August 2009  << Back  

  Can't open the PDF? Click here for help.

 

 RESEARCH ARTICLE

Effects of open-label atomoxetine on African-American and Caucasian pediatric outpatients with attention-deficit/hyperactivity disorder

Todd M. Durell, MD

Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN, USA

Andres J. Pumariega, MD

Department of Psychiatry, The Reading Hospital and Medical Center, Reading, PA, USA

Eugenio M. Rothe, MD

Robert Stempel School of Public Health, Florida International University, Miami, FL, USA

Jorge M. Tamayo, MD, BMSS

Department of Psychiatry, CES University, Medellín, Colombia, Department of Psychiatry, University of Puerto Rico, San Juan, PR

David Baron, MSEd, DO

Professor and Chair, Department of Psychiatry, Temple University School of Medicine, Philadelphia, PA, USA

David Williams, MS

Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN, USA

BACKGROUND: Data on the management of attention-deficit/hyperactivity disorder (ADHD) in African-American children and adolescents are limited.

METHODS: This study sought to evaluate the tolerability, safety, and efficacy of atomoxetine hydrochloride in the management of ADHD in African-American children and adolescents by conducting a post hoc subgroup analysis of 2 multicenter, open-label studies.

RESULTS: Atomoxetine was safe and well tolerated, with ≤3.0% of African-Americans and Caucasians discontinuing treatment because of adverse events. A significantly higher proportion of Caucasians reported ≥1 treatment-emergent adverse event, including vomiting (7.2% vs 1.2%; P=.037) and fatigue (6.1% vs 0%; P=.012). No serious safety concerns were observed. Changes from baseline in height, weight, and hemodynamic variables were modest and similar in both racial subgroups. African-Americans and Caucasians showed significant improvement from baseline to end point in the mean ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored (ADHDRS-IV-P:I). Scores decreased by 20.1 in African-Americans and by 19.55 in Caucasians, without significant between-group differences. Patients in both racial groups experienced similar, significant improvements in ADHDRS-IV-P:I inattention and hyperactivity-impulsivity symptoms, Clinical Global Impression-ADHD-Severity, and Conners’ Parent Rating Scale-Revised: Short Form.

CONCLUSIONS: Atomoxetine exhibited similar tolerability, safety, and efficacy profiles in African-Americans and Caucasians with ADHD.

KEYWORDS: attention-deficit/hyperactivity disorder, atomoxetine, African-American, Caucasian, race

ANNALS OF CLINICAL PSYCHIATRY 2009;21(1):26–37

  INTRODUCTION

Attention-deficit/hyperactivity disorder (ADHD) is a chronic condition with childhood onset that affects 3% to 8% of children and adolescents.1,2 By one estimate, approximately 2.5 million US children age 4 to 17 reported having a diagnosis of ADHD or were taking medications for ADHD in 2003.1

Approximately equivalent proportions of African-American and Caucasian children and adolescents have ADHD. Prevalence of a reported ADHD diagnosis among US children and adolescents age 4 to 17 was 8.6% among Caucasians and 7.7% among African-Americans in 2003.1 In one study of African-American and Latino persons living in public housing in Los Angeles, CA, ADHD was one of the most common chronic health problems, affecting 15% of the Latino and African-American children in these households.3

Core symptoms of ADHD include inattention, hyperactivity, and impulsiveness, which are frequently accompanied by behaviors that are developmentally inappropriate for the patient’s age.2,4 ADHD also is frequently accompanied by conduct disorders, oppositional defiant disorder (ODD), mood disorders, anxiety disorders, borderline personality disorder, learning disorders, and/or Tourette’s syndrome.5,6 As a result of its core symptoms and/or these comorbidities, ADHD often compromises social functioning, academic performance, and self-esteem in childhood and/or later in life.7-10

Although the prevalence of ADHD is similar for African-Americans and Caucasians, the toll the disorder takes on African-American families may be greater in part because African-Americans are less likely to be evaluated for (and diagnosed with) ADHD and, once diagnosed, African-Americans may also be less likely to begin treatment.11,12 In 2003, 5.0% of Caucasian children and adolescents with ADHD were taking medication, compared with 3.7% of African-Americans of similar ages.1 In the 1997 Medical Expenditure Panel Survey, Caucasian children age 3 to 18 with ADHD were more than twice as likely to be treated for their ADHD as their African-American (and Latino) counterparts.13

Cultural attitudes may underlie some of these racial disparities.14 African-American parents report more negative expectations about ADHD treatment than their Caucasian counterparts.11 Among certain inner-city African-American families, diagnosis and treatment of ADHD may be hindered by mistrust toward the school system, a belief that Caucasian educators do not adequately appreciate African-American culture, a belief that ADHD carries a social stigma, as well as concerns that treatment with stimulants will result in later drug abuse problems.15 African-American parents also report significantly less knowledge about ADHD than do Caucasian parents,16,17 and are less likely to describe symptoms as a medical problem.18 In addition, many African-American parents feel that their children are overdiagnosed with ADHD compared with their Caucasian counterparts and that teachers blame ADHD for their children’s learning or behavioral problems significantly more often than they do with Caucasians.16

