Neuropsychological Outcomes Across the Day in Children with Attention-Deficit/Hyperactivity Disorder Treated with Atomoxetine: Results from a Placebo-Controlled Study Using a Computer-Based Continuous Performance Test Combined with an Infra-Red Motion-tracking Device

Abstract

The effect of atomoxetine (ATX) on executive function has been assessed by means of questionnaires only. The aim of this study was therefore to evaluate the efficacy of ATX using standard variables of a computer-based continuous performance test (cb-CPT) combined with an infra-red motion-tracking device at different times of the day. One hundred twenty-eight girls and boys aged 6 to 12 years with a diagnosis of ADHD according to DSM-IV-TR criteria were randomized in the study. The primary efficacy measures were the q-scores of the cb-CPT combined with an infra-red motion-tracking device. The test comprises 13 neuropsychological variables that can be taken to reflect hyperactivity, inattention, or impulsivity. One hundred five patients completed the study (ATX group: n=54; placebo group: n=51). ATX (target dose 1.2 mg/kg/day) over 8 weeks was significantly superior to placebo in reducing hyperactivity, inattention, and impulsivity as measured by q-scores of 10 primary variables of the cb-CPT. Both groups of patients showed a circadian pattern of neuropsychological outcomes across the day as reflected by the cb-CPT combined with an infra-red motion-tracking device. In summary, this study demonstrated a positive effect of ATX on some aspects of executive function, inhibitory control, and hyperactivity compared with placebo.

Introduction

A ttention-deficit/hyperactivity disorder (ADHD) is one of the most common psychiatric disorders in childhood and adolescence (Biederman and Faraone 2005) that affects around 5.26% of school-age children (Polanczyk et al. 2007). The disorder is characterized by age-inappropriate inattention, hyperactivity, and impulsivity Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV-TR) (American Psychiatric Association 2000). Genetic and imaging studies have linked the symptoms of ADHD to abnormalities of the dopaminergic and adrenergic transmitter systems (di Michele et al. 2005). Results from animal and human studies implicate a dysregulation of frontal-subcortical-cerebellar catecholaminergic circuits, leading to impaired neurotransmission (Arnsten 2009) and to impaired executive function, including working memory, response inhibition, set shifting, abstraction, planning, organization, fluency, and attention (Doyle 2006; Arnsten 2009). Executive function involves the interaction of cortical (frontal, temporal, parietal), subcortical (limbic, basal ganglia), and functional systems, including connections between the basal ganglia, thalamus, and frontal lobes (Riccio et al. 2002). Continuous performance tests (CPTs) have demonstrated sensitivity to executive dysfunction (Riccio et al. 2002) and therefore represent one method for studying different components of executive function, especially the attention system (Teicher et al. 2004).

As ADHD usually affects the patient in various situations and at various times of the day (Coghill et al. 2008; Wehmeier et al. 2009), long-acting medications have increasingly been used for treatment (Wolraich et al. 2001; Döpfner et al. 2004; Kelsey et al. 2004; Swanson et al. 2004; Banaschewski et al. 2006). The clinical efficacy of these medications may, however, vary over the day, depending on the respective pharmacokinetic and pharmacodynamic profiles (Döpfner et al. 2004; Kelsey et al. 2004; Sonuga-Barke et al. 2004; Wehmeier et al. 2009). Atomoxetine (ATX) is one such long-acting treatment option (Banaschewski et al. 2004; Cheng et al. 2007), the effectiveness of which has been shown to persist into the evening hours and until the next morning (Kelsey et al. 2004; Wehmeier et al. 2009).

The effect of ATX on executive function has been assessed by means of questionnaires only (Maziade et al. 2009). The aim of this study was therefore to evaluate the efficacy of ATX using standard variables of a computer-based CPT (cb-CPT) combined with an infra-red motion-tracking device. The primary objective was to compare ATX to placebo in a randomized, double-blind, placebo-controlled study using an objective methodology to measure ADHD symptoms (Wehmeier et al. in press). A secondary objective was to investigate the cb-CPT results at different times of the day. The results of this secondary objective will be reported here.

Methods

Study design

This randomized, double-blind, placebo-controlled, two-arm, multicenter study was approved by an ethics review board (University of Cologne) and conducted (Clinical Trial Registry Number: NCT00546910) according to the International Conference on Harmonization (ICH) Good Clinical Practice guideline. The 16 study sites were located all over Germany and included 3 child and adolescent psychiatric university clinics, 1 child and adolescent psychiatric clinic, and 12 private practices for child and adolescent psychiatry and/or pediatrics. Informed consent was obtained from the legal representatives and informed assent from the child. Eligible patients were randomized to 8 weeks of treatment with ATX starting at 0.5 mg/kg per day for 1 week, followed by 7 weeks on the standard target dose of 1.2 mg/kg per day or placebo (administered in capsules with identical appearance to study drug). Medication was given once daily in the morning. The cb-CPT was carried out in the morning (before taking the medication to assess the effect approximately 24 hours after intake), at noon, and at the late afternoon/early evening. Efficacy and tolerability assessments were performed at baseline and after 1, 2, 4, 6, and 8 weeks.

Patients

Eligible were girls and boys aged 6 to 12 years with a diagnosis of ADHD according to DSM-IV-TR criteria (American Psychiatric Association 2000). The diagnosis was confirmed using the “Diagnose-Checkliste Hyperkinetische Störungen” (Diagnostic Checklist for Hyperkinetic Disorders), a structured instrument that is routinely used for the diagnostic assessment of ADHD in Germany (Döpfner and Lehmkuhl 2000). The items of this instrument correspond to those of the ADHD Rating Scale (ADHD-RS) (DuPaul et al. 1998; Faries et al. 2001). Exclusion criteria comprised previous treatment with ATX, treatment with psychotropic medication other than the study drug, clinically relevant over- and underweight, a history of bipolar disorder, psychosis, pervasive developmental disorder, seizure disorder (other than febrile seizures), serious suicidal risk, and other relevant acute or unstable medical conditions. Psychotherapy initiated before the study was acceptable.

