Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 May 29.
Published in final edited form as: J Psychiatr Res. 2018 Mar 27;102:57–64. doi: 10.1016/j.jpsychires.2018.03.009

Response inhibition and emotional cognition improved by atomoxetine in children and adolescents with ADHD: The ACTION randomized controlled trial

Kristi R Griffiths a,1, John E Leikauf b,1, Tracey W Tsang a,e, Simon Clarke a,c,d, Daniel F Hermens f, Daryl Efron g, Leanne M Williams b,*,2, Michael R Kohn a,c,e,2
PMCID: PMC9148271  NIHMSID: NIHMS1708527  PMID: 29674270

Abstract

Although the non-stimulant medication atomoxetine is effective for attention-deficit hyperactivity disorder (ADHD) in children and adolescents, there are still significant gaps in our knowledge about whether atomoxetine improves anxiety symptoms or cognition in children. Furthermore, while cognition has been proposed as an intermediate phenotype for ADHD dysfunction, the relationships between clinical and cognitive outcomes are not yet understood. We addressed these knowledge gaps in a controlled trial using objective assessments of both general and emotional cognitive functions implicated in ADHD and in anxiety, which commonly co-occurs with ADHD. A total of 136 children and adolescents with ADHD (ages 6–17years; 80% male; 31.6% with a comorbid anxiety disorder) were enrolled in a randomized double-blind, placebo-controlled, cross-over trial of 6-weeks treatment with atomoxetine. Of these, 109 completed the second cross-over phase. Selected cognitive domains associated with ADHD and anxiety disorders (Sustained attention, response inhibition and fearful face identification) were assessed using a normed, computerized test battery. Symptom outcomes were assessed by parent reports on the ADHD Rating Scale-IV and Conners’ Anxious-Shy subscale. For completers, atomoxetine caused a greater improvement in the primary cognitive outcomes of response inhibition and fear identification compared to placebo, but not in sustained attention. Atomoxetine also improved ADHD and anxiety symptoms. Anxiety symptoms improved most for ADHD and anxiety disorder combined, but presence of an anxiety disorder did not moderate any other outcomes. Changes in cognitive and clinical outcomes were not correlated. These findings contribute to the foundations of measurement-based treatment planning and offer targets for probing the mechanisms of atomoxetine action.

1. Introduction

Atomoxetine reduces clinical symptoms of Attention-Deficit/ Hyperactivity Disorder (ADHD) (Childress, 2016; Clemow and Bushe, 2015; Schwartz and Correll, 2014). ADHD is frequently associated with deficits in cognitive functions that are theorized to relate to core symptoms of the disorder as intermediate phenotypes (Nigg, 2005; Willcutt et al., 2005) and that are associated with additional functional impairment (Biederman et al., 2006). There is conflicting evidence as to whether atomoxetine is efficacious for improving these specific cognitive functions. Furthermore, the relationships between changes in cognition and in clinical symptoms are currently undescribed. Knowledge of these relationships will inform our understanding of the pathophysiology of ADHD symptoms and may also inform personalized treatment approaches.

Deficits in behavioral inhibition and attention are core diagnostic criteria for ADHD (Amercian Psychological Association, 2013). Clinical deficits in emotional functioning are not currently part of DSM criteria but have long been recognized as often being present in persons with ADHD (Shaw et al., 2014). Measured objectively with performance-based tests, response inhibition and sustained attention deficits are among the most common cognitive deficits associated with ADHD (Gau and Shang, 2010a; b; Nigg, 2005; Williams et al., 2010c). Previously, using standardized behavioral tests, we have also shown that children and adolescents with ADHD have problems identifying emotional expressions, particularly those depicting expressions of threat, such as fear (Williams et al., 2008).

The question of whether atomoxetine improves response inhibition, sustained attention, or emotional cognition for children has not yet been examined using randomized, controlled designs. In adults with ADHD, atomoxetine has been found to improve response inhibition and sustained attention (Chamberlain et al., 2007; Faraone et al., 2005; Ni et al., 2013; Spencer et al., 1998a). In children, non-placebo-controlled designs have shown both positive (Barry et al., 2009; Gau and Shang, 2010a; b; Shang and Gau, 2012) and null (Bedard et al., 2015; Cubillo et al., 2014; Spencer et al., 2001) results. No previous study has examined atomoxetine’s effects on facial expression identification, despite evidence for impairments in this aspect of cognition in ADHD (Williams et al., 2008).

Thus, our first aim was to investigate the effect of atomoxetine on objective measures of response inhibition, sustained attention and fearful face identification within a large sample of young people with ADHD using a stringent, double-blind, placebo-controlled design.

While these constructs have been and are defined in different ways throughout the literature, we simplify here for ease of communication by referring to the measures collectively as ‘cognition’ or ‘cognitive.’ Within the NIH Research Domain Criteria (RDoC) framework, our measures of cognition are operationalized at the behaviour level (computerized testing) and index constructs within the cognitive systems domain (response inhibition in the construct of cognitive control and sustained attention in the construct of attention) and the negative valence system domain (facial emotion identification).

Additionally, more than 40% of children and adolescents with ADHD also meet criteria for an anxiety disorder (Steinhausen et al., 2006; Tsang et al., 2015). This comorbidity exacerbates functional disability (Jensen et al., 2001a; b; Steinhausen et al., 2006; Tsang et al., 2015). Anxiety also has opposite effects to ADHD on inhibition and reaction times measured with cognitive tests, and may add to impairments in emotion recognition, especially fear (Bloemsma et al., 2013; Manassis et al., 2000; Schatz and Rostain, 2006; Vloet et al., 2010; Williams et al., 2008). There is growing evidence that atomoxetine is effective for clinical symptoms of anxiety in youth who also have ADHD (Adler et al., 2009; Gabriel and Violato, 2011; Garnock-Jones and Keating, 2011; Geller et al., 2007; Kratochvil et al., 2005), and anxiety is an indication for considering atomoxetine as a first-line treatment option (Pliszka et al., 2006) because anxiety moderates response to stimulant and behavioral treatments (Hinshaw, 2007; Jensen et al., 2001a; b; Williams et al., 2008). However, it is still unknown whether anxiety moderates the cognitive or clinical effects of atomoxetine in youth with ADHD. The second aim of our study was therefore to directly compare the effect of anxiety disorder comorbidity on atomoxetine efficacy for both cognitive and symptom outcomes.

