et al. 2012). Previous animal studies have reported that phthalate compounds might cause hyperactivity and impulsivity in rats (Masuo et al. 2004). In humans, Engel et al. (2010) reported that prenatal phthalate exposure was associated with problems with aggression, conduct, and attention, as well as with poor executive functioning and emotional control. Previously, we have reported an association between urine DEHP or DBP metabolite levels and attention deficit hyperactivity disorder (ADHD) symptoms (Kim et al. 2009) and low intelligence (Cho et al. 2010) in a cross-sectional study of community-dwelling Korean children. However, no studies have been conducted in children who have a confirmed diagnosis of ADHD obtained through meticulous diagnostic testing. At this point, additional research is needed to investigate the contribution of phthalates to the pathophysiological mechanisms leading to ADHD, related externalizing behaviors, and cognitive problems. First, case-control studies are needed to examine whether phthalate metabolite concentrations are higher in children with ADHD than in healthy controls. Second, studies of clinical sample of ADHD should be conducted to investigate which phenotypes of ADHD were best attributed to the phthalates. Third, neurobiological studies are required to analyze how phthalates affect the neural system, leading to cognitive and/or behavioral changes. Recent animal
toxicity studies have indicated that exposure to phthalates has a negative impact on the development of the hippocampus in male rats (Smith et al. 2011); however, the effects of phthalates on human brain development have not been established. In this study, three levels of analyses were conducted. First, we conducted a case-control study comparing DEHP or DBP metabolite levels between children with and without ADHD. Second, we investigated urine DEHP or DBP metabolite concentrations in relation to the clinical subtype of ADHD diagnosed or each of the individual behavioral measures conducted in a sample of children with ADHD. Third, we investigated the correlation between urine phthalate metabolite concentration and cortical thickness (CT). Based on the results of previous studies (Kim et al. 2009; Engel et al. 2010), we hypothesized that DEHP or DBP metabolites in urine would be higher in children with ADHD than in those without ADHD and that these metabolites would be associated with symptom severity. We also hypothesized that phthalate metabolite levels would be negatively correlated with CT in the right frontal and temporal lobes, which are the cortical regions most closely related to aggression and externalizing behaviors (Wahlund & Kristiansson, 2009; Fahim et al. 2011) and where CT has been found to be reduced in ADHD patients (Seidman et al. 2005; Makris et al. 2007; Almeida Montes et al. 2013).
Materials and method Participants and procedures We recruited 180 medication-naïve children with ADHD from the child psychiatric clinic of the Seoul National University Hospital in South Korea. The recruited children were aged between 6 and 15 years and had been diagnosed with ADHD according to DSM-IV criteria, as ascertained by a child psychiatrist and a semi-structured interview. To diagnose ADHD and any co-morbid disorder, we used the Korean Kiddie-Schedule for Affective Disorders and Schizophrenia – Present and Lifetime Version (K-SADS-PL; Kim et al. 2004). Exclusion criteria included the following: (1) a history of pervasive developmental disorder, mental retardation, bipolar disorder, psychotic disorder, obsessive compulsive disorder, or Tourette’s syndrome; (2) a history of organic brain disease, seizure disorder, or other neurological disorder; (3) an IQ < 70; (4) the presence of learning disabilities or language disorders; (5) the presence of major depressive disorder, anxiety disorder, or tic disorder requiring drug therapy; and (6) any previous course of methylphenidate treatment lasting more than 1 year or occurring within the last 4 weeks.
