Effects of Atomoxetine in Individuals with Attention-Deficit/Hyperactivity Disorder and Low-Functioning Autism Spectrum Disorder

Abstract

Objectives:

This naturalistic, retrospective study investigated the effects of atomoxetine (ATX) on attention-deficit/hyperactivity disorder (ADHD) symptoms and autistic features in children with autism spectrum disorders (ASDs) and intellectual disability (ID).

Methods:

Participants (n = 37, age range 6–17 years, mean: 10.16 ± 3.60) were assessed at baseline, 4th and 12th weeks using Clinical Global Impressions (CGI) scales, DSM-IV-based ADHD-rating scale (ADHD-RS), and amended Turkish version of Aberrant Behavior Checklist (ABC). The primary outcome measure was a treatment response defined by a CGI-improvement score of 1 or 2 together with a decrease of at least 25% in the parent-rated ADHD-RS total score at the end of 12th week.

Results:

Five patients (13.5%) stopped medication at 4 weeks due to ineffectivity (2) and intolerable side effects (increased motor activity and talkativeness [n = 1], irritability [n = 2], temper outbursts [n = 2], and increased blood pressure [n = 1]). Sixteen patients (43.2%) were judged to be responders according to primary outcome measure. Improvement rate on CGI scale was 48.8%. On ADHD-RS, there were significant reductions between baseline and 4th week and between baseline and 12th week in both hyperactivity and inattention, and between baseline and 12th week in impulsivity scores. Decrease was significant in hyperactivity and social withdrawal subscales of the parent-reported ABC. Responders based on primary outcome measure were not significantly different from nonresponders in terms of sociodemographic features or clinical parameters, including intellectual, language, autism symptom, and ADHD symptom levels.

Conclusion:

In this chart review, ATX appears to be safe and effective for social withdrawal and ADHD symptoms in children with ASD and ID.

Introduction

Autism spectrum disorders (ASDs) are characterized by impaired communication and social interaction skills, as well as repetitive and restricted behavior and interests. Attention-deficit/hyperactivity disorder (ADHD) is defined by impairing symptoms of inattention, hyperactivity, and impulsivity. Recent studies using clinical samples revealed that 41%–78% of children with ASDs also meet diagnostic criteria for ADHD (Murray 2010). The symptoms of ADHD can be reliably distinguished from those of ASDs (Gadow et al. 2006), and individuals with co-occurring ASDs and ADHD are more severely impaired (Murray 2010). Coexisting ADHD was demonstrated to be associated with higher rates of difficulties in social processing, adaptive functioning, and executive control in addition to increased autistic traits and general psychopathology in children with ASDs (Holtmann et al. 2007; Yerys et al. 2009). Treatment of the ADHD symptoms may help to improve their educational achievement, socialization, and behavioral management (Reichow et al. 2013).

Psychostimulants are the most extensively studied medications in children with comorbid ASD and ADHD (ASD+ADHD). A fourth of the patients with ASDs who applied to clinical services in the United States were prescribed stimulants during the year 2002 (Oswald and Sonenklar 2007), and a recent study in Denmark showed that prescription rates were increasing very rapidly in that country (Dalsgaard et al. 2013). However, the ASD+ADHD population on stimulants have lower response and higher discontinuation rates due to adverse effects than the non-ASD+ADHD population (Aman et al. 2008). Decreased appetite and higher levels of social withdrawal, insomnia, irritability, and depressive symptoms are frequently reported in children with ASD+ADHD (Reichow et al. 2013).

Atomoxetine (ATX) is a norepinephrine reuptake inhibitor approved for ADHD treatment in children older than 6 years. As summarized in Table 1, studies found that 43%–75% of the subjects show improvement in ADHD symptoms with ATX (Jou et al. 2005; Arnold et al. 2006; Posey et al. 2006; Troost et al. 2006; Zeiner et al. 2011; Harfterkamp et al. 2012; Fernandez-Jaen et al. 2013). These were all small-scale studies with 12–24 patients except the one conducted by Harfterkamp et al. (2012). Most of the studies reported that ATX is effective in controlling ADHD symptoms. The only study with negative results was performed on children with severe autistic symptoms (Charnsil 2011); however, the doses administered were lower than the suggested doses due to greater side effects.

Table 1.

Summary of Studies of Atomoxetine in Patients with Autism Spectrum Disorders and Attention-Deficit/Hyperactivity Disorder

	Study design	Subjects	Main outcome measures	Response	Premature termination
Jou et al. (2005)	Retrospective (1–36 weeks)	20 children, mean age: 11.5 (10 with ID)	CGI-I, CPRS	CGI-I ≤2: 60%	1 patient (5%) (severe mood swings)
Posey et al. (2006)	Open label (8 weeks)	16 children, mean age: 7.7 (all IQ >70)	CGI-I, SNAP-IV, VABS	CGI-I ≤2: 75%	3 patients (19%) (irritability and emergence of inability to swallow pills)
Troost et al. (2006)	Open label (10 weeks)	12 children, mean age: 10.2 (all IQ >70)	CGI-ADHD, ADHD-RS, CPRS, ABC	CGI-ADHD-I ≤2: 58%	5 patients (42%) (GIS symptoms, anxiety, agitation)
Arnold et al. (2006)	Placebo controlled crossover, (6 weeks each)	16 children, mean age 9.3 (intellectual age ≥1.5 years)	CGI-I, hyperactivity score of ABC, SNAP-IV	CGI-I ≤2 + 25% decrease in H-ABC: 57% with ATX, 25% placebo	1 patient (6%) at 4 weeks (aggressive behavior)
Charnsil (2011)	Open label (10 weeks)	12 children with severe autistic symptoms, mean age: 10.7	CGI-I, ABC	CGI-I ≤2: 42%, however, no significant change in ABC	3 patients (25%) abdominal discomfort and irritable mood
Zeiner et al. (2011)	Open label (10 weeks)	14 boys, mean age: 11.6 (all IQ >70)	ADHD-RS	50%	2 patients (14%)
Harfterkamp et al. (2012)	Randomized, placebo controlled (8 weeks)	ATX group, 48 children, mean age: 9.9, (IQ ≥65)	CGI-I, ADHD-RS, CTRS	ATX>placebo^a; CGI-I ≤2: 21% (ATX group)	1 patient (2%) due to fatigue
Fernandez-Jaen et al. (2013)	Open label (16 weeks)	24 patients, mean age: 8.8 (19 with ID)	CGI-I, ADHD-RS, CPRS	CGI-I ≤2: 50% ADHD-RS ≥25% decrease: 50% (of 24)	3 patients (12.5%)