While deficits in cognitive functions are not unique to ADHD, such objective, reproducible measures are of interest because they can probe neural circuits (Castellanos et al., 2006; Insel et al., 2010; Karalunas et al., 2014; Williams et al., 2010c). These measures are commonly used in neuroimaging and other studies to draw inferences about underlying circuit dysfunctions in disorders such as ADHD. However, interpretation is still in many ways limited by our lack of knowledge regarding the relationships between the dynamics of cognitive measures and symptoms measured by clinical rating scales—particularly in response to pharmacologic treatment (Mueller et al., 2017; Williams et al., 2010b). These cognitive constructs could define subtypes with different response profiles, mediate the effects of medications on symptoms, or represent important treatment targets in their own right. Our third exploratory aim was to determine the relationships between any changes in cognition and changes in clinical symptoms.

We addressed our three aims in the study called “A Randomized Controlled Trial Investigation Of a Non-stimulant in attention-deficit/hyperactivity disorder (ACTION)” (Tsang et al., 2011). ACTION followed a placebo-controlled, double-blinded, crossover design in a large sample of children and adolescents with ADHD, of whom a representative subset also had an anxiety disorder. We hypothesized that atomoxetine (compared with placebo) would be associated with greater improvement in our primary outcomes of response inhibition, sustained attention and fearful face identification in addition to clinical symptoms of both ADHD and anxiety. We further hypothesized that presence of an anxiety disorder would be associated with less improvement in response inhibition and sustained attention, greater improvement in fear recognition, and greater improvement in anxiety but not ADHD symptoms. Our overall objective was to understand the effect of atomoxetine on objective cognitive outcomes in order to identify targets for future personalized treatment and to advance understanding of the underlying mechanisms of treatment action.

2. Material and methods

2.1. Trial design

ACTION was a double-blind, randomized, placebo-controlled crossover trial undertaken at three academic sites in Australia; Sydney, Melbourne, and Adelaide (Tsang et al., 2011). Enrolment commenced in February, 2008 and was completed in April, 2010.

All participants were assessed at Baseline (Week 0), post-treatment phase 1 (Week 6), and post-treatment phase 2 (Week 13). There was a one-week washout period during Week 7, prior to commencement of the second phase. We used a pair-wise randomization technique, ensuring an even allocation of participants to receiving atomoxetine or placebo first while also taking into account the expected ratios across five weight groups. The randomization master list was generated by a hospital biostatistician who forwarded this list to the compounding chemist responsible for packaging and labeling the blister cards for the study. This protocol ensured that investigators, research personnel and participants were blind to treatment phase allocation.

2.2. Participants

Children aged 6–17 years with ADHD were referred to the study by clinicians at each of the three sites. Diagnoses of ADHD were established by clinical consensus (MRK, SC, DE) according to DSM-IV (American Psychiatric Association, 1994) criteria. Inclusion criteria were primary diagnosis of ADHD and English language fluency. Participants were excluded if they had concurrent stimulant use, contraindications to atomoxetine (e.g. cardiac abnormalities or severely underweight), or alcohol and substance use. Research psychologists confirmed diagnosis and ADHD subtype (predominantly inattentive, predominantly hyperactive-impulsive, or combined) using the parentrated ADHD Rating Scale IV (ADHD-RS IV (DuPaul, 1991)), indicated by a score > 2 in six or more items in the Inattentive and/or Hyperactive-Impulsive sections of the scale.

The presence or absence of a comorbid anxiety disorder was assessed using the Anxiety Disorders Interview Schedule for Children (ADISC)(Silverman et al., 2001). We applied DSM-IV criteria to assess for separation anxiety disorder, generalized anxiety disorder and/or social anxiety disorder (Geller et al., 2007). We also used the ADISC to assess for other comorbid internalizing disorders, including major depressive disorder, specific phobias, obsessive-compulsive disorder and post-traumatic stress disorder, and for other comorbid externalizing disorders, including oppositional defiant disorder and conduct disorder.

Institutional Review Board approval was obtained at each site prior to enrolment of participants and all study procedures accorded with the Helsinki Declaration. Investigators obtained written informed assent from each participant and consent from their parent/guardian before undertaking any study-related procedures, after providing written and verbal explanations about the study aims, methods, and potential benefits and risks.

2.3. Study treatment

After baseline assessments, ADHD participants were randomized to receive either atomoxetine or placebo first. Atomoxetine dose was determined according to participant body mass, as per prescribing guidelines (mean dose was 1.35 mg.kg−1; range 1.0–1.4 mg.kg−1) (Tsang et al., 2011). Atomoxetine and placebo preparations were purchased from Eli Lilly. Both atomoxetine and placebo doses were packaged externally into blister packs to maintain double blinding. The atomoxetine and placebo pill and packaging were identical in appearance. Compliance to treatment was monitored by a daily record booklet completed by parents/guardian and the return of empty blister packs.

2.4. Assessment of response inhibition, sustained attention and fearful face emotion identification

Response inhibition, sustained attention and fearful face identification were assessed at baseline and at the end of each 6-week treatment phase (active atomoxetine and placebo). An average of errors of commission and response reaction time in the Go-NoGo task was used as a measure of response inhibition, while an average of errors of omission, response reaction time and reaction time variability in the Continuous Performance Task was used as a measure of sustained attention. Average percent accuracy and reaction times for correct identification of facial expressions of fear in an emotion identification task were used to assess threat-related emotion identification. Raw scores for each measure were transformed into standardized scores using peer regression modeling, using established norms for populations aged six years and above (Williams et al., 2010a). This transformation enabled comparison across ages, provided a reference against normative functioning, and allowed computation of summary scores across different scales of measurement (i.e. accuracy and reaction time). Z scores ranged from −5 to +5 after winsorizing any outliers to these bounds, with 0 indicating average normative performance relative to age and gender. Raw scores for each task variable used to compute response inhibition, sustained attention and fear identification capacity scores are reported in Table S1.