Efficacy measures

The primary efficacy measures were the q-scores of the cb-CPT combined with an infra-red motion-tracking device (QbTest, provided by Qbtech, Gothenburg, Sweden; www.qbtech.se/products/qbtest). This method (referred to only as “cb-CPT” below) was developed on the basis of an earlier approach to measuring ADHD core symptoms, assessing attention, impulsivity, and hyperactivity objectively (Teicher 1995; Teicher et al. 1996, 2000, 2003, 2004; Heiser et al. 2004, 2006; Faedda and Teicher 2005). The test was shown to have ecological validity in that it correlates well with clinical response (Tabori-Kraft et al. 2007) and could identify the medication dose that produced the best overall clinical results (Teicher et al. 2008).

For the test, the patient was seated in front of a screen, wearing a head reflector. An infra-red camera, placed above the computer monitor, tracked all movements made during the test. The cb-CPT task was to press a button on the appearance of a target (gray circle) on the screen, but not on the appearance of a nontarget (crossed gray circle). The test duration was 15 minutes, with 450 stimuli appearing at random (equal numbers of targets and nontargets) at two-second intervals. The response profile, the kind and number of errors, and the motion pattern indicate the type and severity of symptoms.

The cb-CPT comprises the following 13 neuropsychological variables that can be taken to reflect hyperactivity, inattention, or impulsivity:

(a) Hyperactivity:

• Time active (TA) is the time (%) during which the patient has moved his or her head for a distance of >1 cm/second. TA is an inverse measure of the patient's ability to sit still.

• Distance (DIS) is the length of the path described by the headband reflector during the test. DIS is an indicator of the amount of total activity during the test.

• Area (AR) is the area covered by the headband reflector movements during the test. AR reflects how vivid the movements were during the test.

• Microevents (MEs) are the number of position changes of the head of >1 mm during the test. A large number of ME indicate a high degree of activity.

• Motion simplicity (MS) reflects the degree to which head movements include directional change. MS provides an inverse measure of the complexity of movement patterns (Paulus and Geyer 1991).

(b) Inattention:

• Mean reaction time (mRT) is the average elapsed time from stimulus presentation to response. It reflects the speed of information processing and response execution (Tannock 1998). mRT relates to the time component of executive function (Denckla 1996).

• Reaction time variation (RTV) is the standard deviation (SD) of the mRT. This measure reflects the fluctuation in reaction time performance. High performance variability reflects symptoms such as difficulty in sustaining attention, forgetfulness, disorganization, and careless errors (Castellanos et al. 2005).

• Normalized variation of reaction time (nVRT) is RTV expressed as the SD in units of mRT (RTV/mRT). Slow mRT often results in high variability (Salthouse 1993). nVRT corrects for this effect and is therefore a validity measure for RTV. A normal score on nVRT together with a high score on RTV indicates that high RTV is a result of a long mRT.

• Omission error (OE) is the proportion of nonresponses registered when the stimulus was a target (as opposed to the stimulus being a nontarget). OE represents inattention and inability to remain focused on the task. Increased OE is related to selective attention and deficient arousal (Barkley 1991).

• Commission error (CE) is the proportion of responses registered when the stimulus was a nontarget (as opposed to the stimulus being a target). CE is a result of anticipatory or incomplete processing of the stimulus and can be considered a measure of impulsive behavior (Halperin et al. 1991).

• Anticipatory response (ANT) is a response given shortly before a stimulus is presented. ANT reflects anticipation of a target rather than being a response to a stimulus (Rubia et al. 1999).

(d) Further variables:

• Error rate (ER) is the overall proportion of incorrect responses (OEs plus CEs) to the correct responses (Halperin et al. 1991).

• Multiresponse (MR) occurs when the response button was pressed more than once after the presentation of a stimulus. MR reflects hyper-responsivity (Tannock 1998).

The scores derived from the cb-CPT are standardized scores, so-called q-scores, which are adjusted for age and gender effects. In the general population, the q-scores have a mean of 0 and an SD of 1. Higher scores indicate more severe symptoms for all cb-CPT variables (i.e., all variables are calculated to point into the same direction).

Further efficacy measures assessed in this study were as follows: (a) ADHD-RS-IV-Parent Version: Investigator-Administered and Scored (DuPaul et al. 1998; Faries et al. 2001); (b) Clinical Global Impressions-Severity (CGI-S) Scale (Guy 1976; NIMH 1985); (c) Weekly Ratings of Morning and Evening Behavior-Revised-Investigator Rated (WREMB-R-Inv) Scale, which reflects ADHD-related behavioral problems in the evening and the morning (Wehmeier et al. 2009). The WREMB was used to obtain information on symptoms across the day. The scale allows the weekly assessment of behavioral symptoms as rated by the investigator, based on parent questioning. The WREMB has three subscales: “Morning,” “late noon and evening,” and a single item “difficulty falling asleep” (Kelsey et al. 2004; Carlson et al. 2007; Wehmeier et al. 2009). The scale is based on the Daily Rating of Evening and Morning Behavior-Revised (DPREMB-R) Scale (Kelsey et al. 2004). The DPREMB-R was validated for the assessment of ADHD-related behaviors (Sutton et al. 2003) and has been used in several studies (Michelson et al. 2002; Kelsey et al. 2004). The WREMB-R measures 11 specific behaviors (e.g., getting up and out of bed, doing or completing homework, sitting through dinner). The 11 items are shown in Table 1. The “late noon and evening” and “morning” subscores comprise seven and three items, respectively. The “difficulties falling asleep” subscore is based on one item. The possible score for each item ranges from 0 (no difficulty) to 3 (a lot of difficulty).

Table 1.

The 11 Items of the Weekly Rating of Evening and Morning Behavior—Revised Scale (Kelsey et al. 2004)

Late afternoon/evening items (evening subscore)

Problems with homework/tasks

Difficulty sitting through dinner

Difficulty playing quietly in the afternoon/evening

Inattentive and distractable in the afternoon/evening

Difficulty transitioning

Arguing or struggling in the afternoon/evening

Difficulty settling at bedtime

Difficulty falling asleep (subscore)

Early morning items (morning subscore)

Difficulty getting out of bed

Difficulty getting ready

Arguing or struggling in the morning

Tolerability assessments

Tolerability was assessed by open-ended questioning for adverse events (AEs) and the assessment of vital signs and body weight.