ABC, Aberrant Behavior Checklist; ADHD, attention-deficit/hyperactivity disorder; ADHD-RS, ADHD Rating Scale; ATX, atomoxetine; CGI-I, Clinical Global Impressions-Improvement scale; CPRS, Conner's Parent rating scale; CTRS, Conner's Teacher rating scale; H-ABC, hyperactivity subscore of ABC; ID, intellectual disability; IQ, intelligence quotient; SNAP-IV, DSM-IV-based Swanson, Nolan, and Pelham Scale; VABS, Vineland Adaptive Behavior Scale.

Improvement in ADHD symptoms was greater with ATX than placebo.

Studies using ATX mostly included high-functioning subjects and less is known about patients with coexisting intellectual disability (ID). In a retrospective study, Jou et al. reviewed 20 cases with ADHD+ASD, half of whom had comorbid ID. Response rate was 60% for both ID and non-ID groups, defined as very much or much improved according to the improvement score of Clinical Global Impressions (CGI-I). In a small randomized placebo controlled study, Arnold et al. evaluated 16 patients with ATX and placebo, 6 weeks each. The patients were required to be above the developmental level of 1.5 years; however, number or response rates of the children with ID were not reported. Fernandez-Jaen et al. investigated 24 children with ASD+ADHD and most of the cases had associated ID. Half of the participants demonstrated clinical improvement (CGI-I ≤2) with the drug and the authors reported that no statistically significant difference was detected between the intellectually disabled (n = 19) and nondisabled (n = 5) patients. A retrospective study examined the association between level of cognitive functioning and clinical response during treatment with ATX (Mazzone et al. 2011) in children with ADHD without ASD. The authors demonstrated that children and adolescents with intelligence quotient (IQ) scores below 85 were nearly four times less likely to be responders than those with an IQ scores more than 85 (20.71% vs. 76.9%).

The first aim of this study was to investigate the effectiveness and tolerability of ATX on the ADHD symptoms and autistic features in a larger group of low-functioning children with ASD+ADHD. The second aim was to compare the clinically responding and nonresponding patients in terms of sociodemographic features, ASD type, intellectual and language level, baseline Clinical Global Impressions-Severity (CGI-S) score, level of autistic and ADHD symptoms, and presence of epilepsy.

Methods

Participants

In this naturalistic and retrospective study, medical records of all children (6–18 years of age) with the diagnosis of ASD and ADHD, who had been on ATX treatment, were reviewed. All of the participants met Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) (American Psychiatric Association 1994) criteria for autistic disorder or pervasive developmental disorder-not otherwise specified (PDD-NOS) and ADHD, and received ATX to improve the core symptoms of ADHD (i.e., inattention, hyperactivity, and impulsivity) for at least 1 month. Participants taking other psychotropic medications were included only if their medications were unchanged 2 weeks before or during ATX use.

The patients in this study were being followed up in the Autism Spectrum Disorders Clinic of our department. The diagnoses were made by the following child psychiatry residents and based on interviews with the parents and child, review of school reports, psychological testing and laboratory assessments, and confirmed by the last author who has expertise in working with patients with ASDs. The forms and checklists are routinely used and ratings are done by the clinician at the times of assessments.

Forty-three children were identified and a total of 37 cases who had associated ID (IQ <70) were included in the study. The group consisted of 28 males (75.8%) and 9 females (24.2%), with a mean chronological age of 10.16 ± 3.60 years. Twenty-three (62%) of them presented with autistic disorder and 14 (38%) had PDD-NOS. Besides clinical interview, the diagnosis of ASDs was supported by the Childhood Autism Rating Scale (CARS) with a score of ≥30 in all cases (Schopler et al. 1980). The diagnosis of ADHD was confirmed by a thorough clinical interview and the ADHD-rating scale (ADHD-RS).

Thirty children (81%) were using psychotropic medications before ATX was started. Risperidone was being used in 25 (67.5%) children. Eighteen of them were using risperidone only and seven in combination (with a SSRI [n = 3], sodium valproat [n = 3], and lamotrigine [n = 1]). Five children were using antiepileptics only (sodium valproate [n = 4], carbamazepine with levetiracetam [n = 1]). Among the 12 patients with epilepsy, 4 had complex partial seizures. Generalized seizures including tonic–clonic (n = 2), absence (n = 1), atypical absence (n = 1), tonic (n = 1), and multiple types (n = 3) were also noted.