These data were collected using a computerized cognitive performance battery (IntegNeuro™ (Clark et al., 2006)). This battery has been established as sensitive for identifying cognitive problems in ADHD (Williams et al., 2010a) and also has demonstrated test-retest reliability over a time frame of six weeks (Williams et al., 2005). In addition to the response inhibition, sustained attention and emotion identification tasks, IntegNeuro also includes tests of verbal memory, working memory, information processing speed, and planning.

2.5. Assessment of ADHD and anxiety symptom severity

Symptom ratings were obtained via rating scales administered to the participant and parent/guardian by study clinicians. ADHD symptom severity was assessed by using the total score on the parentcompleted ADHD-RS IV. The ADHD-RS-IV assesses symptoms according to the 18 DSM-IV criteria, with each symptom rated from 0 to 3. Anxiety symptom severity was rated using the Anxious-Shy subscale of the Conners Parent Rating Scale – Revised: Long version (CPRS; (Conners et al., 1998)). Standardized T-scores were used, with a score of 50 reflecting the age- and gender- normed median score, and every 10 points representing 1SD.

2.6. Treatment outcomes

Our primary study outcomes were differences in response inhibition, sustained attention, and fear facial emotion identification after 6 weeks of atomoxetine relative to 6 weeks of placebo (Tsang et al., 2011). Secondary outcome measures were differences in ADHD symptoms (ADHD-RS IV total score) and anxiety symptoms (CPRS anxiousshy score). Exploratory analyses were also conducted on the effect of atomoxetine on other cognitive domains assessed within the computerized battery (verbal memory – delayed recall accuracy on list of 12 words; working memory – maximum digit span; information processing speed – average of accuracy and completion time of switching of attention task; and planning – average of accuracy and completion time on a maze task).

2.6.1. Statistical analysis

Analyses were performed using IBM SPSS Statistics (version 24.0). We used a modified intent-to-treat approach, including data from all participants who completed the first treatment phase (Brams et al., 2012{Hainque, 2016 #963). This method was chosen as at least one outcome endpoint is required in order to apply restricted maximum likelihood estimation to analyze incomplete data using linear mixed models. To test the primary hypothesis that atomoxetine would improve response inhibition, sustained attention and fear identification, we used mixed linear models with sequence order (atomoxetine-placebo, placebo-atomoxetine), phase (1, 2) and treatment (atomoxetine, placebo) as fixed effects, and subject-within-sequence as a random effect. Linear mixed models are recommended for analyzing cross-over clinical trial data because they model longitudinal effects and handle missing data effectively. Restricted maximum likelihood estimation was used to analyze incomplete data. Cohen’s d effect sizes are reported for primary and secondary outcome measures.

The linear mixed model outlined above was also used to test our hypothesis that atomoxetine would improve clinical symptoms of both ADHD and anxiety relative to placebo. Cohen’s d effect sizes are reported for primary and secondary outcome measures.

Our second hypothesis was that atomoxetine would differentially impact response inhibition, sustained attention and fear identification for individuals with and without comorbid anxiety disorders. To test this, a treatment-by-anxiety disorder comorbidity term (presence, ANX + or absence, ANX-) was included as a fixed effect in the models.

Our third exploratory aim was to determine the relationships between any changes in cognition and changes in symptom scales. Absolute change scores were used for z scored cognitive measures, while percent change scores ((ATX - Placebo)/Placebo) were used for symptom measures. Relationships were assessed using Pearson’s bivariate correlations.

Exploratory follow-up analyses were conducted to evaluate potential covariates of response to atomoxetine, including age, baseline ADHD symptom severity, and baseline performance on the three primary outcome measures. These factor-by-treatment interaction terms (e.g. age x treatment) were added in separate models as a fixed effect. In exploratory analyses we also assessed atomoxetine efficacy for other general cognitive domains of verbal memory, working memory, information processing and planning using the basic linear mixed model described above.

3. Results

3.1. Participant characteristics

Of the 198 participants referred to the study, 136 (mean age 11 ± 2.6years, 80% males) began treatment with either atomoxetine (n = 63) or placebo (n = 73). The first phase was completed by 116 participants (modified intent-to-treat sample), of whom 109 completed the second, crossover phase. Forty-three participants (31.6%) had a comorbid anxiety disorder. The CONSORT diagram is shown in Fig. 1. Participant characteristics, including ADHD subtype and presence of comorbidities are presented in Table 1. At baseline, participants with a comorbid anxiety disorder had higher CPRS anxiety symptom severity, t (57.7) = −4.25, p < 0.001, and ADHD symptom severity, t (114) = −2.1, p = 0.038. There were no differences between those with comorbid anxiety versus those without in baseline capacity for response inhibition, sustained attention or fear identification.

Fig. 1.

Fig. 1.

Open in a new tab

CONSORT diagram.

Table 1.

Summary of demographic characteristics and baseline clinical scores, and response inhibition and fear identification capacity scores for total sample and those with (+) and without (−) comorbid anxiety. Mean (SD).

Total sample (n = 116) ADHD + ANX (n = 38) ADHD –ANX (n = 78)
Age, mean (SD) 11.29 (2.5) 11.2 (2.7) 11.3 (2.5)
Males, N (%) 91 (78.4) 28 (73.7) 63 (80.8)
ADHD subtype, N (%)
 Inattentive 45 (38.8) 11 (28.9) 34 (43.6)
 Hyperactive 4 (3.4) 1 (2.6) 3 (3.8)
 Combined 67 (57.7) 26 (68.4) 41 (52.6)
ADHD-RS IV Total/54 36.6 (10.0) 39.4 (9.2)* 35.3 (10.2)*
CPRS-L: Anxious-Shy/90 60.7 (15.1) 69.3 (16.5)* 56.5 (12.4)*
Anxiety Disorder
 Separation anxiety disorder 14 (12.1)
 Generalized anxiety disorder 23 (19.8)
 Social anxiety disorder 26 (22.4)
Other comorbidities, N (%)
 Conduct Disorder 14 (12.1) 5 9
 Major Depression 3 (2.6) 1 2
 Obsessive-Compulsive 10 (8.6) 7 3
 Oppositional Defiant 46 (40.0) 21 25
 Post-traumatic Stress Disorder 7 (6.0) 2 5
Behavioral Performance Measures
Response Inhibition Z-scorea −0.21 (0.56) −0.34 (0.48) −0.26 (0.50)
Sustained Attention Z-scorea −1.04 (0.85) −1.11 (0.67) −0.95 (0.81)
Fear Identification Z-scorea −0.56 (1.19) −0.59 (1.05) −0.54 (1.26)
Open in a new tab
a

Z-scores are expressed relative to healthy age and gender norms, with 0 representing healthy control average and higher scores indicating better performance. CPRS-L subscales are expressed as standardized T-scores, with a score of 50 reflecting the age- and gender- normed median score, and every 10 points representing 1SD.