Statistical analyses

Details of the primary analysis have been published elsewhere (Wehmeier et al. in press). As there were no data available from placebo-controlled trials using this cb-CPT when this study was planned, the sample size calculation was based on the following generic approach: For other endpoints (e.g., the ADHD-RS) effect sizes (ESs) of at least 0.5 were assumed for treatment with ATX. Patients were randomized in a 1:1 ratio between ATX and placebo treatment using a computer-generated randomization scheme via a centralized telephone-based system. For a two-sided t-test with level 5%, 128 patients overall were needed to achieve a power of 80% to detect an ES of 0.5. The random allocation sequence was generated by a group independent from the study team. Patients were enrolled by the investigators, who used the telephone randomization system for assigning patients to interventions. Investigators, patients, and study team were blinded until the data were locked.

The focus of this article is on the effects of ATX on the variety of neuropsychological outcome parameters across the day. These were specified a priori as secondary analyses. The q-scores of the cb-CPT were analyzed using a repeated measures analysis based on the restricted maximum likelihood method assuming an unstructured covariance matrix (mixed model repeated measures [MMRM]). A covariance matrix using fewer parameters (first choice was autoregressive type) was prespecified in case of nonconvergence. The independent fixed effects in the model were treatment, visit, baseline value, baseline value-by-visit interaction, treatment-by-visit interaction, time of day (categorized as morning, noon, evening), and daytime-treatment-visit interaction. ESs comparing the treatment groups at week 8 with the respective p-values were reported. Analysis of covariance (ANCOVA) was used to perform sensitivity analyses based on last observation carried forward (LOCF) and per protocol analyses were done to support the primary analyses. All efficacy and tolerability analyses were conducted on the full analysis set (FAS), which included all randomized patients who took at least one dose of study medication (ATX or placebo). Secondary measures (WREMB-R, ADHD-RS, and CGI-S-ADHD) were analyzed using the same MMRM approach as described above. AE rates were evaluated descriptively for each treatment arm. All statistical tests were done using a two-sided level of 5%. Ninety-five percent confidence intervals are reported. No adjustments for multiple comparisons were done as these were secondary analyses. (See Wehmeier et al. (in press) regarding the multiple comparisons adjustment for the primary analyses.) The data analysis for this article was performed using SAS software version 8.2 or higher (Copyright, SAS Institute Inc.; SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC).

Results

Patient disposition

The first patient entered the study on October 26, 2007, and the last patient completed the study on May 26, 2009. Of 135 patients initially assessed, 128 patients were randomly assigned to treatment; 125 patients received at least one dose of study drug (ATX n=63, placebo n=62; FAS). Of these patients, 105 (84.0%) patients completed the study (Fig. 1). The most common reasons for early discontinuation were lack of efficacy (9.6%), followed by AEs (4.0%), patient decision (1.6%), and physician decision (0.8%). The retention rate was slightly lower in the placebo group (82.3%) than in the ATX group (85.7%).

FIG. 1.

Patient disposition.

Baseline characteristics

In the overall sample the mean (SD) age was 9.0 (1.79) years and 77.6% of the patients were male. The two treatment groups were comparable in terms of baseline characteristics and baseline ADHD severity as measured by the ADHD-RS (Table 2).

Table 2.

Baseline Characteristics

	Atomoxetine, n=63	Placebo, n=62
Demographics
Age (years), mean, SD	9.1 (1.93)	8.9 (1.64)
Age ≥12 years, n (%)	4 (6.3)	4 (6.5)
Gender: Male, n (%)	47 (75.6)	50 (80.6)
BMI (kg/m²), mean (SD)	17.7 (2.83)	17.9 (2.91)
ADHD subtype, n (%)
Combined	40 (63.5)	48 (77.4)
Predominantly inattentive	17 (27.0)	11 (17.7)
Predominantly hyperactive-impulsive	6 (9.5)	3 (4.8)
Psychiatric co-morbidity, n (%)
At least one psychiatric co-morbidity	25 (39.7)	25 (40.3)
Oppositional defiant disorder	20 (31.7)	19 (30.6)
Conduct disorder	9 (14.3)	12 (19.4)
ODD or CD	25 (39.7)	25 (40.3)
Tic disorder	1 (1.6)	0 (0.0)
Mood disorder	0 (0.0)	1 (1.6)
Previous treatment
Previous stimulant exposure, n (%)	13 (20.6)	18 (29.0)
Symptom severity, mean score (SD)
ADHD-RS total score	37.30 (10.62)	36.68 (12.53)
ADHD-RS, inattention subscore	19.98 (5.22)	19.13 (6.05)
ADHD-RS, hyperactivity/impulsivity subscore	17.32 (6.77)	17.55 (7.51)
CGI-S ADHD	5.11 (1.02)	5.05 (1.11)
WREMB-R total score	21.70 (7.64)	21.58 (7.91)
WREMB-R evening subscore	14.54 (4.76)	15.39 (4.74)
WREMB-R morning subscore	5.56 (2.66)	4.63 (2.92)
WREMB-R difficulty falling asleep	1.60 (1.23)	1.56 (1.26)

ADHD=attention-deficit/hyperactivity disorder; ADHD-RS=ADHD Rating Scale; BMI=body mass index; CD=conduct disorder; CGI-S=Clinical Global Impression-Severity; ODD=oppositional defiant disorder; SD=standard deviation; WREMB-R=Weekly Ratings of Morning and Evening Behavior-Revised.

Efficacy outcome: cb-CPT scores

The primary efficacy analyses (MMRM) showed that treatment with ATX once daily (target dose 1.2 mg/kg/day) over 8 weeks was significantly superior to placebo in reducing hyperactivity, inattention, and impulsivity as measured by q-scores of 10 primary variables of the cb-CPT (Table 3). The respective MMRM analyses based on the per-protocol population (excluding tests where the longest period “off task” exceeded 45 seconds) and the supportive ANCOVA analysis using LOCF (data not shown) yielded corresponding results.