ATX was started with a dose of 0.3–0.5 mg/(kg·day), titrated slowly to 1–1.2 mg/(kg·day) in 4 weeks. The daily doses were between 1.0 and 1.4 mg/(kg·day) according to clinical opinion for the side effects and efficacy of the treatment (mean 1.20 ± 0.11 mg/[kg·day]). Patients were seen biweekly for the first month and once per month afterwards. Five (13.5%) children stopped ATX at 4 weeks due to ineffectivity (2) and intolerable side effects (increased motor activity and talkativeness [n = 1]), irritability (n = 2), temper outbursts (n = 2), and increased blood pressure (BP; n = 1). Baseline, 4th- and 12th-week assessments were evaluated in the study. Children with history of ATX use, comorbid psychotic, or bipolar disorder, and IQ ≥70 were excluded. Comorbid psychiatric diagnosis such as anxiety disorders, depressive disorder, oppositional defiant disorder, and conduct disorder were allowed in the study.

The routine assessments included baseline medical and laboratory evaluation of the children for significant medical problems, indicating contraindication to ATX together with the efficacy and side effect measures at the beginning, 4th week, and 12th week of treatment. In the laboratory examinations, complete blood count, liver, kidney, and thyroid function tests, fasting glucose levels, cardiac pulse, systolic and diastolic BP, and body weight measurements were reviewed. The most recent IQ testing was obtained from the files and the children were assigned to the following three levels of intelligence: mild, moderate, and severe ID, based on information from clinical assessments and standardized intelligence tests (Wechsler 1974; Thorndike et al. 1986).

The baseline severity of the symptoms was assessed using the parent-rated ADHD-RS (DuPaul 1998) and Aberrant Behavior Checklist (ABC) (Aman et al. 1985) in addition to clinician-rated CGI-S scale (Guy 1976). Information about the improvement in clinical features and side effects was collected at 4th and 12th weeks using CGI-I, ADHD-RS, ABC, and Barkley Side Effect Rating Scale (BSERS) (Barkley et al. 1990).

Measures

The ADHD-RS consists of 18 items that correspond to the diagnostic symptoms of the disorder, according to Diagnostic and Statistical Manual of Mental Disorders, 4th ed., Text Revision (DSM-IV-TR) (American Psychiatric Association 2000) criteria. Each item is scored from 0 to 3; higher scores on the scale are related to greater symptomatic intensity.

The ABC is used to investigate maladaptive behavior in individuals with developmental disabilities. The original checklist consists of 58 items. The psychometric properties of the Turkish version of the ABC (ABC-T) were investigated on 435 children and adolescents with mental retardation (Sucuoglu 2003). Because 12 items loaded on more than one factor, they were excluded from the ABC-T. The scale with 46 items included the following subscales: hyperactivity, social withdrawal/lethargy, self-injurious behaviors, stereotypic behaviors, and other behaviors. Although factor structure of the Turkish version is different than the original checklist, it is a reliable and valid instrument for assessing behavior problems of people with mental retardation.

The BSERS is a 17-item, parent-rated scale of potential stimulant side effects, where each item is rated from 0 “absent” to 9 “serious.” Following the recommendations of Barkley et al. (1990), ratings of 7 and higher were considered “severe.”

The CARS is a 15-item behavior rating scale that was developed to determine the presence and degree of autistic symptoms as well as to distinguish the cases with ASDs from those with other developmental disorders (Schopler et al. 1980; Sucuoglu et al. 1996). The total score range is between 15 and 60, and scores higher than 30 points indicate ASDs (31–37 mild to moderate and >37 severe symptoms).

The CGI-S is a 7-point scale that requires the clinician to rate the severity of the patient's illness from 1 (normal) to 7 (extremely ill) at the time of assessment. The CGI-I assesses how much the patient's illness has improved or worsened relative to a baseline state at the beginning of the intervention, and rated as “very much improved” (scored 1) to “very much worse” (scored 7). The CGI-I was determined by decrease in the severity of target symptoms (i.e., hyperactivity, inattention, and impulsivity) and careful review of clinicians' impressions.

Data analysis

SPSS v22.0 software package (SPSS) was used for statistical analysis. If a patient stopped medication before 12 weeks, efficacy data were examined by using the last observation carried forward (LOCF) method. Descriptive statistics were presented as number (%) or mean ± standard deviation (SD). The distribution of each variable was evaluated using the Kruskal–Wallis test. Differences between baseline, 4th week, and 12th week were examined with repeated measures ANOVA or Friedman tests, according to the distribution of the variables. Bonferrroni correction was used for multiple comparisons and p values of less than 0.05 (two-tailed) were accepted for statistical significance. Effect sizes were calculated to estimate the magnitude of the significant changes. Effect sizes (i.e., partial eta-squared (η²) and the correlation coefficient (r), which was calculated as the Z value (of the Wilcoxon Signed Rank test) divided by √N (N = total number of subjects)) were reported to estimate the magnitude of the significant changes. Cohen (1988) suggests that η ² = 0.01 or r = 0.1 constitutes a small effect size, η ² = 0.06 or r = 0.3 is a medium effect size, and η ² = 0.14 or r = 0.5 is a large effect size.

The primary outcome measure was a treatment response defined by a CGI-I score of 1 or 2 (“very much improved” or “much improved) together with a decrease of at least 25% in the parent-rated ADHD-RS total score. Only those individuals who could sustain treatment till the end of 12th week were judged as treatment responders. The baseline characteristics considered as predictors of response were sociodemographic parameters (i.e., age, gender, and monthly family income), intellectual and language level, baseline CGI-S, level of autistic and ADHD symptoms, and presence of epilepsy. Monthly family income was considered as a parameter related to socioeconomic level and presence of epilepsy because of its probable relationships with ID and severity of autistic symptoms (El Achkar et al. 2015).