Phthalate exposure and brain change in ADHD children 3 The comparison group included participants of a research project entitled ‘The effects of pollution on neurobehavioral development and future policies to protect our children’, who were aged 8–11 years and recruited from ordinary primary schools in Korea from 2008 to 2009 (Cho et al. 2010). Phthalate metabolites have been known to be influenced by sex and age (CDC, U.S. Department of Health and Human Services, 2005). Because the patient and control groups differed with respect to age ranges and gender ratios, we used a selection method to match the age and gender distributions. First, among the 180 children with ADHD, 87 who were aged 8–11 years were selected. Among these 87 children with ADHD, only nine were female. Consequently, we excluded female patients entirely and compared the remaining 78 male patients with male controls. Among 473 boys from the control group who had participated in the diagnostic interview, 35 boys met DSM-IV criteria for ADHD (21 inattentive type, nine hyperactiveimpulsive type, five combined type). These boys were excluded from the control group and were used as a non-referred ADHD group. The remaining 438 boys comprised the non-ADHD group. The parents completed the Korean version of the ADHD Rating Scale-IV (ADHD-RS; So et al. 2002), the Disruptive Behavioral Disorder Rating Scale according to DSM-IV (DBDS; Silva et al. 2005), and the Children’s Behavior Checklist (CBCL; Oh et al. 1997). Certified child and adolescent psychiatrists who had completed inter-rater reliability training courses conducted the clinical assessments of children with ADHD using the Clinical Global Severity (CGI) scale (National Institute of Mental Health, 1970). The children with ADHD were administered the Korean version of the computer-based Continuous Performance Test (CPT; Greenberg & Waldman, 1993), named ADHD Diagnostic System (Shin et al. 2000). Raw data on the CPT variables were converted to aged-adjusted T scores. Higher T scores indicated poorer test performance. Among 180 children with ADHD, a subset of 115 participants completed all assessments, including magnetic resonance imaging (MRI). This subset did not differ from the original participant pool in age, sex, or clinical characteristics (Table 1). We provided detailed information about the study to the parents and children and then obtained written informed consent before any child entered the study. The study protocol was approved by the institutional review board of Seoul National University Hospital. Phthalate level measurements The spot urine samples were collected in 50 ml sterile specimen containers at the outpatient clinics in the
morning and were refrigerated at −20 °C immediately. Refrigerated specimens were transported to the laboratory within 2 h. The metabolites measured in this study included one primary metabolite of DBP, monon-butyl phthalate (MBP), and two secondary metabolites of DEHP, mono-2-ethylhexyl phthalate (MEHP) and mono(2-ethyl-5-oxohexyl)phthalate (MEOP). The urine phthalate metabolite concentrations were measured with high-performance liquid chromatography tandem mass spectrometry (Agilent 6410 triple Quad LCMS; Agilent, USA). Five hundred microliters of urine were buffered with 30 μl of 2.0 mol/l sodium acetate (pH 5.0) and then spiked with a mixture of isotope phthalate monoester standards (100 ng/ml) and 10 μl of β-glucuronidase. The sample was incubated at 37 °C for 3 h to deconjugate the glucuronidated phthalate metabolites. After incubation, 100 μl of 2 nmol/l hydrogen chloride was added to collect phthalate monoester. The extract was dried with nitrogen gas and reconstituted with 1 ml of high-performance liquid chromatography-grade water in a 2-ml glass vial. One blank and one quality control (QC) sample were included in each batch of samples. The QC sample was spiked with pooled urine and a mixture of phthalate monoester standard (100 ng/ml). The supernatants were purified by solid phase extraction with disposable Agilent C18 1.8 μm (2.1 × 50 mm). The mobile phase was 0.1% acetic acid water: 0.01% acetic acid acetonitrile (90:10, v/v) at a flow rate of 0.25 ml/ min, and the eluates were monitored at target masses of 221, 293, and 291 and internal standard masses of 225, 297, and 95. We used the value (μg/l)/creatinine (g/l) for dilution correction in the analyses. Because the concentrations were not normally distributed, we used their natural log-transformed values. Image acquisition and processing Whole-brain structural MRI was acquired with a T1-weighted magnetization-prepared rapid acquisition gradient echo (MPRAGE) scan on a 3 T Siemens scanner (Siemens Magnetom Trio Tim Syngo MR B17, Germany). Images were acquired with the following parameters: TR = 1900 ms, TE = 3.13 ms, inversion time 900 ms, flip angle = 9°, voxel size 0.9 mm3, FOV = 230 mm, slices 176. T1-weighted images were registered in the ICBM 152 average template using a linear transformation and corrected for intensity non-uniformity artifacts. The images were then classified into white matter (WM), gray matter (GM), cerebrospinal fluid (CSF) and background using an advanced neural net classifier. Hemispheric cortical surfaces were automatically extracted from each T1-weighted image using the Constrained Laplacianbased Automated Segmentation with Proximities
4 S. Park et al. Table 1. Demographic and clinical characteristics of total population of ADHD (N = 180) and subgroup of ADHD with brain imaging (N = 115)
Variable Gender, male/female, no. (%) Age, mean (S.D.), years ADHD subtype, no. (%) Combined Predominantly inattentive Predominantly hyperactive-impulsive Not otherwise specified Psychiatric co-morbidity, no. (%) Oppositional defiant disorder Anxiety disorder Tic disorder Enuresis Intelligence quotient, mean (S.D.) Parent ADHD rating scale, total, mean (S.D.) Clinical Global Impression – Severity, mean (S.D.) DBDS, total, mean (S.D.) CBCL, externalizing problems, mean (S.D.)