In a recent survey of parents’ attitudes toward stimulant medication, many African-American caregivers expressed the belief that stimulant medication leads to substance abuse (17%), that ADHD medications have detrimental side effects (21%), and that more children are medicated for ADHD than is necessary (22%), with many preferring counseling to medications (21%).19 Despite these attitudes, most parents surveyed reported that they trusted recommendations made by physicians as to the effectiveness and safety of ADHD medications.19

Mainstays in the management of ADHD include the psychostimulants methylphenidate and amphetamine, as well as the nonstimulant atomoxetine hydrochloride. Atomoxetine has been demonstrated to be effective, safe, and well tolerated in children and adolescents20-22 and adults23,24 with ADHD. However, data on the use of atomoxetine in the management of ADHD in African-American children and adolescents are limited. To date, we know of only one previous analysis of atomoxetine therapy in the African-American population.25 This trial demonstrated efficacy with atomoxetine but greater improvement with methylphenidate OROS (osmotic release oral system). However, this study was limited to a 3-week treatment period.25

To evaluate the safety, tolerability, and efficacy of atomoxetine when administered in African-American compared with Caucasian outpatients, we conducted a post hoc subgroup analysis of 2 Lilly-sponsored studies (internal study identifiers B4Z-MC-LYAB and B4Z-MCLYBB). These studies were virtually identical in design, purpose, and eligibility criteria. Both were phase III multicenter, open-label studies designed to assess the safety, tolerability, and efficacy of atomoxetine when administered for up to 11 weeks at daily doses of up to 1.8 mg/kg/d or 120 mg/d (whichever was lower). The primary objective of the analysis was to assess the tolerability and safety of atomoxetine in African-American children and adolescents with ADHD. The secondary objective was to assess atomoxetine’s efficacy in this population via the ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored (ADHDRS-IV-P:I).26 We hypothesized that African-American children and adolescents with ADHD who received atomoxetine treatment for up to 11 weeks would exhibit no significant differences compared with Caucasian children and adolescents in tolerability, safety, or efficacy profiles at therapeutic doses. A similar analysis was previously conducted in Latino ADHD patients.27

  METHODS

Patients

In total, 1173 African-American and Caucasian outpatients were enrolled in the B4Z-MC-LYAB and B4Z-MCLYBB studies, which were conducted from March 2000 to December 2001 across 95 sites in the United States, Canada, and Puerto Rico. Eligible for inclusion were outpatients age 6 to <18 with a diagnosis of ADHD as defined by criteria in the DSM-IV-TR.4 This diagnosis was corroborated using the Kiddie Schedule for Affective Disorders and Schizophrenia, Present and Lifetime Version, Behavioral Disorders Supplement (K-SADSPL:Behavioral).28 Other eligibility criteria were a score on the ADHDRS-IV-P:I that was ≥1.0 standard deviation (SD) above age and gender norms and normal intelligence (IQ ≥80). Patients with a history of bipolar disorder or alcohol or drug abuse were excluded.

Before performing study procedures or administering the study drug, informed consent was obtained from each subject’s parent or legal guardian. In addition, children or adolescents whom investigators deemed capable of reading and understanding a written assent form gave written assent before participating (unless prohibited by the ethical review board or applicable law). The study protocol, informed consent document, and assent forms were reviewed and approved by local ethical review boards.

Treatment

After an 8- to 42-day screening and drug-washout phase (8 to 28 days in study LYBB and 15 to 42 days in study LYAB), patients who met enrollment criteria and signed an informed consent/child assent form received open-label treatment with atomoxetine 0.5 to 1.8 mg/kg/d (35 to 120 mg/d) for up to 11 weeks. Doses could be adjusted according to scores on the Clinical Global Impressions-ADHD-Severity (CGI-ADHD-S) instrument. Patients with a score ≥2 were eligible for increases in their atomoxetine dose. Atomoxetine was administered on an open-label basis twice daily, once before and once after school.

Assessments and Outcomes

Tolerability. At each weekly visit, any treatment-emergent adverse events were recorded. These were elicited through open-ended questioning. An adverse event was defined as any untoward medical occurrence reported by a patient after informed consent has been obtained, without regard to the possibility of a causal relationship with treatment. Serious adverse events were defined as events that were fatal or life-threatening, resulted in initial or prolonged hospitalization, or caused severe or permanent disability.

Safety. At baseline and study end point, safety was assessed via laboratory tests; monitoring of vital signs, height, and weight; and ECG. The primary outcome measure was the change from baseline to end point in mean values for these safety parameters in African-American and Caucasian patients. CYP2D6 enzyme function was assessed at baseline, and patients were classified as poor or extensive metabolizers.

Efficacy. The primary efficacy outcome variable was the change from baseline to end point in the ADHDRS-IV-P:I among African-American and Caucasian patients. Secondary efficacy outcome variables included changes from baseline to end point in the ADHDRS-IV-P:I subscales of inattention and hyperactivity/impulsivity; the CGI-ADHD-S, a single-item clinician rating of the severity of ADHD symptoms; and the Conners’ Parent Rating Scale-Revised: Short Form (CPRS-R:S), including subscale scores for Oppositional, Cognitive Problems, Hyperactivity, and ADHD Index.29

Other secondary efficacy outcome measures included the numbers (%) of African-American and Caucasian patients achieving (1) a clinical response, which was defined and analyzed in 2 different ways: as an ADHDRS-IV-P:I total score at end point that was ≤75% of the baseline score, or as a ≥2 point reduction in CGI-ADHD-S from baseline to end point; or (2) remission, which was also defined and analyzed in 2 different ways: as an ADHDRS-IV-P:I total T-score ≤ 60, or as a CGI-ADHD-S score ≤2 at study end point.