Table 3.

Effect Sizes and Respective p-values at the Three Time Points Over the Day for the Computer-Based Continuous Performance Test q-Scores Based on Mixed Model Repeated Measures at Week 8

	Morning		Noon		Evening		Day-time-treatment-visit interaction ^a
q-score	ES	p-value	ES	p-value	ES	p-value	p-value
Hyperactive
Time active	0.59	<0.001	0.86	<0.001	0.64	<0.001	0.389
Distance	0.85	<0.001	1.10	<0.001	0.75	<0.001	0.693
Area	1.09	<0.001	1.31	<0.001	0.84	<0.001	0.677
Microevents	0.93	<0.001	1.19	<0.001	0.89	<0.001	0.799
Motion simplicity	0.47	0.008	0.37	0.035	0.32	0.079	0.613
Inattention
RT variation	0.72	<0.001	0.70	<0.001	0.72	0.001	1.000
Omission error rate	0.74	<0.001	0.66	0.001	0.70	0.001	0.995
Reaction time	0.42	0.043	0.31	0.136	0.49	0.022	0.998
Normalized RT variation	0.47	0.019	0.60	0.003	0.43	0.042	0.954
Impulsivity
Commission error rate	0.61	<0.001	0.50	0.002	0.37	0.023	0.944
Anticipatory^b	0.10	0.416	0.21	0.099	0.20	0.120	0.480
Other
Error rate^b	1.00	<0.001	0.91	<0.001	0.91	<0.001	0.985
Multiresponse^b	−0.03	0.858	0.24	0.108	0.14	0.360	0.307

Positive effect sizes point in favor of atomoxetine.

p-values are based on treatment group comparisons at that day time for the week 8 visit.

A significant interaction test would provide evidence that the treatment effect varies with the day-time.

These endpoints were not primary endpoints.

ES=effect size; RT, reaction time.

None of the MMRM analyses revealed a significant interaction between treatment, visit, and the time of the day (Table 3). The analyses for differences between the treatment groups within the three times of the day revealed significant differences between the treatment groups for most of the endpoints for most of the times of the day.

The analyses of q-scores of the cb-CPT using general additive models (GAMs) allow more detailed insight into the variation of symptoms over the day and differences between the treatment groups over the day (Fig. 2). At baseline, the courses across the day for the five parameters reflecting hyperactivity (TA, DIS, AR, MEs, MS) appear very similar. Symptoms appear to be more or less severe over the entire day, with very little variation. After 8 weeks of treatment, however, the two groups clearly separate. The curves for the ATX group consistently reflect better values over the day than the curves for the placebo group. Interestingly, the shape of the curves changes over time, indicating higher variability of symptoms over the course of the day. Hyperactivity increases in the morning, decreases during the early afternoon, increases again in the late afternoon, although to a lesser degree, and finally decreases in the early evening.

FIG. 2.

Computer-based continuous performance test q-scores over the day, baseline and week 8 results based on general additive models analyses.

A similar pattern was observed for the OE rate and the nVRT, which both indicate inattention. The other two variables that reflect inattention, RTV and mRT, show different patterns. RTV shows considerable variation over the day. nVRT decreases during the morning and increases again, with a clear peak at the early afternoon. At baseline, the ATX group showed worse values than the placebo group over the day. At week 8, the shapes of the curves are similar to the shapes seen for the variables that reflect hyperactivity. At week 8, the ATX group shows better values over the day than the placebo group. The profile of the RT within the day was similar to that of the RTV at baseline. At week 8, the curves for both groups were flat over the entire day.

The CE rate indicates impulsivity. The profile within the day also decreases during the morning and increases toward noon, with a peak in the early afternoon. Little variation within the day can be seen after 8 weeks when the ATX group consistently showed better values compared with the placebo group.

A number of graphs show a peculiar isolated peak. Such peaks should be disregarded as they are the result of outliers both on the time and the q-score axes. As a result smoothing by splines is unstable in these time windows of the day where observations were sparse.

Efficacy outcomes: Weekly Ratings of Morning and Evening Behavior-Revised, Attention-Deficit/Hyperactivity Disorder-Rating Scale, and Clinical Global Impression-Severity-Attention-Deficit/Hyperactivity Disorder

The WREMB-R total score and the subscores also showed statistically significant differences between the treatment groups at week 8 (Table 4). ATX was significantly superior to placebo in reducing ADHD symptom severity as measured by ADHD-RS and CGI-S scores. Sensitivity tests using ANCOVA supported these results (data not shown).

Table 4.

Changes of Attention-Deficit/Hyperactivity Disorder-Rating Scale (ADHD-RS), Weekly Ratings of Morning and Evening Behavior-Revised, and Clinical Global Impressions-Severity-(C&I-S) Attention-Deficit/Hyperactivity Disorder; Mixed Model Repeated Measures Analyses at Week 8

	LS-mean difference	95% CI	Effect size	p-value
ADHD-RS
Total score	11.60	8.20 to 14.99	1.30	<0.001
Inattention subscore	5.12	3.30 to 6.94	1.07	<0.001
Hyperactivity/impulsivity subscore	6.55	4.74 to 8.35	1.37	<0.001
CGI-S ADHD	1.11	0.76 to 1.46	1.11	<0.001
WREMB-R
Total score	5.74	3.55 to 7.92	1.00	<0.001
Morning subscore	1.18	0.42 to 1.93	0.59	0.002
Late noon and evening subscore	3.96	2.49 to 5.44	1.02	<0.001
Difficulty falling asleep	0.62	0.31 to 0.93	0.62	<0.001

CI=confidence interval; LS=least square; MMRM=mixed model repeated measures.