3.2. Primary general and emotional cognitive outcomes

3.2.1. Response inhibition

Atomoxetine significantly improved response inhibition, assessed using the Go-NoGo test [F(1,115) = 13.22, p < 0.001, Cohen’s d = 0.42]. Specifically, improvements in the composite score were due to the reduction in NoGo errors of commission [F(1,117) = 9.31, p = 0.003] (Fig. 2).

Fig. 2. Treatment effects on response inhibition, sustained attention and fear identification.

Fig. 2.

Open in a new tab

Mean normalized scores (SEM) in each cognitive domain showing atomoxetine relative to placebo. Z-scores are expressed relative to healthy age and gender norms, with 0 representing healthy control average and higher scores indicating better performance. * significant treatment effect, p < 0.05.

3.2.2. Sustained attention

There were trend level effects of atomoxetine on sustained attention [F(1,102) = 3.56, p = 0.06, Cohen’s d = 0.17], due to the combination of small, statistically non-significant improvements in omission errors [F(1,104) = 1.00, p = 0.32], response reaction time [F(1,102) = 2.54, p = 0.11] and reaction time variability [F(1,109) = 2.94, p = 0.09] (Fig. 2).

3.2.3. Fear emotion identification

Atomoxetine significantly improved the identification of facial expressions of fear [F(1,109) = 4.59, p = 0.03, Cohen’s d = 0.22]. Although mean results showed that both accuracy and reaction time for fear identification improved with atomoxetine (Fig. 2), this effect was due primarily to a reduction of the reaction time for correctly identifying fearful faces [F(1,110) = 3.95, p = 0.04] (Fig. 2).

3.3. Secondary ADHD and anxiety symptom outcomes

Atomoxetine was associated with significantly reduced symptom severity for both ADHD [F(1,107) = 13.02, p < 0.001] and for anxiety [F(1,103) = 4.2, p = 0.043] (Table 2).

Table 2.

Summary of clinical outcomes after each treatment phase. ADHD-RS, Conners Parent Rating Scale (CPRS) Anxious-Shy and Clinical Global Impressions (CGI) scores were significantly reduced on atomoxetine relative to placebo in the total cohort. Mean (SD) and Cohen’s d effect sizes (ES).

Atomoxetine Placebo Cohen’s d effect size
ADHD-RS Total cohort 29.14 (13.1) 33.57 (12.0) 0.35
ADHD -ANX 27.54 (13.2) 31.48 (11.7) 0.31
ADHD + ANX ANXANX + +ANX 33.64 (12.2) 38.60 (11.3) 0.46
CPRS Anxiou-s-Shy Total cohort 56.8 (13.7) 59.6 (14.6) 0.20
ADHD -ANX 54.66 (12.6) 55.3 (11.7) 0.05
ADHD + ANX + ANX+ + +ANX + ANX 62.8 (14.8) 70.6 (15.9) 0.51
Open in a new tab

3.4. Effect of anxiety comorbidity on primary outcomes

Participants with ADHD “alone” versus ADHD with comorbid anxiety did not differ on primary outcomes for efficacy of atomoxetine [inhibition [F(1,116) = 0.05, p = 0.82]; sustained attention [F (1,102) = 1.28, p = 0.26]; or fear identification [F(1,109) = 0.24, p = 0.63]].

3.5. Effect of anxiety comorbidity on secondary outcomes

Atomoxetine was associated with improved anxiety severity significantly more for those with comorbid anxiety disorders than those without [ADHD –ANX, Cohen’s d = 0.05, ADHD + ANX, Cohen’s d = 0.51; F(1, 104) = 5.6, p = 0.02]. There were no differences between those with and without comorbid anxiety disorders in the magnitude of atomoxetine-related ADHD symptom improvement [F(1,108) = 0.00, p = 0.99].

3.6. Relationships between treatment-related changes in cognition and symptom scales

There were no significant correlations between treatment-related change in ADHD severity and change in response inhibition (r = 0.05, p = 0.65), sustained attention (r = 0.03, p = 0.76) or fear identification (r = 0.06, p = 0.54). Change in anxiety was also not significantly correlated with change in response inhibition (r = −0.00, p = 0.99), sustained attention (r = −0.17, p = 0.11), or fear identification (r = −0.02, p = 0.87).

3.7. Exploratory analysis of potential covariates and additional cognitive outcomes

For response inhibition and sustained attention, treatment efficacy was not moderated by age, baseline symptom severity or baseline cognitive performance. Treatment efficacy for fear identification was moderated by age [F(1,108) = 6.24, p = 0.01] and baseline capacity for fear identification [F(1,116) = 5.15, p = 0.03]. Younger children and those with poor fear identification capacity at baseline experienced significant improvements, while older children/adolescents and those without baseline impairments did not (see supplementary materials).

Atomoxetine did not impact other aspects of general cognition, including verbal memory [F(1,100) = 0.17, p = 0.68], working memory [F(1,103) = 0.40, p = 0.53], information processing speed [F(1,104) = 3.67, p = 0.06] or planning [F(1,107) = 2.15, p = 0.15] composite scores (Fig. S1).

There was a significant effect of phase for sustained attention [F(1,102) = 8.92, p = 0.004] and ADHD symptom severity [F(1,107) = 7.4, p = 0.01] (Table S2).