Treatment adherence was assessed by patient self-report and defined a priori as taking the prescribed dose on ≥70% of days between visits.

Statistical methods

Continuous data were expressed as mean (± SD) and categorical data as number (%). Statistical tests were 2-tailed at α=.05.

Analyses were conducted on an intent-to-treat basis. Baseline characteristics in African-Americans and Caucasians were compared using Fisher’s exact test for categorical data and 2-sample t tests for continuous data. Numbers of patients (%) discontinuing treatment and treatment adherence rates also were compared between racial subgroups using Fisher’s exact test. Mean modal doses of atomoxetine received by African-Americans and Caucasians were compared using a 2-sample t test.

Numbers (%) of African-American and Caucasian patients with treatment-emergent adverse events and abnormal laboratory values were compared using Fisher’s exact test. Safety analyses included all patients who took ≥1 dose of atomoxetine. Changes from baseline to end point in mean values for vital signs, ECG, height, and weight were analyzed using 2-sample t tests.

The primary efficacy analysis included all patients who had a baseline and ≥1 postbaseline measurement. Comparison of the change from baseline in ADHDRS-IV-P:I total score among African-Americans and Caucasians also was carried out using the last observation carried forward (LOCF) convention and included computation of the mean, SD, and 2-sided 95% confidence interval (CI). Changes from baseline to end point also were compared across racial subgroups using repeated-measures mixed models. Secondary efficacy measures, including changes from baseline to end point in the ADHDRS-IV-P:I inattention and hyperactivity/impulsivity subscale scores, the CPRS-R:S and subscale (Oppositional, Cognitive Problems, Hyperactivity, and ADHD Index) scores, and CGI-ADHD-S scores, were analyzed in a similar manner. For data on questionnaires, such as the ADHDRS-IV-P:I and CPRS-R:S, if >1 item of a subscale was missing, the total score was considered missing. If 1 item was missing, its value was imputed as the mean of the remaining items. The numbers (%) of patients meeting response and remission criteria at end point were compared using Fisher’s exact test, and 2-sided 95% confidence intervals of the odds ratios were computed.

  RESULTS

Patient characteristics and medication usage

For this subanalysis, 83 African-Americans and 1090 Caucasians were included. The mean age of patients in both groups was approximately 11, and ≥76% of patients were male (TABLE 1). A significantly lower proportion of African-Americans (vs Caucasians) had prior stimulant use. African-American patients had significantly higher mean body weight and height, whereas Caucasians had a significantly higher mean pulse rate.

The proportion of poor CYP2D6 metabolizers was lower among African-American (3.6%; 3/83) compared with Caucasian patients (7.5%; 82/1087), but this difference was not statistically significant. No significant differences were observed in mean modal dose of atomoxetine in African-Americans (1.24 [± 0.42] mg/kg/d) and Caucasians (1.27 [± 0.42] mg/kg/d) or in mean maximum doses. The treatment adherence rate was slightly lower in African-American (57.8%) compared with Caucasian (68.7%) patients (P=.051).


TABLE 1

Patient characteristics*

Characteristic African-Americans (n=83) Caucasians (n=1090) P value
Mean (SD) age, years 11.25 (2.80) 11.23 (2.82) .947
Gender no. (%), male 64 (77.11) 831 (76.24) 1.000
Mean (SD) height, cm 149.54 (14.86) 145.08 (17.06) .021
Mean (SD) weight, kg 48.99 (18.83) 41.72 (17.48) .0003
Mean (SD) pulse rate, bpm 77.58 (11.10) 80.59 (11.89) .025
No. (%) with ADHD subtype
  Hyperactivity/Impulsivity
  Inattention
  Combined
3 (3.61)
27 (32.53)
53 (63.86)
32 (2.94)
391 (35.87)
667 (61.19)
.720
No. (%) with previous stimulant use 53 (63.86) 854 (78.49)a .004
No. (%) poor metabolizers (CYP2D6) 3 (3.61) 82 (7.54)b .269
Mean (SD) ADHDRS-IV-P:I total score 39.24 (9.73) 37.75 (10.02) .189
Mean (SD) CPRS-R:S total score 55.59 (15.80) 53.81 (15.04) .301
Mean (SD) CGI-ADHD-S 4.77 (0.65) 4.87 (0.86) .197
*Percents may not add to 100 because of rounding.
aTotal number of patients was 1088.
bTotal number of patients was 1087.
ADHD: attention-deficit/hyperactivity disorder; ADHDRS-IV-P:I: ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored; CGI-ADHD-S: Clinical Global Impressions-ADHDSeverity; CPRS-R:S: Conners’ Parent Rating Scale-Revised: Short Form; CYP2D6: cytochrome P450 2D6 isoenzyme; SD: standard deviation.
Patient disposition

Overall rates of discontinuation were similar among African-American (20.5%) and Caucasian patients (20.7%). Similar proportions of patients also discontinued because of adverse events (2.4% of African-Americans vs 3.0% of Caucasians), perceived lack of efficacy (8.4% vs 10.1%), loss to follow-up (2.4% vs 1.5%), patient decision (4.8% vs 3.2%), physician decision (1.2% vs 0.6%), and other reasons (1.2% vs 2.3%) (FIGURE 1).