Results

Out of 37 participants, 32 completed the 12-week period on ATX. Treatment completers did not differ from the premature termination group in age, gender, severity of autistic symptoms, and levels of intelligence and speech. The efficacy data of the five children at 4th week were carried forward to the 12th week. A total of 18 subjects (48.6%) were considered “very much improved” (n = 5) and “much improved” (n = 13) with the CGI-I by the end of 12th week. Only five (15.2%) of the subjects were evaluated as “very much” or “much” improved by the end of 4th week.

On the parent-rated ADHD-RS, there were significant reductions between baseline and 4th week, and between baseline and 12th week in total, hyperactivity, and inattention scores. Besides, the difference between baseline and 12th week was also significant in impulsivity (Table 2). Twenty-one (56.8%) of the patients showed ≥25% improvement in ADHD-RS total score. The percentage of patients showing ≥25% improvement in inattention, hyperactivity, and impulsivity scores was 48.6%, 45.9%, and 45.9%, respectively. In the parent-rated ABC, decrease was only significant in the hyperactivity and social withdrawal subscales, and 48.6% and 43.2% of the patients showed ≥25% improvement.

Table 2.

Comparison of the Side Effect and Efficacy Measures at Baseline, 4th Week, and 12th Week

	n	Mean (SD)	p (effect size)	Significant bivariate comparisons
Weight
Baseline	37	39.3 (16.5)	0.29^a
4th week	37	39.9 (16.4)	(η ² = 0.08)
12th week	32	39.8 (16.9)
Systolic BP
Baseline	37	100.0 (12.4)	0.043^b	BL-4th week, p = 0.004^c
4th week	37	103.4 (15.8)		(r = 0.33)
12th week	32	99.5 (19.5)
Diastolic BP
Baseline	37	69.7 (11.9)	0.47^b
4th week	37	68.4 (7.8)
12th week	32	68.4 (7.1)
ADHD-RS-total
Baseline	37	38.8 (9.9)	<0.001^a	BL-4th week, p = 0.001^c
4th week	37	33.0 (9.31)	(η ² = 0.42)	BL-12th week, p < 0.001^c
12th week	37	28.8 (11.8)
ADHD-RS-inattention
Baseline	37	20.9 (4.5)	<0.001^a	BL-4th week, p = 0.001^c
4th week	37	17.5 (4.7)	(η ² = 0.45)	BL-12th week, p < 0.001^c
12th week	37	15.4 (6.0)
ADHD-RS-hyperactivity
Baseline	37	12.9 (5.1)	0.003^a	BL-4th week, p = 0.013^c
4th week	37	11.3 (4.9)	(η ² = 0.29)	BL-12th week, p = 0.002
12th week	37	10.0 (5.2)
ADHD-RS-impulsivity
Baseline	37	4.9 (2.7)	0.007^a	BL-12th week,
4th week	37	4.2 (2.5)	(η ² = 0.25)	p = 0.005^c
12th week	37	3.5 (2.3)
ABC-hyperactivity
Baseline	37	26.7 (10.3)	<0.001^a	BL-4th week, p < 0.001^c
4th week	37	22.6 (9.2)	(η ² = 0.40)	BL-12th week, p = 0.001^c
12th week	37	20.2 (10.4)
ABC-social withdrawal
Baseline	37	25.3 (8.7)	0.005^a	BL-4th week, p = 0.02^c
4th week	37	22.5 (8.3)	(η ² = 0.26)	BL-12th week, p = 0.005^c
12th week	37	19.8 (8.7)
ABC-stereotypy
Baseline	37	8.2 (4.7)	0.31^a
4th week	37	7.4 (4.3)	(η ² = 0.07)
12th week	37	7.1 (5.1)
ABC-self mutilation
Baseline	37	2.2 (2.9)	0.69^a
4th week	37	2.0 (2.8)	(η ² = 0.02)
12th week	37	1.9 (2.6)
ABC-other
Baseline	37	5.5 (3.1)	0.14^a	BL-4th week, p = 0.03
4th week	37	5.1 (2.6)	(η ² = 0.11)	BL-12th week, p = 0.02
12th week	37	4.5 (2.8)

Repeated measures ANOVA.

Friedman test.

Statistically significant bivariate comparisons (after Bonferroni corrections).

η ², partial eta-squared; ABC, Aberrant Behavior Checklist; BP, blood pressure; n, number of cases; r, correlation coefficient; SD, standard deviation.

According to CARS, 19 of the patients had severe autistic symptoms. When the analyses were repeated for this subgroup, statistically significant gains were observed in all parameters of ADHD-RS, and ABC hyperactivity and social withdrawal sores.

Sixteen (43.2%) of the patients fulfilled the response criteria of our study (i.e., CGI-I score of 1 or 2 and ≥25% decrease in the ADHD-RS total score). When responders and nonresponders were compared according to selected variables, no statistically significant difference was detected for sociodemographic parameters (gender, age, and monthly family income), intellectual and language level, presence or absence of epilepsy, baseline CGI-S, and levels of autistic and ADHD symptoms (Table 3).

Table 3.