Tolerability

Patients received study treatment for a median of 55 days in both groups, with quartiles of 53 and 55 days in the ATX group and quartiles of 52 and 55 days in the placebo group. Treatment emergent AEs (TEAEs) were reported in 32 (50.8%) patients treated with ATX and in 27 (43.5%) patients receiving placebo. Drug-related TEAEs were observed in 14 (22.2%) and 7 (11.3%) patients, respectively. Two patients treated with ATX discontinued the study due to TEAEs (3.2%), in the placebo group, three patients discontinued due to TEAEs (4.8%). The most commonly reported TEAEs (>5%) were abdominal pain (ATX: 11.1%; placebo: 3.2%), nausea (ATX: 9.5%; placebo: 3.2%), fatigue (ATX: 6.3%; placebo: 1.6%), upper respiratory tract infection (ATX: 6.3%; placebo: 0.0%), pharyngolarygeal pain (ATX: 6.3%; placebo: 0.0%), headache (ATX: 4.8%; placebo: 8.1%), and aggression (ATX: 0.0%; placebo: 6.5%). No serious AEs were reported. More detailed tolerability results have been published elsewhere.

Discussion

The baseline characteristics of the patients in this sample (Table 2) are comparable to those of other placebo-controlled ATX studies (Cheng et al. 2007). While previous studies have largely relied on Clinical Rating Scales to measure improvement of ADHD symptoms, this was the first randomized, placebo-controlled study using a cb-CPT combined with an infra-red motion-tracking device (QbTest) as an objective approach to measure parameters of executive function, inhibitory control, and motor activity in children on ATX (Wehmeier et al. in press). These functions were assessed by means of 13 specific variables that have been used to objectively measure response to stimulants both in routine clinical practice and in one-armed studies (Teicher et al. 2003, 2004, 2008; Heiser et al. 2004; Tabori-Kraft et al. in press). The repeated assessment of these 13 neuropsychological variables allowed the identification of patterns of executive function, inhibitory control, and motor activity across the day.

Both groups of patients showed a circadian pattern of neuropsychological outcomes across the day. Neuropsychological outcomes as reflected by the 13 parameters of the QbTest at week 8 generally showed an improvement of performance in the morning, with a presumable performance peak at around 10 am. The peak is followed by a decline in performance that reaches a trough at around 2 pm. After this point in time, performance improved again, reaching a second peak at around 5 pm. Performance then declined toward the evening hours and presumably into the night and until early the following morning. These findings are in line with the finding that the degree to which the various times of the day are found to be challenging fluctuates over the day (Coghill et al. 2008). It would be interesting to speculate whether these fluctuations are a result of circadian peaks and troughs in executive performance.

Remarkably, in our study all primary efficacy outcome parameters (i.e., q-scores of the 10 standard variables of the cb-CPT) showed statistically significant superiority of ATX treatment over placebo. Further, three additional (secondary) cb-CPT outcome parameters showed this difference too. Thus, all 13 variables captured by the cb-CPT showed statistically significant superiority of ATX treatment over placebo.

The ESs for the 13 cb-CPT parameters, however, varied considerably. The highest ES were seen for parameters relating to hyperactivity, followed by those related to inattention and impulsivity. Interestingly, these findings are also reflected in the ADHD-RS subscores. The ES of 1.30 seen in the ADHD-RS total score is high. The same high ES was also seen in another placebo-controlled study (Svanborg et al. 2009). This is remarkable because several other placebo-controlled studies with ATX have shown lower ES (Faraone 2009). One reason for the large ES in the present study might be the close oversight of the patient by the physician (three times a day on visit days), possibly resulting in more precise clinical rating. This may have resulted in better differentiation between nonresponders in the placebo group and responders in the ATX group. This effect might have been reinforced by information available to the physician through the cb-CPT report. Finally, the high ES may be a result of a relatively small placebo effect. As shown in another study, patients with inattentive ADHD subtype and medication-naïve patients may have higher placebo response rates (Newcorn et al. 2009).

The MMRM results of the cb-CPT showed no indication of a differential treatment effect over the day. This supports the notion of sustained efficacy over 24 hours because the morning scores before administration of that day's medication are not relevantly lower than the noon and evening scores after administration of medication. Thus, continuous efficacy appears to be present over the entire day, including the following morning.

The analyses of the q-scores of the cb-CPT using GAMs rather than MMRM give a more detailed insight into the variation of the symptoms over the day and the difference between the treatment groups over the day. The variables characterizing the hyperactive symptoms (i.e., TA, DIS, AR, MEs, and MS) show similar graphs across the day at baseline with considerable symptom severity being present over the entire day. This finding suggests that the patients in this study were more hyperactive than children and adolescents without ADHD as measured by the infra-red motion-tracking device. The constantly high scores over the day suggest very low fluctuation of hyperactivity over the day. Ceiling effects could also explain this finding.

After 8 weeks, the two groups clearly separated. The curve of the ATX group has consistently better results over the entire day. Interestingly, also the shape of the curves changes showing higher intra-day variability. The hyperactivity rises in the morning, decreases after lunch, increases again to a lesser degree in the late afternoon, and finally lessens in the early evening. A similar characteristic was observed for OE rate and nVRT, both reflecting inattention. Based on face-validity, the other two variables characterizing inattention, Reaction time variation and Reaction time, result in different curves. The variation of Reaction time shows considerable variation within the day. The variation of the Reaction time decreases during the morning and increases again, with a subsequent peak in the early afternoon. At baseline, the ATX group showed worse results than the placebo group over the day. At week 8, the shape of the curves is similar to the shape seen for the variables reflecting hyperactivity. The ATX group showed better results over the whole day than the placebo group. The profile of Reaction time within the day was similar to that of the Variation of the reaction time at baseline. At week 8, the curves for both groups were flat over the entire day.

The CE rate serves as an indicator for impulsivity. Its profile within day also showed a decrease during the morning, only to increase again later on, leading to a peak at the early afternoon. Little variation within the day can be seen after 8 weeks. At this point in time, the ATX group consistently showed better results than the placebo group.

Further, all primary cb-CPT variables showed significant treatment effects at week 8 for the morning evaluation. This supports findings that ATX is still efficacious after 24 hours, as the cb-CPT was performed after approximately 24 hours of the last dose of ATX and before the next dose. This is the first time that this result was found using an objective measure evaluating the efficacy at a specific time point. All previous findings were based on questionnaires.