Figure 1 Patient disposition

Tolerability and safety

Atomoxetine was well tolerated by patients of both races. A significantly higher proportion of Caucasians (vs African-Americans) reported ≥1 treatment-emergent adverse event (TABLE 2). Headache was the most frequently reported treatment-emergent adverse event in both groups. Other common adverse events included nasopharyngitis, upper-abdominal pain, and decreased appetite. A significantly higher proportion of Caucasians compared with African-Americans reported treatment-emergent vomiting (7.2% vs 1.2%; P=.037) and fatigue (6.1% vs 0%; P=.012).

There were 7 serious adverse events, 1 of which was considered to be related to atomoxetine. This event involved a Caucasian patient who was eventually diagnosed with bipolar disorder. Other serious adverse events included hospitalization for increased ODD in a 15-year-old Caucasian boy (who continued in the study afterward), as well as 1 episode each of compound fracture, pneumonia, accidental injury/trauma, depression (in a patient with a history of depression) among Caucasian youngsters, as well as nonaccidental stab wounds in an 8-year-old Latino boy.

No cases of suicidal behavior were reported in either group. Treatment with atomoxetine was not associated with significant changes from baseline in the mean corrected QT (QTc) interval, body weight, or height (TABLE 3). Both groups exhibited a significant increase in pulse rate with atomoxetine: 6.2 bpm in African-Americans and 6.6 bpm in Caucasians. Caucasians, but not African-Americans, also showed a modest but statistically significant increase in systolic and diastolic blood pressure with atomoxetine: these values increased by 2.4 and 2.1 mmHg, respectively, in Caucasians and by 1.6 and 0.5 mmHg, respectively, in African-Americans. None of the statistically significant differences between African-Americans and Caucasians in hemodynamic parameters were deemed to be clinically significant on review by one of the investigators (T.M.D.), and values for both groups remained within normal ranges generally. A small number of abnormal laboratory values were found within the African-American cohort compared with Caucasians. Upon individual medical review, these mostly modest changes were found, on review by one of the investigators (T.M.D.), to be generally transient and deemed to be of no clinical significance.


TABLE 2

Incidence of treatment-emergent adverse events*

Adverse event African-Americans (n=82) Caucasians (n=1085) P value
Patient with ≥1 AE, n (%) 36 (43.9) 640 (59.0) .010
Headache, n (%) 14 (17.1) 208 (19.2) .770
Nasopharyngitis, n (%) 8 (9.8) 71 (6.5) .254
Upper abdominal pain, n (%) 5 (6.1) 116 (10.7) .258
Decreased appetite, n (%) 5 (6.1) 77 (7.1) 1.000
Somnolence, n (%) 5 (6.1) 44 (4.1) .384
Throat pain, n (%) 4 (4.9) 58 (5.3) 1.000
Diarrhea, n (%) 4 (4.9) 44 (4.1) .770
Cough, n (%) 3 (3.7) 64 (5.9) .620
Upper respiratory tract infection, n (%) 3 (3.7) 53 (4.9) .792
Insomnia, n (%) 3 (3.7) 51 (4.7) 1.000
Irritability, n (%) 3 (3.7) 45 (4.1) 1.000
Nausea, n (%) 2 (2.4) 76 (7.0) .164
Pyrexia, n (%) 2 (2.4) 45 (4.1) .768
Vomiting, n (%) 1 (1.2) 78 (7.2) .037
Nasal congestion, n (%) 1 (1.2) 37 (3.4) .513
Fatigue, n (%) 0 (0.0) 66 (6.1) .012
AE: adverse event.
*Events occurring in ≥3% of patients in either racial subgroup. Only patients who received ≥1 dose of study drug are included.

TABLE 3

Mean changes from baseline in height, weight, and cardiovascular outcomes

Variable African-Americans (n=83) Caucasians (n=1090) P value
Mean (SD) height, cm
  Baseline
  Change to end point
N
82
73

149.54 (14.86)
1.03 (1.97)a
N1081
973

145.08 (17.06)
1.32 (2.65)a

.021
.244
Mean (SD) weight, kg
  Baseline
  Change to end point

83
82

48.99 (18.83)
0.01 (1.97)

1090
1077

41.72 (17.48)
0.32 (2.34)a

.0003
.183
Mean (SD) pulse rate, bpm
  Baseline
  Change to end point

83
82

77.58 (11.09)
6.24 (11.60)a

1088
1079

80.59 (11.88)
6.56 (13.05)a

.025
.829
Mean (SD ) systolic blood pressure, mmHg
  Baseline
  Change to end point

83
82

105.93 (10.19)
1.58 (9.00)

1087
1078

104.19 (11.01)
2.40 (9.92)a

.162
.466
Mean (SD ) diastolic blood pressure, mmHg
  Baseline
  Change to end point

83
82

67.86 (8.41)
0.53 (9.03)

1087
1078

66.30 (8.21)
2.06 (8.80)a

.096
.131
Mean (SD) QTc interval, msec
  Baseline
  Change to end point

82
82

404.76 (16.03)
1.56 (16.14)