Comparison of the Responder and Nonresponder Groups for the Baseline Characteristics

	Responders (n = 16)	Nonresponders (n = 21)	p
Age year, mean (SD)	10.6 (3.9)	9.9 (3.4)	0.63
Male sex (%)	13 (81.3)	15 (71.4)	0.70
ASD diagnosis, n (%)
Autistic disorder	8 (50.0)	15 (71.4)	0.32
PDD-NOS	8 (50.0)	6 (28.6)
Intellectual level (%)
Mild ID	4 (25.0)	6 (28.6)	0.49
Moderate ID	9 (56.3)	8 (38.1)
Severe ID	3 (18.8)	7 (33.3)
Level of speech (%)
No phrasal speech	5 (31.3)	10 (47.6)	0.50
Phrasal or fluent speech	11 (68.8)	11 (52.4)
Epilepsy positive (%)	6 (37.5)	6 (28.6)	0.72
CARS score, mean (SD)	36.4 (5.6)	39.5 (7.7)	0.19
Baseline CGI-S, mean (SD)	4.3 (0.9)	4.7 (0.9)	0.20
Baseline ADHD-RS-IV total, mean (SD)	39.5 (10.1)	38.2 (9.9)	0.62

ADHD, attention-deficit/hyperactivity disorder; ADHD-RS-IV, DSM-IV-based ADHD-rating scale; ASDs, autism spectrum disorders; CARS, Childhood Autism Rating Scale; CGI-S, Clinical Global Impressions-Severity scale; ID, intellectual disability; PDD-NOS, pervasive developmental disorder-not otherwise specified; SD, standard deviation.

With slow and careful dose titration and close monitoring, all the patients were taking the drug at the 4th week evaluation. There was no significant change in weight and diastolic BP during the study; however, there appeared a significant increase in systolic BP between baseline and 4th week (Table 2). Forty-four percent showed no change; 34% 5–10 mm Hg increase; 13% 20–40 mm Hg increase; and 9% 5–10 mm Hg decrease. A 13-year-old boy with a systolic BP of 140 mm Hg was decided to stop taking the drug at the 4th week. He also had palpitations and increased anxiety. His baseline systolic BP was 130 and no concomitant medications were used.

Comparison for total BSERS scores at baseline, 4th week, and 12th week revealed no significant difference. Mean and SD scores of each item can be seen on Table 4. The patients had higher levels of dizziness at 4th week than at baseline (p = 0.02); however, the difference was no longer significant at 12th week. Interestingly, BSERS revealed significant decrease in being uninterested in others between baseline and 12th week (p = 0.008). In addition, there was near significant increase in stomach ache (p = 0.08; increased at both 4th and 12th weeks) and drowsiness (p = 0.09; increased at 4th week).

Table 4.

Mean and Standard Deviation Scores of the Barkley Side Effect Rating Scale Items at Baseline, 4th Week, and 12th Week

	Baseline ^a	4th Week ^a	12th Week ^a	p
Trouble sleeping	2.11 (2.48)	2.41 (2.65)	2.42 (2.70)	0.29
Poor appetite	1.78 (2.40)	2.51 (2.59)	1.69 (1.99)	0.19
Irritable	3.22 (2.92)	3.16 (2.79)	3.11 (2.86)	0.45
Proneness to crying	1.65 (2.5)	2.22 (3.2)	1.94 (3.12)	0.43
Anxiousness	1.62 (2.34)	1.70 (2.03)	1.11 (2.12)	0.34
Sadness/unhappiness	1.30 (2.23)	1.81 (2.8)	1.58 (2.63)	0.42
Headaches	0.68 (1.49)	0.62 (1.77)	0.61 (1.73)	0.13
Stomachaches	0.57 (1.07)	1.05 (1.86)	1.19 (2.14)	0.08^b
Nightmares	0.54 (1.24)	0.78 (1.75)	0.47 (1.36)	0.26
Daydreams	1.14 (2.36)	1.32 (2.0)	0.72 (1.5)	0.09^b
Talking little with others	2.24 (3.27)	1.86 (2.86)	1.56 (2.69)	0.40
Uninterested in others	2.89 (3.36)	2.22 (2.67)	1.61 (2.59)	0.006^{b c}
Drowsiness	0.89 (2.08)	1.43 (2.39)	0.92 (1.7)	0.09^b
Biting fingernails	1.38 (2.99)	1.30 (2.72)	0.89 (2.16)	0.58
Unusually happy	1.70 (2.63)	1.65 (2.62)	1.06 (1.96)	0.16
Dizziness	0.16 (0.5)	0.70 (1.7)	0.47 (1.06)	0.025^{b d}
Tics or nervous movements	2.03 (3.35)	2.22 (3.07)	1.86 (2.76)	0.81

Range 0–9 points.

p ≤ 0.10 with Friedman test.

Statistically significant difference between baseline and 12th week.

Statistically significant difference between baseline and 4th week.

Discussion

In this study, efficacy and tolerance of ATX were studied in a group of low-functioning children and adolescents with ASDs and ADHD. In a naturalistic treatment setting and in 12-week period, overall treatment success based on both clinician's opinion and standard parental rating scales was 43.2%. Clinician-only observation of improvement revealed a rate of 48.8%. The patients obtained significant benefits for inattention, hyperactivity, impulsivity, and social withdrawal from baseline to 12th week. Concerning the discontinuation rates, none of the patients stopped before the 4th week visit, and five patients stopped (13.5%) the medication before the 12-week period.

Although some studies reported much higher efficacy rates for ATX in the ASD group, clinician-based improvement of 48.8% in our study is comparable to the majority of the former studies (Table 1). A recent review reported that median response rate of ATX is 50% in children with developmental disabilities, most of whom had ASDs (Aman et al. 2014). Posey et al. who found a response rate of 75% included only children with nonverbal IQ scores above 70 and stated that the rate might be lower for children with mental retardation. In a group of children predominantly with mental retardation (19 out of 24), clinical improvement was found as 50%, which was very similar to ours (Fernandez-Jaen et al. 2013). Also mean decreases in the scores of parent-rated ADHD-RS for inattention (5.6 vs. 5.5) and hyperactivity/impulsivity (4.0 vs. 4.3) were very similar to those in Fernandez-Jaen's and our studies.