4. Discussion

We report on the first placebo-controlled trial to examine the effects of atomoxetine on objective measures of response inhibition, sustained attention and identification of fearful faces in youth with ADHD with and without comorbid anxiety disorders. Relative to placebo, atomoxetine treatment resulted in moderate improvement in response inhibition but no significant improvement in sustained attention. Atomoxetine improved identification of fearful faces, particularly in younger children and those with pre-existing deficits. For the one-third of participants with a comorbid anxiety disorder, atomoxetine was more effective at reducing symptoms of anxiety than in those with ADHD alone. The presence of a comorbid anxiety disorder did not affect the impact of treatment on other outcome measures. Atomoxetine-related changes in cognition were not significantly correlated to changes in clinical symptom measures.

Our first finding was that atomoxetine improved response inhibition in young people with ADHD as assessed by behavioral performance on a Go-NoGo test. This finding extends upon previous observations from open-label designs that atomoxetine impacts inhibitory aspects of executive function in both children (Gau and Shang, 2010a; b) and adults (Chamberlain et al., 2007; Faraone et al., 2005; Spencer et al., 1998b) with ADHD.

We also found that atomoxetine, relative to placebo, improved the identification of fearful faces in the same young people with ADHD. This effect was on averaged accuracy and reaction time performance, but was driven primarily by a decrease in reaction time. Previously we have observed that stimulants, in contrast, have little impact on fear identification in young people with ADHD (Williams et al., 2008).

We did not find significant improvements in sustained attention with atomoxetine relative to placebo. Despite a trend-level improvement in sustained attention, participants remained impaired relative to age-matched peers. This is in harmony with findings from Bedard et al. (2015) that methylphenidate is more likely to be effective for improving sustained attention than atomoxetine.

Atomoxetine reduced symptoms of both ADHD and anxiety. The effect size of improvement in ADHD symptoms was lower than that found in a meta-analysis of previous studies (−0.35 relative to −0.64) (Schwartz and Correll, 2014). This difference may be exaggerated by the fact that some previous studies excluded participants who responded to placebo during a lead-in phase, potentially enhancing the contrast between atomoxetine and placebo in those studies (e.g. (Adler et al., 2009; Geller et al., 2007)). Additionally, we observed a modest phase effect in our study that tended to reduce the overall effect size. Atomoxetine was also effective in reducing anxiety symptom severity, with an effect size consistent with previous studies (Geller et al., 2007; Kratochvil et al., 2005; Sumner et al., 2006).

Contrary to expectations, we did not observe a differential impact of atomoxetine on the primary cognitive outcomes in ADHD as a function of comorbid anxiety disorders. This null effect was not accounted for by any observed baseline differences in cognitive capacity between subgroups of patients with and without comorbid anxiety disorders. Rather, our findings suggest that atomoxetine may have a specific effect on response inhibition and fear identification regardless of whether young people with ADHD also meet criteria for an anxiety disorder.

Our data suggest a general dissociation in the effects of atomoxetine on cognitive and clinical symptom measures. This finding replicates and extends upon the previously reported lack of concordance between treatment-related improvements in clinical behaviour and CPT performance (Bedard et al., 2015). While this result suggests that the cognitive effects are not mediating clinical improvement, cognition may be a valid treatment target in itself given the association with additive functional impairment, high test-retest reliability and the ability to map cognitive functions to specific neural circuits (Weafer et al., 2013; Williams et al., 2005). It is also possible that cognition mediates more complex interactions between specific symptom changes, rather than an overall clinical scale, or that clinical manifestation is driven by complex environmental interactions with cognitive capacities.

Future studies might consider stratifying patients according to cognitive profiles and undertake trials that determine prospectively whether individuals with specific cognitive problems respond preferentially to atomoxetine. Studies of ADHD youth suggest that data-driven approaches applied at the pre-treatment baseline differentiate clusters of individuals defined by response inhibition deficits from those with sustained attention deficits (Leikauf et al., 2017; van Hulst et al., 2015). Such approaches may be valuable for pursuing biomarkerguided treatment outcomes. Additional biomarkers could be incorporated to flesh out the neural circuit and neuro-genetic basis of atomoxetine effects, including neuroimaging probes and variation in genes that modulate the adrenergic system (Yang et al., 2013).

We recognize limitations in the design of the ACTION study. A 1-week washout period between phases may have been too brief, resulting in a carryover effect of atomoxetine and thus less sensitivity to observe improvements in our outcome measures. We did observe a phase effect in our data such that ADHD symptom severity was lower in phase 2 than 1, irrespective of being on atomoxetine or placebo. Nonetheless, our results were reproduced when analyzing between-group treatment effects in phase one only. A second potential design limitation is that the 6-week trial period may have been of insufficient duration to reveal the full therapeutic effects of atomoxetine. This duration was chosen to reduce unnecessary burden to participants of an extended placebo phase, as recommended by our institutional review board. We note that studies published subsequent to our design choice have shown that atomoxetine-related improvements can accrue for 12 weeks (Clemow and Bushe, 2015). In future studies, the scientific value and longer-term benefit of investigating the effect of atomoxetine needs to be evaluated relative to the practicalities and patient burden of increasing the duration of washout and placebo periods. Third, while the study was adequately powered to evaluate the primary aims, the sample size was not sufficient to determine whether atomoxetine has particular benefit on cognition for individual anxiety disorders comorbid with ADHD. Consistent with previous work, we grouped together separation anxiety, generalized anxiety disorder and social anxiety disorder (Geller et al., 2007). Future studies could be designed to test moderating effects of individual disorders or other comorbidities.

The present findings advance our knowledge of the impact of atomoxetine on improving specific aspects of general and emotional cognition in children and adolescents with ADHD. Because a stringently controlled design was employed we have confidence in the indications from this study that atomoxetine improved response inhibition and fear identification relative to placebo. These findings contribute to the foundation for developing personalized approaches to planning ADHD intervention choices according to the cognitive needs of each young person experiencing ADHD.

Supplementary Material

Supplementary Tables and Appendix
Supplementary Figure F1

Acknowledgments

We thank Elizabeth Barnes for her assistance with statistical analyses.

Funding and disclosures

This study is funded by a National Health and Medical Research Council (NHMRC) grant (ID: 457424) by the Australian Government. Dr Griffiths is supported by the Tenix Fellowship from the Westmead Medical Research Foundation and a National Health and Medical Research Council Early Career Researcher Fellowship [1122842]. John Leikauf’s effort was supported by the National Institute of Mental Health training grant [5 T32 MH19908 22].