1082
1082

405.37 (14.32)
0.13 (14.19)

.711
.386
aWithin-group changes from baseline using paired t tests, P < .001.
QTc: corrected QT interval; SD: standard deviation.
Efficacy

Treatment with atomoxetine in African-American patients resulted in significant reductions (improvements) from baseline to end point in ADHDRS-IV-P:I total score (–20.1; P < .001 vs baseline), inattention symptoms (–10.15; P < .001) and hyperactivity-impulsivity symptoms (–9.95; P < .001) (TABLE 4). These effects were of similar magnitude to changes observed in Caucasian patients (–19.55, –10.71, and –8.84, respectively; all P < .001). Reductions in ADHDRS-IV-P:I total score (FIGURE 2) as well as hyperactivity-impulsivity symptoms (FIGURE 3) were of significantly greater magnitude in Caucasian compared with African-American patients at the first office assessment (visit 4), but not at subsequent visits. Data for inattention symptoms are presented in FIGURE 4.


TABLE 4

Efficacy measures: Mean change from baseline to LOCF end point by racial origin group*

Treatment group N BASELINE CHANGE P value (A)a P value (B)a
Mean SD Mean SD
ADHDRS-IV-P:I total score
  Caucasian
  African-American

1082
82

37.73
39.06

10.03
9.65

–19.55
–20.10

12.67
13.39

<.001
<.001

.943
ADHDRS-IV-P:I Inattention Subscale
  Caucasian
  African-American

1082
82

21.69
21.84

4.53
4.46

–10.71
–10.15

7.29
7.62

<.001
<.001

.433
ADHDRS-IV-P:I Hyperactivity/Impulsivity Subscaleb
  Caucasian
  African-American

1082
82

16.04
17.22

7.61
6.76

–8.84
–9.95

6.95
6.96

<.001
<.001

.432
CGI-ADHD-S
  Caucasian
  African-American

1082
82

4.87
4.76

0.86
0.64

–1.90
–1.77

1.32
1.20

<.001
<.001

.684
CPRS-R:S total scorec
  Caucasian
  African-American

1082
82

53.75
55.48

15.03
15.87

–24.41
–25.84

18.64
20.82

<.001
<.001

.774
CPRS-R:S Oppositionalc
  Caucasian
  African-American

1082
82

9.40
9.26

5.14
5.27

–3.34
–3.50

4.73
4.67

<.001
<.001

.602
CPRS-R:S Cognitive Problemsc
  Caucasian
  African-American

1082
82

14.14
14.67

3.94
3.82

–6.19
–6.55

5.53
5.61

<.001
<.001

.961
CPRS-R:S Hyperactivityc
  Caucasian
  African-American

1082
82

9.74
10.63

5.49
5.26

–5.31
–5.94

4.77
5.44

<.001
<.001

.732
CPRS-R:S ADHD Indexc
  Caucasian
  African-American

1082
82

27.63
28.28

6.26
6.55

–12.75
–12.82

9.38
10.14

<.001
<.001

.773
ADHD: attention-deficit/hyperactivity disorder; ADHDRS-IV-P:I: ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored; ANCOVA: analysis of covariance; CGI-ADHD-S: Clinical global Impressions-ADHD-Severity; CPRS-R:S: Conners’ Parent Rating Scale-Revised: Short Form; LOCF: last observation carried forward; SD: standard deviation.
*Negative values indicate improvement from baseline.
a (A) P values by Student t test; (B) Between-group P value is from baseline to endpoint (LOCF) scores using least-squares means from an ANCOVA model with terms for origin, protocol, and baseline as covariates.
b Sum of the scores on the even-numbered items.
c Oppositional (items 2,6,11,16,20,24), Cognitive Problems (items 3,8,12,17,21,25), Hyperactivity (items 4,9,14,18,22,26), and ADHD Index (items 1,5,7,10,13,15,17,19,21,23,25,27).

In both African-American and Caucasian children and adolescents, treatment with atomoxetine also resulted in significant improvements in CGI-ADHD-S, CPRS-R:S total score and Oppositional, Cognitive Problems, Hyperactivity, and ADHD Index subscales from baseline. There were no racial differences in any of these efficacy measures (TABLE 4).

Response rates were similar in African-American and Caucasian patients, including proportions of patients with reductions in ADHDRS-IV-P:I total scores of ≥25% from baseline (80.5% vs 81.2%, P=.884) and reductions in CGI-ADHD-S scores of ≥2 points (64.6% vs 62.2%; P=.724). Remission rates also were similar between groups, including proportions of patients with an ADHDRS-IV-P: I total T-score ≤60 (34.2% vs 38.0%; P=.555) and a CGI-ADHD-S score of ≤2 (59.8% vs 64.0%; P=.476).

Figure 2 Changes from baseline in ADHDRS-IV-P:I total score over time with atomoxetine*

ADHD: attention-deficit/hyperactivity disorder; ADHDRS-IV-P:I: ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored.
*Reductions indicate improvements.
aStatistically significant (P=.035).

Figure 3 Changes from baseline in ADHDRS-IV-P:I hyperactivity-impulsivity symptoms over time with atomoxetine*

ADHD: attention-deficit/hyperactivity disorder; ADHDRS-IV-P:I: ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored.
*Reductions indicate improvements.
aStatistically significant (P = .035).