The Weekly Ratings of Morning and Evening Behavior-Revised-Investigator Rated (WREMB-R-Inv) Scale was used to assess ADHD-related behavioral problems in the morning, late noon, and evening, as well as difficulties falling asleep (Kelsey et al. 2004; Wehmeier et al. 2009). Findings from this questionnaire suggest that morning and evening behavior, as reflected by the WREMB-R improve over time with once daily ATX treatment. This effect persisted over the observation period of 8 weeks. This finding is consistent with the findings from other studies that used the WREMB-R to assess evening and morning behavior in children and adolescents on ATX (Kelsey et al. 2004, 2007; Wehmeier et al. 2009). This is an important finding because children with ADHD have considerably greater problems in the mornings and the evenings than children without ADHD (Coghill et al. 2008).

ADHD symptom severity as measured by the ADHD-RS and CGI, and WREM Scales was also reduced to a statistically significantly greater degree in the ATX group compared with the placebo group. Thus, the Clinical Rating Scale scores (ADHD-RS, CGI-S) corroborate the results of the cb-CPT.

ATX was generally well tolerated. No serious AEs were reported at any time during the study. The tolerability profile of ATX was seen to be consistent with the findings from previous studies, and reflects the known tolerability profile of ATX.

The limitations of this study include the relatively short duration of the observation period. Further, the cb-CPT used in this study is still far from being a standard tool for assessing symptom severity and treatment response in clinical practice. The relationship between clinical symptoms of ADHD and objectively measured CPT results is still unclear (Crosbie and Schacher 2001; Nigg et al. 2005). Therefore, the concordance and discordance between the standard Clinical Rating Scale ADHD-RS and the cb-CPT used in this particular study is currently under investigation and will be the topic of another publication.

The high number of variables observed both using the cb-CPT and the questionnaires with its total scores and subscores may lead to problems with multiple testing, that is, false-positive findings. However, the high consistency of findings based on the various variables suggests that significant p-values are based on real differences and not observed by pure chance alone.

A further limitation could be the close oversight of the patient by the physician (three times a day on visit days), which may have had an influence on Clinical Rating Scale scores. This may have resulted in better differentiation between nonresponders in the placebo group and responders in the ATX group. This effect might have been reinforced by information available to the physician through the cb-CPT report as the investigators were not blinded to the cb-CPT results. However, not blinding the investigators to the cb-CPT results more closely reflects clinical settings where this test is used as opposed to a more experimental approach with investigators being blinded toward the cb-CPT results.

The combination of a cb-CPT with a motion-tracking device may result in the development of enhanced techniques to predict treatment response and facilitate the clinical management of patients on ADHD medication. Objective measurements of ADHD-related symptoms with techniques such as the one used in this study may play an important role in the future. The cb-CPT used in this study can be considered a biological marker according to the definition of the Food and Drug Administration, as it objectively measures a biological process (neuropsychological function) and evaluates pharmacological response (Biomarkers Definitions Working Group 2001).

In summary, this study demonstrated that ATX reduces ADHD symptom severity as measured by various Clinical Rating Scales. Most importantly, this study clearly showed the positive effect of ATX on executive function, inhibitory control, and hyperactivity as measured by the particular cb-CPT and motion-tracking device used here. Further, the long duration of ATX was shown using an objective method to measure the ADHD core symptom efficacy of ATX over the day.

Footnotes

Acknowledgments

First and foremost we would like to thank all patients and their parents for participating in this study. We would also like to thank the investigators for their contribution to making this study successful: Dr. Alfred (Munich), Dr. Behre (Kehl), Dr. Busse (Berlin), Dr. Dieffenbach (Datteln), Dr. Fischbach (Solingen), Dr. Geraets (Düsseldorf), Dr. Kühle (Giessen), Prof. Lehmkuhl (Cologne), Dr. Meyers (Dorsten), Dr. Niemeyer (Hannover), Dr. Peters-Pasztor (Eberswalde), Dr. Schiekirka (Wolfenbüttel), and Prof. Schulz (Freiburg). Finally, we would like to thank Fredrik Ulberstad and Petter Knagenhjelm (Qbtech, Gothenburg, Sweden) for their invaluable technical support in planning and carrying out the study as well as interpreting the QbTest results.

Disclosures

The study was funded by Lilly Deutschland, the German affiliate of Eli Lilly and Company. Data were analyzed by PSI, St. Petersburg, Russia, under the oversight of Alexander Schacht. The article was drafted by Peter M. Wehmeier (P.M.W) and Alexander Schacht. A.S. is a full-time employee of Lilly Deutschland. Christian Wolff, Walter R. Otto, Ralf W. Dittmann (R.W.D), and Tobias Banaschewski (T.B.) made important contributions to the article. All authors reviewed the final draft and provided important intellectual content. R.W.D is a former employee of Eli Lilly & Co. and now holds the Eli Lilly Endowed Chair of Pediatric Psychopharmacology at the Central Institute of Mental Health Mannheim, University of Heidelberg, Germany. R.W.D. was involved in clinical trials conducted by Lilly and is a member of a Lilly Advisory Board. T.B. has served as an advisor or consultant for Desitin, Lilly, Medice, Novartis, Pfizer, Shire, UCB, and Viforpharma. He received conference attendance support and conference support or received speaker's fee by Lilly, Janssen, McNeil, Medice, Novartis, Shire, and UCB. He is involved in clinical trials conducted by Lilly, Shire, and a study on ADHD care management conducted by Novartis. A.S. and R.W.D. own Eli Lilly & Co. stock.

References

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th Text Revision (DSM-IV-TR) Washington, DC: American Psychiatric Association, 2000.

Arnsten

. Toward a new understanding of attention-deficit hyperactivity disorder pathophysiology: An important role for prefrontal cortex dysfunction. CNS Drugs, 23,Suppl 1:33–41. 2009.

Banaschewski

, Coghill

, Santosh

, Zuddas

, Asherton

, Buitelaar

, Danckaerts

, Döpfner

, Faraone

, Rothenberger

, Sergeant

, Steinhausen

, Sonuga-Barke

EJS

, Taylor

. Long-acting medications for the hyperkinetic disorders. A systematic review and European treatment guidelines. Eur Child Adolesc Psychiatry, 15:476–495. 2006.