MRK is a member of the Strattera (Atomoxetine) Advisory Board for Eli Lilly Pty. Ltd. The atomoxetine used in the ACTION study was purchased from Eli Lilly Pty. Ltd. The remaining authors declare they have no competing interests.

Footnotes

Trial registration

Trial Registry Name:Australia and New Zealand Clinical Trial Registry.

Registration Number: ANZCTR 12607000535471.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jpsychires.2018.03.009.

References

  1. Adler LA, Liebowitz M, Kronenberger W, Qiao M, Rubin R, Hollandbeck M, Deldar A, Schuh K, Durell T, 2009. Atomoxetine treatment in adults with attention-deficit/hyperactivity disorder and comorbid social anxiety disorder. Depress. Anxiety 26 (3), 212–221. [DOI] [PubMed] [Google Scholar]
  2. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. American Psychiatric Pub. [Google Scholar]
  3. Association, A.P, 2013. Diagnostic and Statistical Manual of Mental Disorders (DSM-5®). American Psychiatric Pub. [Google Scholar]
  4. Barry RJ, Clarke AR, Hajos M, McCarthy R, Selikowitz M, Bruggemann JM, 2009. Acute atomoxetine effects on the EEG of children with attention-deficit/hyperactivity disorder. Neuropharmacology 57 (7), 702–707. [DOI] [PubMed] [Google Scholar]
  5. Bedard AC, Stein MA, Halperin JM, Krone B, Rajwan E, Newcorn JH, 2015. Differential impact of methylphenidate and atomoxetine on sustained attention in youth with attention-deficit/hyperactivity disorder. JCPP (J. Child Psychol. Psychiatry) 56 (1), 40–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Biederman J, Petty C, Fried R, Fontanella J, Doyle AE, Seidman LJ, Faraone SV, 2006. Impact of psychometrically defined deficits of executive functioning in adults with attention deficit hyperactivity disorder. Am. J. Psychiatr. 163 (10), 1730–1738. [DOI] [PubMed] [Google Scholar]
  7. Bloemsma JM, Boer F, Arnold R, Banaschewski T, Faraone SV, Buitelaar JK, Sergeant JA, Rommelse N, Oosterlaan J, 2013. Comorbid anxiety and neurocognitive dysfunctions in children with ADHD. Eur. Child Adolesc. Psychiatr. 22 (4), 225–234. [DOI] [PubMed] [Google Scholar]
  8. Brams M, Turnbow J, Pestreich L, Giblin J, Childress A, McCague K, Muniz R, 2012. A randomized, double-blind study of 30 versus 20 mg dexmethylphenidate extended-release in children with attention-deficit/hyperactivity disorder: late-day symptom control. J. Clin. Psychopharmacol. 32 (5), 637–644. [DOI] [PubMed] [Google Scholar]
  9. Castellanos FX, Glaser PE, Gerhardt GA, 2006. Towards a neuroscience of attentiondeficit/hyperactivity disorder: fractionating the phenotype. J. Neurosci. Meth. 151 (1), 1–4. [DOI] [PubMed] [Google Scholar]
  10. Chamberlain SR, Del Campo N, Dowson J, Muller U, Clark L, Robbins TW, Sahakian BJ, 2007. Atomoxetine improved response inhibition in adults with attention deficit/hyperactivity disorder. Biol. Psychiatr. 62 (9), 977–984. [DOI] [PubMed] [Google Scholar]
  11. Childress AC, 2016. A critical appraisal of atomoxetine in the management of ADHD. Therapeut. Clin. Risk Manag. 12, 27–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Clark CR, Paul RH, Williams LM, Arns M, Fallahpour K, Handmer C, Gordon E, 2006. Standardized assessment of cognitive functioning during development and aging using an automated touchscreen battery. Arch. Clin. Neuropsychol. 21 (5), 449–467. [DOI] [PubMed] [Google Scholar]
  13. Clemow DB, Bushe CJ, 2015. Atomoxetine in patients with ADHD: a clinical and pharmacological review of the onset, trajectory, duration of response and implications for patients. J. Psychopharmacol. 29 (12), 1221–1230. [DOI] [PubMed] [Google Scholar]
  14. Conners CK, Sitarenios G, Parker JDA, Epstein JN, 1998. The revised Conners’ Parent Rating Scale (CPRS-R): factor structure, reliability, and criterion validity. J. Abnorm. Child Psychol. 26 (4), 257–268. [DOI] [PubMed] [Google Scholar]
  15. Cubillo A, Smith AB, Barrett N, Giampietro V, Brammer MJ, Simmons A, Rubia K, 2014. Shared and drug-specific effects of atomoxetine and methylphenidate on inhibitory brain dysfunction in medication-naive ADHD boys. Cerebr. Cortex 24 (1), 174–185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. DuPaul GJ, 1991. Parent and teacher ratings of ADHD symptoms: psychometric properties in a community-based sample. J. Clin. Child Psychol. 20 (3), 245–253. [Google Scholar]
  17. Faraone S, Biederman J, Spencer T, Michelson D, Adler L, Reimherr F, Seidman L, 2005. Atomoxetine and Stroop task performance in adult attention-deficit/hyperactivity disorder. J. Child Adolesc. Psychopharmacol. 15 (4), 664–670. [DOI] [PubMed] [Google Scholar]
  18. Gabriel A, Violato C, 2011. Adjunctive atomocaxetine to SSRIs or SNRIs in the treatment of adult ADHD patients with comorbid partially responsive generalized anxiety (GA): an open-label study. ADHD Atten. Deficit Hyperactivity Disord. 3 (4), 319–326. [DOI] [PubMed] [Google Scholar]
  19. Garnock-Jones KP, Keating GM, 2011. Spotlight on atomoxetine in attention-deficit hyperctivity disorder in children and adolescents. CNS Drugs 24 (1), 85–88. [DOI] [PubMed] [Google Scholar]