The level of ID (i.e., mild, moderate, or severe) did not have significant effect on clinical response rates in our study. Mazzone et al. reported that higher IQ is associated with higher response rate to ATX in a sample of children with ADHD without ASD. However, a recent article reported that higher IQ did not predict clinical response in children and adolescents with ASDs and concomitant ADHD symptoms (Harfterkamp et al. 2015). The latter study included children with IQ scores more than 60; however the mean IQ score was 92.6 with an SD of 17.1. Our study supported a similar finding in children and adolescents with much more lower IQ scores. Besides, both ours and Harfterkamp's study found that type of ASD (autistic disorder versus other DSM-IV-TR pervasive developmental disorders, such as PDD-NOS) did not predict clinical response.

Severity of the autistic symptoms was proposed to affect the response rate (Zeiner et al. 2011). No significant difference between the responder and nonresponder groups was observed for the CARS scores in our study. However, the only study with negative results with ATX was performed on children with severe autistic symptoms (Charnsil 2011). In that study, no statistical significant improvement in the ADHD parameters was detected in 12 children with CARS scores above 37. Our study included 19 (51.4%) participants with severe autistic symptoms, and when the analyses were repeated for this subgroup, significant gains were observed in all parameters of ADHD-RS, and ABC hyperactivity and social withdrawal sores. Charnsil reported that besides the severity of autistic symptoms, inability to increase the dose of ATX to optimal doses (mean dose was 0.98 mg/[kg·day]) could explain the lack of benefit in their group. Our study suggests that ATX is effective in children with severe autistic symptoms when adequate doses are used.

Previous studies report contrasting results for the effect of ATX on social withdrawal in subjects with ASDs. Arnold et al. (2006) and Posey et al. (2006) showed statistically significant improvement with ATX when compared with placebo and baseline scores, respectively. However, some studies detected no change (Troost et al. 2006; Charnsil 2011; Harfterkamp et al. 2014). The parents in our study reported significant decrease in “being uninterested in others” with the BSERS, together with a decrease in the ABC withdrawal scores between baseline and 12th week. As a recent review demonstrated that stimulant treatment is associated with increased rates of social withdrawal in children with ASDs (Reichow et al. 2013), ATX may be a better alternative for these children.

Although placebo-controlled studies revealed discontinuation rates as low as 2% or 6% (Arnold et al. 2006; Harfterkamp et al. 2012), the rate found in our study is compatible with majority of studies, which found rates between 10% and 20% in children with ASDs. However, these rates are much higher than the rates (around 3%) reported for the typically developing children with ADHD (Kratochvil et al. 2008). In the present study, the most commonly reported side effects causing discontinuation of ATX were irritability and temper outbursts. However, after 2 months of cessation of ATX, these symptoms decreased in two patients, did not change in one patient, and increased in one patient. So it is important to bear in mind that these symptoms may not always be related to medication side effects.

One patient stopped medication at 4 weeks due to increased systolic BP, palpitations, and increased anxiety. The increase in BP returned toward baseline and the palpitations disappeared upon drug discontinuation. In addition, there was significant increase in systolic BP between baseline and 4th week. Previous studies reveal that ATX produces only a minimal, if any, increase in BP (Schwartz and Correll 2014). It is hypothesized that blockade of norepinephrine reuptake with ATX would increase neurotransmitter concentrations in the neuroeffector junction, which would cause a pressor response in the periphery. However, this mechanism seems to be counteracted by a central sympatholytic action through activation of alpha-2 adrenoreceptors (Shibao et al. 2007). None of the studies on children with ASDs and ADHD reported any patients who discontinued ATX due to increase in BP, nor any significant increase between baseline and after treatment (Table 1). In light of the present observations, we underline the necessity of measuring BP at baseline and periodically while on therapy to identify children and adolescents who are at heightened risk for increased BP.

The present study has several limitations to note. The most important of these are the retrospective, uncontrolled, and unblinded nature of the data and concomitant medication use during ATX treatment. The follow-up period of 12 weeks might have been short to detect the exact response rate because a recent study suggested that continued treatment with ATX up to 28 weeks further improved ADHD symptoms in children and adolescents with ASDs (Harfterkamp et al. 2013). The sample size in our study was adequate to detect ATX efficiency in improving ADHD symptoms in children with ASDs and ID, but it might be low to investigate the possible predictors of response. Nonetheless, a recent, larger study also found that age, gender, IQ level, and ASD type did not predict clinical response and confirmed our findings in higher functioning children with ASDs (Harfterkamp et al. 2015).

In the LOCF analysis, we assumed that for patients who stopped ATX at 4th week (due to side effect, etc.), no change would have been observed if they continued until 12th week. This analysis has advantages like minimizing the number of subjects eliminated from the analysis and examining trends over time, rather than focusing simply on the endpoint (Streiner and Geddes 2001). However, if patients' data would have continued to improve or worsen after they stopped medication, then LOCF might have underestimated the average improvement or worsening (Prakash et al. 2008).

We used the Turkish (amended) version of ABC because this is the only standardized form in our language. We gave it as a complementary tool to the DSM-IV-based ADHD-RS and evaluated a probable change in withdrawal. The ABC scores in our study are not readily comparable to scores in other studies that used the original English version. However, the readers can infer the level of ADHD symptoms (e.g., hyperactivity) from the ADHD-RS scores and follow the symptom change with time. Another limitation is the lack of a diagnostic assessment with tools accepted as gold standard for ASDs such as the Autism Diagnostic Interview-Revised (Lord et al. 1994) and the Autism Diagnostic Observation Schedule (Lord et al. 2000). Instead, we used CARS in the absence of more validated instruments for the Turkish population. Finally, reliance on prerecorded IQ scores with different tests may be seen as a limitation. However, all the tests were conducted by qualified psychologists experienced with patients with ASDs. In addition, the presence and level of ID were also confirmed clinically with other information in the reports just as school and educational reports from school or special education teachers and reviewing of the adaptive skills.