Figure 4 Changes from baseline in ADHDRS-IV-P:I inattention symptoms over time with atomoxetine*

ADHD: attention-deficit/hyperactivity disorder; ADHDRS-IV-P:I: ADHD Rating Scale-IV-Parent Version: Investigator Administered and Scored.
*Reductions indicate improvements.

  DISCUSSION

In the first study of its kind to evaluate the effects of atomoxetine on patients with ADHD of different racial origins, we have confirmed our initial hypothesis that African-American children and adolescents with ADHD receiving atomoxetine treatment for up to 11 weeks would have similar responses to treatment compared with their Caucasian counterparts in terms of tolerability, safety, and efficacy at therapeutic doses. In particular, there were no new or unexpected adverse events or safety concerns in African-American patients compared with their Caucasian counterparts, and neither the frequency nor severity of adverse events differed significantly between racial groups. Although approximately 44% of African-American patients and 59% of Caucasian patients reported adverse events, the vast majority of these were mild or moderate and transient. These findings should be reassuring to African-American ADHD patients, who may be less likely in general to be evaluated and treated for ADHD (vs their Caucasian counterparts), as well as their parents, many of whom may harbor more negative treatment expectations and overall distrust of the health care system compared with Caucasians. It is also reassuring to note that atomoxetine was well tolerated by patients of both races because the maximum permitted daily dose of atomoxetine when the study was conducted (1.8 mg/kg or 120 mg) exceeds values in the current US manufacturer’s label for the compound.

Previously, similar responses to atomoxetine were observed in Latino compared with Caucasian children and adolescents with ADHD.27 Latino patients were more likely to report decreased appetite or dizziness, whereas their Caucasian counterparts were more likely to report abdominal or throat pain. The ADHDRS-IV-P: I Hyperactivity/Impulsivity subscale decreased to a significantly greater degree in Latino patients, but these patients also had significantly higher baseline scores.

In the present study, both the African-American and Caucasian subgroups used similar mean modal doses of atomoxetine and had similarly low rates of discontinuation due to adverse events (2.4% and 3.0%, respectively). Significantly lower proportions of African-American (vs Caucasian) patients reported ≥1 treatment-emergent adverse event, including vomiting and fatigue. A relatively lower adverse-event reporting rate in African-American patients might be consistent with a cultural pattern toward stoicism concerning adverse-event reporting, which has been observed in some of the authors’ practices (A.P., E.R.), and/or to distrust of health care providers.

No serious safety concerns, cases of suicidal behavior, or significant changes from baseline in QTc interval were observed following atomoxetine treatment in either racial subgroup. Atomoxetine treatment modestly but significantly increased mean pulse rate from baseline in both groups and significantly increased blood pressure in Caucasian but not African-American patients. Caucasian patients had significantly higher pulse rates than did African-Americans at baseline. The mild pressor effects were of similar magnitude to changes previously observed in short-term atomoxetine studies involving similarly aged children and adolescents.30 Changes in pulse rate and blood pressure associated with atomoxetine are consistent with its modest peripheral noradrenergic activity. These mild effects typically occur early in therapy, then stabilize, and are usually not clinically significant.30

African-American and Caucasian patients experienced similar, significant reductions (improvements) from baseline to end point in ADHDRS-IV-P: I total score, inattention and hyperactivity/impulsivity subscale scores, CGI-ADHD-S, CPRS-R:S total score, and Oppositional, Cognitive Problems, Hyperactivity, and ADHD Index subscale scores. There were no significant racial differences in any of these efficacy measures. Although improvements in ADHDRS-IV-P:I total score and in hyperactivity-impulsivity symptoms were significantly more marked in Caucasian patients (vs African-American patients) at the first postbaseline office visit, this difference did not persist. Response rates and remission rates also were similar in African-American and Caucasian patients. The proportion of poor CYP2D6 metabolizers in African-American (3.6%) compared with Caucasian (7.5%) patients in our study suggests that our patients were representative of larger populations.

Previous use of stimulants to manage ADHD was less frequent in African-American compared with Caucasian patients (63.9% vs 78.5%; P=.004), which is consistent with prior published findings of racial disparities in ADHD treatment.1,11-13 These findings are also consistent with a previous study showing that Caucasian children were more than twice as likely to receive methylphenidate treatment than their African-American counterparts in a population with limited incomes.31 A more recent retrospective chart review involving ADHD patients age 5 to 20 who were treated in an urban referral center demonstrated that both race and insurance type were associated with specific stimulant treatment patterns, such that Medicaid-insured non–African-American children and privately insured African-American children tended to receive higher stimulant doses.32

Further, the lower prior exposure to stimulant medications in African-Americans compared with Caucasians may reflect African-Americans’ lower overall levels of pharmacologic treatment resulting from both inadequate access to appropriate health care services and, perhaps more important, the significant concerns in the African-American population about ADHD medications.19 It is possible that the availability of a nonstimulant in the present study might have facilitated participation by African-Americans, who might otherwise have shunned pharmacologic treatment.