Total population (N = 180)
Subgroup with brain imaging (N = 115)
148/32 (82.2/17.8) 8.99 (2.40)
96/19 (83.5/16.5) 8.85 (2.32)
83 (46.1) 69 (38.3) 7 (3.9) 21 (11.7)
49 (42.6) 46 (40.0) 6 (5.2) 14 (12.2)
22 (12.2) 4 (2.2) 5 (2.8) 4 (2.2) 106.21 (14.43) 24.85 (10.71) 4.46 (0.72) 10.16 (7.87) 57.41 (8.40)
12 (10.4) 3 (2.6) 2 (1.7) 2 (1.7) 105.25 (14.19) 24.45 (11.00) 4.46 (0.72) 9.41 (7.32) 57.69 (8.59)
p value 0.781 0.621 0.910
0.639 >0.99 0.709 >0.99 0.575 0.757 >0.99 0.750 0.782
ADHD, Attention deficit hyperactivity disorder; DBDS, Disruptive Behavioral Disorder Rating Scale according to DSM-IV; CBCL, Child Behavior Checklist.
(CLASP) algorithm, which reconstructed the inner cortical surface by deforming a spherical mesh onto the WM/ GM boundary and then expanding the deformable model to the GM/CSF boundary (MacDonald et al. 2000). The reconstructed hemispheric cortical surfaces consisted of 40 962 vertices, each forming highresolution meshes. The inner and outer cortical surfaces had the same number of vertices, and there was a close correspondence between the counterpart vertices of the inner and outer cortical surfaces. CT was defined using the t-link method, which captures the Euclidean distance between these linked vertices (MacDonald et al. 2000). For group analysis, each individual thickness map was transformed to a surface group template using a 2-dimensional (2D) surface-based registration that aligns variable sulcal folding patterns through sphere-to-sphere warping (Lerch & Evans, 2005). Statistical analysis We conducted three principal sets of statistical analyses. First, we compared the urine phthalate metabolite (e.g. MEOP, MEHP, MBP) concentrations between clinic-referred boys with ADHD and boys without ADHD using independent-sample t tests and analysis of covariance (ANCOVA). In the ANCOVA, years of paternal and maternal education, both of which were
significantly different between groups, were used as covariates. We also compared the urine phthalate metabolite concentrations between non-referred boys with ADHD and boys without ADHD. Second, we conducted a comparison of the urine phthalate metabolite concentrations among the children with ADHD according to subtype. Because only seven children were hyperactive-impulsive subtype, we grouped the hyperactive-impulsive and combined subtypes together for the purpose of statistical analysis; these two subtypes of ADHD are known to have similar phenomenology and pathophysiology (Grizenko et al. 2010).Twenty-one children with ADHD, not otherwise specified were excluded from this analysis. We also conducted linear regression analyses to elucidate the associations of urine phthalate metabolite concentrations with ADHD symptoms, externalizing behaviors, and attentional deficits among children with ADHD. Regression analyses were performed using a set of covariates based on established predictors of children’s cognitive and behavioral functioning: age, gender, years of parental education, yearly income, subtype of ADHD, and IQ. The statistical analyses above were performed using SPSS v. 21.0 (SPSS Inc., USA), with statistical significance defined as alpha level < 0.05. Third, to investigate the correlation between brain CT and phthalate metabolite concentrations, multiple regression analyses were performed with urine
Phthalate exposure and brain change in ADHD children 5 phthalate metabolite concentrations, age, sex, and intracranial volume (ICV) as independent variables and each of the vertices of CT as a dependent variable. All the 81 924 vertices (40 962 per hemisphere) were used in statistical analysis. We utilized the SurfStat (K. Worsley, http://www.math.mcgill.ca/keith/surfstat/), which is a MATLAB toolbox (MathWorks Inc., USA) for the statistical analysis of univariate and multivariate surface data using linear mixed-effects models. The regression equation was defined as follows: Yb0 + b1 urine phthalate metabolite concentrations + b2 Age + b3 Sex + b4 ICV where Y is the CT, b0 is the Y intercept, b1–4 are the regression coefficients, and ε is the residual error. Then, we included the ADHD-RS score and the DBDS score as well as age, sex, and ICV as covariates. The regression equation was defined as follows: Yb0 + b1 urine phthalate metabolite concentrations + b2 Age + b3 Sex + b4 ICV + b5 ADHD
matched control boys without ADHD. The geometric mean concentrations of MEHP (44.46 ± 2.14 v. 25.49 ± 2.41), MEOP (43.65 ± 1.89 v. 20.71 ± 1.94), and MBP (66.00 ± 1.93 v. 50.31 ± 1.79) were significantly higher in boys with ADHD than in boys without ADHD (p < 0.001 each). ANCOVA revealed that all significant group differences that were found in the t tests persisted even after adjusting for years of paternal and maternal education, both of which were significantly different between groups (Table 2). To cross-validate these findings, we compared the urine phthalate metabolite concentrations between nonreferred boys with ADHD and boys without ADHD. The geometric mean concentrations of MEHP (29.68 ± 1.72 v. 25.49 ± 2.41, p = 0.158) and MEOP (23.79 ± 1.83 v. 20.71 ± 1.94, p = 0.230) were not significantly different between two groups. The geometric mean concentrations of MBP (60.13 ± 1.98 v. 50.31 ± 1.79, p = 0.085) were higher, although not statistically significant, in non-referred boys with ADHD than in boys without ADHD.