Banaschewski

, Roessner

, Dittmann

, Santosh

, Rothenberger

. Non-stimulant medications in the treatment of ADHD. A systematic review and European treatment guidelines. Eur Child Adolesc Psychiatry, 13,Suppl 1:I102–I116. 2004.

Barkley, RA:. The ecological validity of laboratory and analogue assessment methods of ADHD symptoms. J Abnorm Child Psychol, 19:149–178. 1991.

Biederman

, Faraone

. Attention-deficit hyperactivity disorder. Lancet, 366:237–248. 2005.

Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints:. Preferred definitions and conceptual framework. Clin Pharm Ther, 69:89–95. 2001.

Carlson

, Dunn

, Kelsey

, Ruff

, Ball

, Ahrbecker

, Allen

. A pilot study for augmenting atomoxetine with methylphenidate: Safety of concomitant therapy in children with attention-deficit/hyperactivity disorder. Child Adolesc Psychiatry Ment Health, 1:10. 2007.

Castellanos

, Sonuga-Barke

, Scheres

, Di Martino

, Hyde

, Walters

. Varieties of attention-deficit/hyperactivity disorder-related intra-individual variability. Biol Psychiatry, 57:1416–1423. 2005.

10.

Cheng

, Chen

, Ko

, Ng

. Efficacy and safety of atomoxetine for attention-deficit/hyperactivity disorder in children and adolescents—meta-analysis and meta-regression analysis. Psychopharmacology, 194:197–209. 2007.

11.

Coghill

, Soutullo

, d'Aubuisson

, Preuss

, Lindback

, Silverberg

, Buitelaar

. Impact of ADHD on the patient and family. Results from a European survey. Child Adolesc. Psychiatry Ment Health, 2:31. 2008.

12.

Crosbie

, Schachar

. Deficient inhibition as a marker for familial ADHD. Am J Psychiatry, 158:1884–1890. 2001.

13.

Denckla

. Research on executive function in a neurodevelopmental context: Application of clinical measures. Dev Neurpsychol, 12:5–15. 1996.

14.

di Michele

, Princhep

, John

, Chabot

. The neurophysiology of attention-deficit/hyperactivity disorder. Int J Psychophysiol, 58:81–93. 2005.

15.

Döpfner

, Gerber

, Banaschewski

, Breuer

, Freisleder

, Gerber-von Müller

, Günter

, Hässler

, Ose

, Rothenberger

, Schmeck

, Sinzig

, Stadler

, Uebel

, Lehmkuhl

. Comparative efficacy of once-a-day extended-release methylphenidate, two-times-daily immediate-release methylphenidate, and placebo in a laboratory school setting. Eur Child Adolesc Psychiatry, 13,Suppl 1:I93–I101. 2004.

16.

Döpfner

, Lehmkuhl

. DISYPS-KJ: Diagnostik-System für psychische Störungen im Kindes- und Jugendalter nach ICD-10 und DSM-IV, Second. Bern, Switzerland: Hans Huber, 2000.

17.

Doyle

. Executive functions in attention-deficit/hyperactivity disorder. J Clin Psychiatry, 67,Suppl 8:21–26. 2006.

18.

DuPaul

, Power

, Anastopoulos

, Reid

. ADHD Rating Scale-IV: Checklists, Norms, Clinical Interpretations. NY: The Guilford Press, 1998; 28–42.

19.

Faedda

, Teicher

. Objective measures of activity and attention in the differential diagnosis of psychiatric disorders of childhood. Essent Psychopharmacol, 6:239–249. 2005.

20.

Faraone

. Using meta-analysis to compare the efficacy of medications for attention-deficit/hyperactivity disorder in youths. PT, 34:678–694. 2009.

21.

Faries

, Yalcin

, Harder

, Heiligenstein

. Validation of the ADHD Rating Scale as a clinician administered and scored instrument. J Atten Disord, 5:39–47. 2001.

22.

Guy

. ECDEU Assessment Manual for Psychopharmacology, 338. Washington, DC: US Department of Health, Education and Welfare, 1976; 534–537.

23.

Halperin

, Wolf

, Greenblatt

, Young

. Subtype analysis of commission errors on the continuous performance test. Dev Neuropsychol, 7:207–217. 1991.

24.

Heiser

, Frey

, Smidt

, Sommerlad

, Wehmeier

, Hebebrand

, Remschmidt

. Objective measurement of hyperactivity, impulsivity, and inattention in children with hyperkinetic disorders before and after treatment with methylphenidate. Eur Child Adolesc Psychiatry, 13:100–104. 2004.

25.

Heiser

, Heinzel-Gutenbrunner

, Frey

, Smidt

, Grabarkiewicz

, Friedel

, Kühnau

, Schmidtke

, Remschmidt

, Hebebrand

. Twin study on heritability of activity, attention, and impulsivity as assessed by objective measures. J Atten Disord, 9:575–581. 2006.

26.

Kelsey

, Sumner

, Casat

, Coury

, Quintana

, Saylor

, Sutton

, Gonzales

, Malcolm

, Schuh

, Allen

. Once-daily atomoxetine treatment for children with attention-deficit/hyperactivity disorder, including an assessment of evening and morning behavior: A double-blind, placebo-controlled trial. Pediatrics, 114:e1–e8. 2004.

27.

Kelsey

, Sutton

, Schuh

, Dell'Agnello

, Sumner

. Once-Daily Atomoxetine for ADHD: Update on Evening and Morning Behavior. Presented at the 13th International Congress of the European Society for Child and Adolescent Psychiatry (ESCAP), Florence, Italy, August 25–29, 2007.

28.

Maziade

, Rouleau

, Lee

, Rogers

, Davis

, Dickson

. Atomoxetine and neuropsychological function in children with attention-deficit/hyperactivity disorder: Results of a pilot study. J Child Adolesc Psychopharmacol, 19:709–718. 2009.

29.

Michelson

, Allen

, Busner

, Casat

, Dunn

, Kratochvil

, Newcorn

, Sallee

, Sangal

, Saylor

, West

, Kelsey

, Wernicke

, Trapp

, Harder

. Once-daily atomoxetine treatment for children and adolescents with attention deficit hyperactivity disorder: A randomized, placebo-controlled study. Am J Psychiatry, 159:1896–1901. 2002.