  20. Gau SS-F, Shang C-Y, 2010a. Executive functions as endophenotypes in ADHD: evidence from the cambridge neuropsychological test battery (CANTAB). JCPP (J. Child Psychol. Psychiatry) 51 (7), 838–849. [DOI] [PubMed] [Google Scholar]
  21. Gau SS, Shang CY, 2010b. Improvement of executive functions in boys with attention deficit hyperactivity disorder: an open-label follow-up study with once-daily atomoxetine. Int. J. Neuropsychopharmacol. 13 (2), 243–256. [DOI] [PubMed] [Google Scholar]
  22. Geller D, Donnelly C, Lopez F, Rubin R, Newcorn J, Sutton V, Bakken R, Paczkowski M, Kelsey D, Sumner C, 2007. Atomoxetine treatment for pediatric patients with attention-deficit/hyperactivity disorder with comorbid anxiety disorder. J. Am. Acad. Child Adolesc. Psychiatry 46 (9), 1119–1127. [DOI] [PubMed] [Google Scholar]
  23. Hainque E, Vidailhet M, Cozic N, Charbonnier-Beaupel F, Thobois S, Tranchant C, Brochard V, Glibert G, Drapier S, Mutez E, 2016. A randomized, controlled, double-blind, crossover trial of zonisamide in myoclonus-dystonia. Neurology 86 (18), 1729–1735. [DOI] [PubMed] [Google Scholar]
  24. Hinshaw SP, 2007. Moderators and mediators of treatment outcome for youth with ADHD: understanding for whom and how interventions work. J. Pediatr. Psychol. 32 (6), 664–675. [DOI] [PubMed] [Google Scholar]
  25. Insel T, Cuthbert B, Garvey M, Heinssen R, Pine DS, Quinn K, Sanislow C, Wang P, 2010. Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am. J. Psychiatr. 167 (7), 748–751. [DOI] [PubMed] [Google Scholar]
  26. Jensen PS, Hinshaw SP, Kraemer HC, Lenora N, Newcorn JH, Abikoff HB, March JS, Arnold LE, Cantwell DP, Conners CK, Elliott GR, Greenhill LL, Hechtman L, Hoza B, Pelham WE, Severe JB, Swanson JM, Wells KC, Wigal T, Vitello B, 2001a. ADHD comorbidity findings from the MTA study: comparing comorbid subgroups. J. Am. Acad. Child Adolesc. Psychiatry 40 (2), 147–158. [DOI] [PubMed] [Google Scholar]
  27. Jensen PS, Hinshaw SP, Swanson JM, Greenhill LL, Conners CK, Arnold LE, Abikoff HB, Elliott G, Hechtman L, Hoza B, March JS, Newcorn JH, Severe JB, Vitiello B, Wells K, Wigal T, 2001b. Findings from the NIMH Multimodal Treatment Study of ADHD (MTA): implications and applications for primary care providers. J. Dev. Behav. Pediatr. 22 (1), 60–73. [DOI] [PubMed] [Google Scholar]
  28. Karalunas SL, Geurts HM, Konrad K, Bender S, Nigg JT, 2014. Annual research review: reaction time variability in ADHD and autism spectrum disorders: measurement and mechanisms of a proposed trans-diagnostic phenotype. JCPP (J. Child Psychol. Psychiatry) 55 (6), 685–710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kratochvil CJ, Newcorn JH, Arnold LE, Duesenberg D, Emslie GJ, Quintana H, Sarkis EH, Wagner KD, Gao H, Michelson D, 2005. Atomoxetine alone or combined with fluoxetine for treating ADHD with comorbid depressive or anxiety symptoms. J. Am. Acad. Child Adolesc. Psychiatry 44 (9), 915–924. [DOI] [PubMed] [Google Scholar]
  30. Leikauf JEGKR, Saggar M, Hong DS, Clarke S, Efron D, Tsang TW, Hermens DF, Kohn MR, Williams LM, July 2017. Identification of biotypes in attention-deficit/hyperactivity disorder a report from a randomized, controlled trial. Pers. Med. Psychiatr. 3, 8–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Manassis K, Tannock R, Barbosa J, 2000. Dichotic listening and response inhibition in children with comorbid anxiety disorders and ADHD. J. Am. Acad. Child Adolesc. Psychiatry 39 (9), 1152–1159. [DOI] [PubMed] [Google Scholar]
  32. Mueller A, Hong DS, Shepard S, Moore T, 2017. Linking ADHD to the neural circuitry of attention. Trends Cognit. Sci. 21 (6), 474–488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Ni HC, Shang CY, Gau SS, Lin YJ, Huang HC, Yang LK, 2013. A head-to-head randomized clinical trial of methylphenidate and atomoxetine treatment for executive function in adults with attention-deficit hyperactivity disorder. Int. J. Neuropsychopharmacol. 16 (9), 1959–1973. [DOI] [PubMed] [Google Scholar]
  34. Nigg JT, 2005. Neuropsychologic theory and findings in attention-deficit/hyperactivity disorder: the state of the field and salient challenges for the coming decade. Biol. Psychiatr. 57 (11), 1424–1435. [DOI] [PubMed] [Google Scholar]
  35. Pliszka SR, Crismon ML, Hughes CW, Corners CK, Emslie GJ, Jensen PS, McCracken JT, Swanson JM, Lopez MD, Texas Consensus Conference Panel on Pharmacotherapy of Childhood Attention Deficit Hyperactivity, 2006. The Texas children’s medication algorithm project: revision of the algorithm for pharmacotherapy of attention-deficit/hyperactivity disorder. J. Am. Acad. Child Adolesc. Psychiatry 45 (6), 642–657. [DOI] [PubMed] [Google Scholar]
  36. Schatz DB, Rostain AL, 2006. ADHD with comorbid anxiety: a review of the current literature. J. Atten. Disord. 10 (2), 141–149. [DOI] [PubMed] [Google Scholar]
  37. Schwartz S, Correll CU, 2014. 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 (2), 174–187. [DOI] [PubMed] [Google Scholar]
  38. Shang CY, Gau SS, 2012. Improving visual memory, attention, and school function with atomoxetine in boys with attention-deficit/hyperactivity disorder. J. Child Adolesc. Psychopharmacol. 22 (5), 353–363. [DOI] [PubMed] [Google Scholar]