The somewhat lower treatment adherence rate among African-American compared with Caucasian patients in the present study might be consistent with such negative perceptions about ADHD medications. Also potentially consistent with more negative treatment expectations was the finding of a significantly lower improvement in the ADHDRS-IV-P:I total score and hyperactivity-impulsivity symptom scores among African-Americans compared with Caucasians at the first postbaseline treatment assessment (FIGURES 2 and 3). However, this possible placebo effect was short lived, and improvements in ADHD with atomoxetine were similar in the 2 racial subgroups at each subsequent visit.

Potential limitations of this analysis include its relatively small number of African-Americans (n=83) compared with Caucasians (n=1090), its open-label design and short treatment duration, and its post hoc nature, which is inherently hypothesis generating rather than hypothesis proving. The open-label design does not allow us to conclusively rule out a placebo response. In addition, the investigators’ unconscious or subconscious attitudes might have biased treatment responses through different behaviors toward patients enrolled in this study to receive atomoxetine, compared with other patients in the investigators’ practices who received atomoxetine or other ADHD medications. The short-term nature of the study also did not enable determination of possible longer-term effects in the 2 populations.

We also did not capture baseline data on socioeconomic status and other related variables, including the parents’ marital and occupational status and their educational attainment, and insurance coverage of the family. The finding that significantly higher proportions of Caucasians compared with African-Americans reported ≥1 treatment-emergent adverse event may be ascribed in part to reluctance of African-American patients to report adverse events, given any sense of stigma or distrust of health care providers. Such effects might have resulted in our overestimating the tolerability of atomoxetine in African-Americans compared with Caucasians.

In addition, none of the instruments used to assess treatment efficacy has been specifically validated in African-American populations. The primary purpose of our study was to assess the tolerability and safety of atomoxetine in African-American compared with Caucasian patients. Our criteria for treatment efficacy, which were consistent with prior randomized controlled trials involving atomoxetine, might be open to challenge, including a ≥25% decrease from baseline in the ADHDRS-IV-P:I total score and a ≥2-point reduction in CGI-ADHD-S. Arguably, a patient with a CGI-ADHD-S score that decreases from 6 at baseline to 4 after atomoxetine treatment would still have moderate symptoms of ADHD. In recent years, researchers have advocated using remission as a central aim of ADHD treatment.33 The present study included remission after atomoxetine treatment as an efficacy end point: an ADHDRS-IV-P:I total T-score ≤60 or a CGI-ADHD-S score ≤2 at study end point. It can be argued that an ADHDRS-IV-total T score of 60 is still 1 standard deviation above age-appropriate norms in the general population. Further, it is not clear whether improvements from baseline to end point or achievement of symptom remission at end point is a more appropriate outcome measure for treatment efficacy in African-American compared with Caucasian patients. However, these were the pre-specified endpoints in the 2 studies, as well as many other randomized controlled trials involving atomoxetine. Our findings warrant further evaluation in longer-term, prospective, randomized controlled trials involving larger numbers of African-Americans assessed by a potentially more liberal range of efficacy outcome variables.

  CONCLUSION

Atomoxetine was well tolerated, safe, and effective in the management of ADHD in African-American children and adolescents, who may bear a disproportionate disease burden because of undertreatment. The availability of an efficacious nonstimulant alternative also may help to facilitate pharmacologic treatment in a population that might otherwise be hesitant to pursue it. Future longer-term, prospective, randomized, placebo-controlled trials that involve larger numbers of African-American patients and that examine not only efficacy and effectiveness, but also treatment acceptability, are warranted to confirm these findings.

DISCLOSURES AND ACKNOWLEDGEMENTS: Assistance in manuscript preparation was provided by Johanna B. Grossman, PhD, and Stephen W. Gutkin, Rete Biomedical Communications Corp. (Wyckoff, NJ). The authors wish to thank Virginia Sutton, PhD, for performing certain statistical analyses. Dr. Sutton was with Eli Lilly and Company when the study was conducted and data analyzed, and is now with i3 Research (Cary, NC). Dr. Durell and Mr. Williams are employees of and (minor) shareholders in Eli Lilly and Company. Dr. Durell also served as an independent paid consultant to Eli Lilly and Company before joining the company as a full-time employee. Dr. Pumariega is a consultant for Eli Lilly and Company through the Multicultural Multibrand Advisory Committee and is on the speakers’ bureau of Forest Pharmaceuticals. Dr. Rothe is a speaker and consultant for McNeil Pharmaceuticals. Dr. Tamayo was an employee of Eli Lilly and Company during the early stages of this study and is now a consultant for Eli Lilly and Company, Pfizer Inc, and Wyeth Pharmaceuticals. Dr. Baron serves or has served on advisory boards for Eli Lilly and Company, Shire, and Pfizer Inc, and also has received grant support from Eli Lilly and Company. These studies and this report were supported by Eli Lilly and Company (Indianapolis, IN).

STUDY IDENTIFIERS: B4Z-MC-LYAB and B4Z-MC-LYBB (studies#5197 and 2161). Although it was not required that these open-label studies be registered on www.clinicaltrials.gov or other government websites, the data are available on the following Eli Lilly and Company corporate website: www.lillytrials.com/results/by_product/results_strattera.html.