− RS + b6 DBDS + ε The results were thresholded at an uncorrected p value of 0.001 and at a false discovery rate (FDR)-corrected p value of 0.05. Results Characteristics and urine phthalate metabolite concentrations among children with ADHD Table 1 shows the demographic and clinical characteristics of 180 clinic-referred children with ADHD (148 males and 32 females, age 6–15 years, mean age 8.99 ± 2.40 years). The combined type (46.1%) was the most common in our sample, followed by the inattentive type (38.3%). The ADHD, hyperactive-impulsive type and the ADHD, not otherwise specified (NOS) type were identified in 3.9% and 11.7% of the sample, respectively. The most common co-morbidity was oppositional defiant disorder (ODD) (12.2%), followed by tic disorder (2.8%). Anxiety disorder and tic disorder were found in 2.2% each of the sample. The geometric mean (ln) concentrations of creatinine-corrected MEHP, MEOP, and MBP were 45.60 μg/g [geometric S.D. (GSD) = 2.41, range 2.95– 892.65], 43.82 μg/g (GSD = 2.18, range 3.70–785.44), and 68.03 μg/g (GSD = 2.10, range 91.90–109.76), respectively. Phthalate metabolite concentration differences between boys with and without ADHD Table 2 shows comparisons between the clinic-referred boys with ADHD (age 8–11 years) and the age-
Phthalate metabolite concentrations according to the subtype of ADHD Table 3 shows comparison of phthalate metabolite concentrations according to the subtype of ADHD. We found that the children with the combined or hyperactive-impulsive subtype of ADHD exhibited higher MEOP and MBP concentrations compared to the children with the inattentive subtype (GM ± GSD: 49.38 ± 2.03 v. 38.08 ± 2.36 for MEOP, p = 0.033 and 76.67 ± 2.01 v. 55.68 ± 2.10 for MBP, p = 0.006, respectively). Associations between phthalate metabolite concentrations and ADHD symptoms, externalizing problems, and CPT scores Table 4 shows associations between phthalate metabolite concentrations and clinical and neuropsychological variables among 180 children with ADHD. After adjusting for age, gender, years of paternal and maternal education, yearly income, subtype, and IQ, significant associations were found between urine DEHP or DBP metabolite concentrations and the CGI-S [B = 0.50, 95% confidence interval (CI) −0.10 to 0.90 for DEHP and B = 0.53, 95% CI 0.10–0.96 for DBP], ODD (B = 4.12, 95% CI 1.23–7.02 for DEHP and B = 6.11, 95% CI 3.03–9.10 for DBP), and total DBDS scores (B = 5.54, 95% CI 1.52–9.55 for DEHP and B = 7.56, 95% CI 3.24–11.87 for DBP). Urine DBP metabolite concentrations were also significantly associated with the aggressive behavior (B = 6.16, 95% CI 1.47–10.85) and externalizing problem (B = 2.16, 95% CI 0.46–10.22) scores. Urine DEHP metabolite
6 S. Park et al. Table 2. Comparison of boys with ADHD and boys without ADHD (age 8–11 years)
Variable
Clinic-referred boys with ADHD (N = 78)a
Boys without ADHD (N = 438)b
9.10 (1.06) 108.78 (15.43) 14.81 (1.84) 14.70 (2.12) 50 (74.6) 25.42 (10.52)
9.09 (0.70) 112.25 (14.58) 13.82 (2.25) 13.26 (2.14) 259 (62.7) 9.51 (7.74)
8.59 (5.50) 1.39 (2.28) 54.93 (7.77) 57.78 (8.64) 57.25 (8.68)
Age, mean (S.D.), years Intelligence quotient, mean (S.D.) Paternal years of education, mean (S.D.) Maternal years of education, mean (S.D.) Yearly income > $25 000, no. (%) ADHD rating scale, mean (S.D.) DBDS, mean (S.D.) Oppositional defiant disorder score Conduct disorder score CBCL, mean (S.D.) Delinquent behavior Aggressive behavior Externalizing problems ADHD subtype, no. (%) Combined Predominantly inattentive Predominantly hyperactive-impulsive Not otherwise specified
42 (53.8) 24 (30.8) 3 (3.8) 9 (11.5)
Phthalate concentrations MEHP MEOP MEHP+MEOP MBP
Adjusted GM (95% CI)a,c 48.18 (41.14–56.43) 43.99 (37.34–51.88) 93.13 (79.60–108.85) 65.96 (56.83–76.63)
t
p
0.13 −1.88 3.92 5.10 3.57 12.19
0.900 0.061