30.

National Institutes of Mental Health (NIMH). Clinical global impressions. Psychopharmacol Bull, 21:839–943. 1985.

31.

Newcorn

, Sutton

, Zhang

, Wilens

, Kratochvil

, Emslie

, D'Souza

, Schuh

, Allen

. Characteristics of placebo responders in pediatric clinical trials of attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry, 48:1165–1172. 2009.

32.

Nigg

, Willcutt

, Doyle

, Sonuga-Barke

. Causal heterogeneity in attention-deficit/hyperactivity disorder: Do we need neuropsychologically impaired subtypes? Biol Psychiatry, 57:1224–1230. 2005.

33.

Paulus

, Geyer

. A temporal and spatial scaling hypothesis for the behavioral effects of psychostimulants. Psychopharmacology, 104:6–16. 1991.

34.

Polanczyk

, Silva de Lima

, Lessa Horta

, Biederman

, Rohde

. The worldwide prevalence of ADHD: A systemtic review and metaregression analysis. Am J Psychiatry, 164:942–948. 2007.

35.

Riccio

, Reynolds

, Lowe

, Moore

. The continuous performance test: A window on the neural substrates for attention? Arch Clin Neuropsychol, 17:235–272. 2002.

36.

Rubia

, Overmeyer

, Taylor

, Brammer

, Williams

, Simmons

, Bullmore

. Hypofrontality in attention deficit hyperactivity disorder during higher-order motor control: A study with functional MRI. Am J Psychiatry, 156:891–896. 1999.

37.

Sonuga-Barke

EJS

, Swanson

, Coghill

, DeCory

, Hatch

. Efficacy of two once daily methylphenidate formulations compared across dose levels at different times of the day. Preliminary indications from a secondary analysis of the COMACS study data. BMC Psychiatry, 4:28. 2004.

38.

Salthouse

. Attentional blocks are not responsible for age-related slowing. J Gerontol, 48:263–270. 1993.

39.

Sutton

, Sumner

, Allen

, Feng

, Schuh

, Michelson

. Validity, reliability, and Responsiveness of the DPREMB-R Scale for ADHD. Scientific Proceedings of the 50th Anniversary Meeting of the American Academy of Child and Adolescent Psychiatry (AACAP), Miami Beach, FL, October 14–19, 2003, 169–170.

40.

Svanborg

, Thernlund

, Gustafsson

, Hägglöf

, Poole

, Kadesjö

. Efficacy and safety of atomoxetine as add-on to psychoeducation in the treatment of attention deficit/hyperactivity disorder: A randomized, double-blind, placebo-controlled study in stimulant-naïve Swedish children and adolescents. Eur Child Adolesc Psychiatry, 18:240–249. 2009.

41.

Swanson

, Wigal

, Sonuga-Barke

, Greenhill

, Biederman

, Kollins

, Nguyen

, DeCory

, Hirshey Dirksen

, Hatch

. the COMACS Study Group: Comparison of once-daily extended release methylphenidate formulations in children with Attention-Deficit/Hyperactivity Disorder in the laboratory school (the COMACS study) Pediatrics, 113:e206–e216. 2004.

42.

Tabori-Kraft

, Sørensen

, Kaergaard

, Dalsgaard

, Thomsen

. Is OPTAx useful for monitoring the effect of stimulants on hyperactivity and inattention? A brief report. Eur Child Adolesc Psychiatry, 16:347–351. 2007.

43.

Tannock

. Attention deficit hyperactivity disorder: Advances in cognitive, neurobiological, genetic research. J Child Psychol Psychiatr, 39:65–99. 1998.

44.

Teicher

. Actigraphy and motion analysis: New tools for psychiatry. Harv Rev Psychiatry, 3:18–35. 1995.

45.

Teicher

, Anderson

, Polcari

, Glod

, Maas

, Renshaw

. Functional deficits in basal ganglia of children with attention-deficit/hyperactivity disorder shown with functional magnetic resonance imaging relaxometry. Nat Med, 6:470–473. 2000.

46.

Teicher

, Ito

, Glod

, Barber

. Objective measurement of hyperactivity and attentional problems in ADHD. J Am Acad Child Adolesc Psychiatry, 35:334–342. 1996.

47.

Teicher

, Lowen

, Polcari

, Foley

, McGreenery

. Novel strategy for the analysis of CPT data provides new insight into the effects of methylphenidate on attentional states in children with ADHD. J Child Adolesc Psychopharmacol, 14:219–232. 2004.

48.

Teicher

, Polcari

, Anderson

, Andersen

, Lowen

, Navalta

. Rate dependency revisited: Understanding the effects of methylphenidate in children with attention deficit hyperactivity disorder. J Child Adolesc Psychopharmacol, 13:41–51. 2003.

49.

Teicher

, Polcari

, McGreenery

. Utility of objective measures of activity and attention in the assessment of therapeutic response to stimulants in children with attention-deficit/hyperactivity disorder. J Child Adolesc Psychopharmacol, 18:265–270. 2008.

50.

Wehmeier

, Dittmann

, Schacht

, Helsberg

, Lehmkuhl

. Morning, evening behavior in children, adolescents treated with atomoxetine once daily for attention-deficit/hyperactivity disorder (ADHD): Findings from two 24-week, open-label studies. Child Adolesc Psychiatry Ment Health, 3:5. 2009.

51.

Wehmeier

, Schacht

, Ulberstad

, Lehmann

, Schneider-Fresenius

, Lehmkuhl

, Dittmann

, Banaschewski

. Does atomoxetine Improve executive function, inhibitory control, and hyperactivity? Results from a placebo-controlled trial using quantitative measurement technology. J Clin Psychopharmacol in press.

52.

Wolraich

, Greenhill

, Pelham

, Swanson

, Wilens

, Palumbo

, Atkins

, McBurnett

, Bukstein

, August

. Randomized, controlled trial of OROS methylphenidate once a day in children with attention-deficit/hyperactivity disorder. Pediatrics, 104:883–892. 2001.