  39. Shaw P, Stringaris A, Nigg J, Leibenluft E, 2014. Emotion dysregulation in attention deficit hyperactivity disorder. Am. J. Psychiatr. 171 (3), 276–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Silverman WK, Saavedra LM, Pina AA, 2001. Test-retest reliability of anxiety symptoms and diagnoses with the Anxiety Disorders Interview Schedule for DSM-IV: child and parent versions. J. Am. Acad. Child Adolesc. Psychiatry 40 (8), 937–944. [DOI] [PubMed] [Google Scholar]
  41. Spencer T, Biederman J, Wilens T, Prince J, Hatch M, Jones J, Harding M, Faraone SV, Seidman L, 1998a. Effectiveness and tolerability of tomoxetine in adults with attention deficit hyperactivity disorder. Am. J. Psychiatr. 155, 693–695. [DOI] [PubMed] [Google Scholar]
  42. Spencer T, Biederman J, Wilens T, Prince J, Hatch M, Jones J, Harding M, Faraone SV, Seidman L, 1998b. Effectiveness and tolerability of tomoxetine in adults with attention deficit hyperactivity disorder. Am. J. Psychiatr. 155 (5), 693–695. [DOI] [PubMed] [Google Scholar]
  43. Spencer TJ, Biederman J, Heiligenstein J, Wilens TE, Faries DE, Prince J, Faraone SV, Rea J, Witcher J, Zervas S, 2001. An open-label, dose-ranging study of atomoxetine in children with attention deficit hyperactivity disorder. J. Child Adolesc. Psychopharmacol. 11 (3), 251–265. [DOI] [PubMed] [Google Scholar]
  44. Steinhausen H-C, Nøvik TS, Baldursson G, Curatolo P, Lorenzo MJ, Pereira RR, Ralston SJ, Rothenberger A, 2006. Co-existing psychiatric problems in ADHD in the ADORE cohort. Eur. Child Adolesc. Psychiatr. 15 (1), i25–i29. [DOI] [PubMed] [Google Scholar]
  45. Sumner C, Sher L, Sutton V, Bakken R, Paczkowski M, Trzepacz P, Kelsey D, 2006. Atomoxetine treatment for pediatric patients with ADHD and comorbid anxiety. Neuropediatrics 37 (S1), TP74. [DOI] [PubMed] [Google Scholar]
  46. Tsang TW, Kohn MR, Efron D, Clarke SD, Clark CR, Lamb C, Williams LM, 2015. Anxiety in young people with ADHD: clinical and self-report outcomes. J. Atten. Disord. 19 (1), 18–26. [DOI] [PubMed] [Google Scholar]
  47. Tsang TW, Kohn MR, Hermens DF, Clarke SD, Clark CR, Efron D, Cranswick N, Lamb C, Williams LM, 2011. A randomized controlled trial investigation of a non-stimulant in attention deficit hyperactivity disorder (ACTION): rationale and design. Trials 12, 77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. van Hulst BM, de Zeeuw P, Durston S, 2015. Distinct neuropsychological profiles within ADHD: a latent class analysis of cognitive control, reward sensitivity and timing. Psychol. Med. 45 (4), 735–745. [DOI] [PubMed] [Google Scholar]
  49. Vloet TD, Konrad K, Herpertz-Dahlmann B, Polier GG, Gunther T, 2010. Impact of anxiety disorders on attentional functions in children with ADHD. J. Affect. Disord. 124 (3), 283–290. [DOI] [PubMed] [Google Scholar]
  50. Weafer J, Baggott MJ, de Wit H, 2013. Test-retest reliability of behavioral measures of impulsive choice, impulsive action, and inattention. Exp. Clin. Psychopharmacol 21 (6), 475–481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Willcutt EG, Doyle AE, Nigg JT, Faraone SV, Pennington BF, 2005. Validity of the executive function theory of attention-deficit/hyperactivity disorder: a metaanalytic review. Biol. Psychiatr. 57 (11), 1336–1346. [DOI] [PubMed] [Google Scholar]
  52. Williams LM, Hermens DF, Palmer D, Kohn M, Clarke S, Keage H, Clark CR, Gordon E, 2008. Misinterpreting emotional expressions in attention-deficit/hyperactivity disorder: evidence for a neural marker and stimulant effects. Biol. Psychiatr. 63 (10), 917–926. [DOI] [PubMed] [Google Scholar]
  53. Williams LM, Hermens DF, Thein T, Clark CR, Cooper NJ, Clarke SD, Lamb C, Gordon E, Kohn MR, 2010a. Using brain-based cognitive measures to support clinical decisions in ADHD. Pediatr. Neurol. 42 (2), 118–126. [DOI] [PubMed] [Google Scholar]
  54. Williams LM, Simms E, Clark CR, Paul RH, Rowe D, Gordon E, 2005. The test-retest reliability of a standardized neurocognitive and neurophysiological test battery: neuromarker. Int. J. Neurosci. 115 (12), 1605–1630. [DOI] [PubMed] [Google Scholar]
  55. Williams LM, Tsang TW, Clarke S, Kohn M, 2010b. An ‘integrative neuroscience’ perspective on ADHD: linking cognition, emotion, brain and genetic measures with implications for clinical support. Expert Rev. Neurother. 10 (10), 1607–1621. [DOI] [PubMed] [Google Scholar]
  56. Williams LM, Tsang TW, Clarke S, Kohn M, 2010c. An ‘Integrative Neuroscience’ perspective on ADHD: linking cognition, emotion, brain and genetics measures with implications for clinical support. Expert Rev. Neurother. 10 (10), 1607–1621. [DOI] [PubMed] [Google Scholar]
  57. Yang L, Qian Q, Liu L, Li H, Faraone SV, Wang Y, 2013. Adrenergic neurotransmitter system transporter and receptor genes associated with atomoxetine response in attention-deficit hyperactivity disorder children. J. Neural. Transm. 120 (7), 1127–1133. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Tables and Appendix
NIHMS1708527-supplement-Supplementary_Tables_and_Appendix.docx (20.2KB, docx)
Supplementary Figure F1
NIHMS1708527-supplement-Supplementary_Figure_F1.jpg (265.4KB, jpg)

RESOURCES