Besides limitations, relatively large number of children with comorbid ASD and ID, assessment by both the clinician and parents, and evaluation of probable side effects at baseline and during treatment rather than spontaneous report of the parents after medication were the strengths of our study.

Conclusions

The results of this study show that ATX may be effective in all of the key symptoms of ADHD, namely inattention, hyperactivity, and impulsivity in children and adolescents with ADHD and low-functioning ASDs. Parent- and clinician-reported response rates are nearly similar to those of previous studies with similar or higher IQ levels. ATX may also be beneficial for social withdrawal. Although it appears safe, BP, especially systolic BP, increase should be followed carefully and may pose problems for some children. Responders were not significantly different from nonresponders in terms of sociodemographic features, intellectual and language level, baseline CGI-S, level of autistic and ADHD symptoms, and presence of epilepsy. Large, prospective, placebo-controlled studies are needed to support these findings.

Clinical Significance

Low-functioning patients with ASDs are frequently excluded from pharmacological treatment studies for ADHD symptoms. This study is important for including a relatively high number of subjects with ASDs and ID. The results showed that ATX may be a safe and effective option for ADHD symptoms in these children. The level of ID did not affect the response rate. So very low-functioning children, with moderate or severe ID, also benefited from ATX as much as those with mild ID. However, clinicians should be careful about potential irritability and elevated BP when using the drug.

Footnotes

Disclosures

Nahit Motavalli Mukaddes is at the advisory board of the Jansen Cilag, Lilly, Bristol-Mayers. Mustafa Deniz Tutkunkardas is working as a Senior Medical Lead in GlaxoSmithKline. Ayse Kilincaslan, Tuba Duzman Mutluer, and Basak Pasabeyoglu have nothing to disclose. The authors received no financial support for the research, authorship, and/or publication of this article.

References

Aman

, Farmer

, Hollway

, Arnold

: Treatment of inattention, overactivity, and impulsiveness in autism spectrum disorders. Child Adolesc Psychiatr Clin N Am, 17:713–738, 2008.

Aman

, Singh

, Stewart

, Field

: The aberrant behavior checklist: A behavior rating scale for the assessment of treatment effects. Am J Ment Defic, 89:485–491, 1985.

Aman

, Smith

, Arnold

, Corbett-Dick

, Tumuluru

, Hollway

, Hyman

, Mendoza-Burcham

, Pan

, Mruzek

, Lecavalier

, Levato

, Silverman

, Handen

: A review of atomoxetine effects in young people with developmental disabilities. Res Dev Disabil, 35:1412–1424, 2014.

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed., Text Revision. Washington, DC: American Psychiatric Association; 2000.

Arnold

, Aman

, Cook

, Witwer

, Hall

, Thompson

, Ramadan

: Atomoxetine for hyperactivity in autism spectrum disorders: Placebo-controlled crossover pilot trial. J Am Acad Child Adolesc Psychiatry, 45:1196–1205, 2006.

Barkley

, McMurray

, Edelbrock

, Robbins

: Side effects of methylphenidate in children with attention deficit hyperactivity disorder: A systemic, placebo-controlled evaluation. Pediatrics, 86:184–192, 1990.

Charnsil

: Efficacy of atomoxetine in children with severe autistic disorders and symptoms of ADHD: An open-label study. J Atten Disord, 15:684–689, 2011.

Cohen

: Statistical Power Analysis for the Behavioral Sciences, 2 ed. Hillsdale (New Jersey), Lawrence Erlbaum Associates, 1988.

10.

Dalsgaard

, Nielsen

, Simonsen

: Five-fold increase in national prevalence rates of attention-deficit/hyperactivity disorder medications for children and adolescents with autism spectrum disorder, attention-deficit/hyperactivity disorder, and other psychiatric disorders: A Danish register-based study. J Child Adolesc Psychopharmacol, 23:432–439, 2013.

11.

DuPaul

, Power

, Anastopoulos

, Reid

: ADHD Rating Scale-IV: Checklists, Norms, and Clinical Interpretations. New York, Guilford Press, 1998.

12.

El Achkar

, Spence

: Clinical characteristics of children and young adults with co-occurring autism spectrum disorder and epilepsy. Epilepsy Behav, 47:183–190, 2015.

13.

Fernandez-Jaen

, Fernandez-Mayoralas

, Calleja-Perez

, Munoz-Jareno

, Campos Diaz Mdel

, Lopez-Arribas

: Efficacy of atomoxetine for the treatment of ADHD symptoms in patients with pervasive developmental disorders: A prospective, open-label study. J Atten Disord, 17:497–505, 2013.

14.

Gadow

, DeVincent

, Pomeroy

: ADHD symptom subtypes in children with pervasive developmental disorder. J Autism Dev Disord, 36:271–283, 2006.

15.

Guy

: ECDEU Assessment Manual for Psychopharmacology, Revised. Rockville (Maryland), United States Department of Health, Education and Welfare, 1976.

16.

Harfterkamp

, Buitelaar

, Minderaa

, van de Loo-Neus

, van der Gaag

, Hoekstra

: Long-term treatment with atomoxetine for attention-deficit/hyperactivity disorder symptoms in children and adolescents with autism spectrum disorder: An open-label extension study. J Child Adolesc Psychopharmacol, 23:194–199, 2013.

17.

Harfterkamp

, Buitelaar

, Minderaa

, van de Loo-Neus

, van der Gaag

, Hoekstra

: Atomoxetine in autism spectrum disorder: No effects on social functioning; some beneficial effects on stereotyped behaviors, inappropriate speech, and fear of change. J Child Adolesc, Psychopharmacol24:481–485, 2014.

18.

Harfterkamp

, van de Loo-Neus

, Minderaa

, van der Gaag

, Escobar

, Schacht

, Pamulapati

, Buitelaar

, Hoekstra

: A randomized double-blind study of atomoxetine versus placebo for attention-deficit/hyperactivity disorder symptoms in children with autism spectrum disorder. J Am Acad Child Adolesc Psychiatry, 51:733–741, 2012.

19.

Harfterkamp

, van der Meer

, van der Loo-Neus

, Buitelaar

, Minderaa

, Hoekstra

: No evidence for predictors of response to atomoxetine treatment of attention-deficit/hyperactivity disorder symptoms in children and adolescents with autism spectrum disorder. J Child Adolesc Psychopharmacol, 25:372–375, 2015.

20.

Holtmann

, Bolte

, Poustka

: Attention deficit hyperactivity disorder symptoms in pervasive developmental disorders: Association with autistic behavior domains and coexisting psychopathology. Psychopathology, 40:172–177, 2007.

21.

Jou

, Handen

, Hardan

: Retrospective assessment of atomoxetine in children and adolescents with pervasive developmental disorders. J Child Adolesc Psychopharmacol, 15:325–330, 2005.

22.

Kratochvil

, Milton

, Vaughan

, Greenhill

: Acute atomoxetine treatment of younger and older children with ADHD: A meta-analysis of tolerability and efficacy. Child Adolesc Psychiatry Ment Health, 2:25, 2008.

23.

Lord

, Risi

, Lambrecht

, Cook

Jr. , Leventhal

, DiLavore

, Pickles

, Rutter

: The autism diagnostic observation schedule-generic: A standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord, 30:205–223, 2000.

24.

Lord

, Rutter

, Couteur

: Autism diagnostic interview-revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord, 24:659–685, 1994.

25.

Mazzone

, Reale

, Mannino

, Cocuzza

, Vitiello

: Lower IQ is associated with decreased clinical response to atomoxetine in children and adolescents with attention-deficit hyperactivity disorder. CNS Drugs, 25:503–509, 2011.

26.

Murray

: Attention-deficit/hyperactivity disorder in the context of autism spectrum disorders. Curr Psychiatry Rep, 123:382–388, 2010.

27.

Oswald

, Sonenklar

: Medication use among children with autism spectrum disorders. J Child Adolesc Psychopharmacol, 17:348–355, 2007.

28.

Posey

, Wiegand

, Wilkerson

, Maynard

, Stigler

, McDougle

: Open-label atomoxetine for attention-deficit/hyperactivity disorder symptoms associated with high-functioning pervasive developmental disorders. J Child Adolesc Psychopharmacol, 16:599–610, 2006.

29.

Prakash

, Risser

, Mallinckrodt

: The impact of analytic method on interpretation of outcomes in longitudinal clinical trials. Int J Clin Pract, 62:1147–1158, 2008.

30.

Reichow

, Volkmar

, Bloch

: Systematic review and meta-analysis of pharmacological treatment of the symptoms of attention-deficit/hyperactivity disorder in children with pervasive developmental disorders. J Autism Dev Disord, 43:2435–2441, 2013.

31.

Schopler

, Reichler

, DeVellis

, Daly

: Toward objective classification of childhood autism: Childhood Autism Rating Scale (CARS). J Autism Dev Disord, 10:91–103, 1980.

32.

Schwartz

, Correll

: Efficacy and safety of atomoxetine in children and adolescents with attention-deficit/hyperactivity disorder: Results from a comprehensive meta-analysis and metaregression. J Am Acad Child Adolesc Psychiatry, 53:174–187, 2014.

33.

Shibao

, Raj

, Gamboa

, Diedrich

, Choi

, Black

, Robertson

, Biaggioni

: Norepinephrine transporter blockade with atomoxetine induces hypertension in patients with impaired autonomic function. Hypertension, 50:47–53, 2007.

34.

Streiner

, Geddes

: Intention to treat analysis in clinical trials when there are missing data. Evid Based Ment Health, 4:70–71, 2001.

35.

Sucuoglu

: The psychometric characteristics of the Turkish form of the aberrant behavior checklist. Turkish J Psychol, 18:93–96, 2003.

36.

Sucuoglu

, Oktem

, Akkok

, Gokler

: A study of the scales for the assessment of the children with autism. J Psychiatry Psychopharmacol Psychol, 4:116–121, 1996.

37.

Thorndike

, Hagen

, Sattler

: The Stanford–Binet Intelligence Scale, 4th ed. Chicago, Riverside Publishing Company, 1986.

38.

Troost

, Steenhuis

, Tuynman-Qua

, Kalverdijk

, Buitelaar

, Minderaa

, Hoekstra

: Atomoxetine for attention-deficit/hyperactivity disorder symptoms in children with pervasive developmental disorders: A pilot study. J Child Adolesc Psychopharmacol, 16:611–619, 2006.

39.

Wechsler

: Wechsler Intelligence Scale for Children–Revised. New York, Psychological Corporation, 1974.

40.

Yerys

, Wallace

, Sokoloff

, Shook

, James

, Kenworthy

: Attention deficit/hyperactivity disorder symptoms moderate cognition and behavior in children with autism spectrum disorders. Autism Res, 2:322–333, 2009.

41.

Zeiner

, Gjevik

, Weidle

: Response to atomoxetine in boys with high-functioning autism spectrum disorders and attention deficit/hyperactivity disorder. Acta Paediatr, 100:1258–1261, 2011.