    References

  1.  Centers for Disease Control and Prevention. Mental health in the United States: prevalence of diagnosis and medication treatment for attention-deficit/hyperactivity disorder—United States, 2003. MMWR Morb Mortal Wkly Rep. 2005;54(34):842–847.
  2. Rappley MD. Attention deficit-hyperactivity disorder. N Engl J Med. 2005;352(2):165–173.
  3. Bazargan M, Calderon JL, Heslin KC, et al. A profile of chronic mental and physical conditions among African-American and Latino children in urban public housing. Ethn Dis. 2005;15(suppl 5):S53–S59.
  4.  American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th ed, text revision. Washington, DC: American Psychiatric Association; 2000.
  5.  American Academy of Pediatrics. Clinical practice guideline: diagnosis and evaluation of the child with attention-deficit/hyperactivity disorder. Pediatrics. 2000;105(5):1158–1170.
  6. Biederman J, Newcorn J, Sprich S. Comorbidity of attention deficit hyperactivity disorder with conduct, depressive, anxiety, and other disorders. Am J Psychiatry. 1991;148(5):564–577.
  7. Brown RT, Borden KA. Hyperactivity at adolescence: some misconceptions and new directions. J Clin Child Psychol. 1986;15(3):194–209.
  8. Hechtman L. Assessment and diagnosis of attention-deficit/hyperactivity disorder. Child Adolesc Psychiatr Clin N Am. 2000;9(3):481–498.
  9. Klein RG, Mannuzza S. Long-term outcome of hyperactive children: a review. J Am Child Adolesc Psychiatry. 1991;30(3):383–387.
  10. Thorley G. Review of follow-up and follow-back studies of childhood hyperactivity. Psychol Bull. 1984;96(1):116–132.
  11. Bussing R, Zima BT, Gary FA, et al. Barriers to detection, help-seeking, and service use for children with ADHD symptoms. J Behav Health Serv Res. 2003;30(2):176–189.
  12. Stevens J, Harman JS, Kelleher KJ. Race/ethnicity and insurance status as factors associated with ADHD treatment patterns. J Child Adolesc Psychopharmacol. 2005;15(1):88–96.
  13. Olfson M, Gameroff MJ, Marcus SC, et al. National trends in the treatment of attention deficit hyperactivity disorder. Am J Psychiatry. 2003;160(6):1071–1077.
  14. Arnold LE, Elliott M, Sachs L, et al. Effects of ethnicity on treatment attendance, stimulant response/dose, and 14-month outcome in ADHD. J Consult Clin Psychol. 2003;71(4):713–727.
  15. Davison JC, Ford DY. Perceptions of attention deficit hyperactivity disorder in one African-American community. J Negro Educ. 2001;70(4):264–274.
  16. Bailey RK, Owens DL. Overcoming challenges in the diagnosis and treatment of attention-deficit/hyperactivity disorder in African Americans. J Natl Med Assoc. 2005;97(10 suppl):5S–10S.
  17. Bussing R, Schoenberg NE, Perwien AR. Knowledge and information about ADHD: evidence of cultural differences among African-Americans and White parents. Soc Sci Med. 1998;46(7):919–928.
  18. Bussing R, Schoenberg NE, Rogers KM, et al. Explanatory models of ADHD: do they differ by ethnicity, child gender, or treatment status? J Emot Behav Disord. 1998;6(4):233–242.
  19. Dosreis S, Butz A, Lipkin PH, et al. Attitudes about stimulant medication for attention-deficit/hyperactivity disorder among African American families in an inner city community. J Behav Health Serv Res. 2006;33(4):423–430.
  20. Kelsey DK, Sumner CR, Casat CD, et al. 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. 2004;114(1):e1–e8.
  21. Michelson D, Buitelaar JK, Danckaerts M, et al. Relapse prevention in pediatric patients with ADHD treatment with atomoxetine: a randomized, double-blind, placebo-controlled study. J Am Acad Child Adolesc Psychiatry. 2004;43(7):896–904.
  22. Spencer T, Heiligenstein JH, Biederman J, et al. Results from 2 proof-of-concept, placebo-controlled studies of atomoxetine in children with attention-deficit/hyperactivity disorder. J Clin Psychiatry. 2002;63(12):1140–1147.
  23. Michelson D, Adler L, Spencer T, et al. Atomoxetine in adults with ADHD: two randomized, placebo-controlled studies. Biol Psychiatry. 2003;53(2):112–120.
  24. Spencer T, Biederman J, Wilens T, et al. Effectiveness and tolerability of atomoxetine in adults with attention deficit hyperactivity disorder. Am J Psychiatry. 1998;155(5):693–695.
  25. Starr HL, Kemner J. Multicenter, randomized, open-label study of OROS methylphenidate versus atomoxetine: treatment outcomes in African-American children with ADHD. J Natl Med Assoc. 2005;97(suppl):11S–16S.
  26. DuPaul GJ, Power TJ, Anastopoulos AD, et al. ADHD Rating Scale-IV: Checklist, Norms, and Clinical Interpretations. New York, NY: Guilford Press; 1998.
  27. Tamayo JM, Pumariega A, Rothe EM, et al. Latino versus Caucasian response to atomoxetine in attention-deficit hyperactivity disorder. J Child Adolesc Psychopharmacol. 2008;18:44–53.
  28. Kaufman J, Birmaher B, Brent D, et al. Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADSPL): initial reliability and validity data. J Am Acad Child Adolesc Psychiatry. 1997;36(7):980–988.
  29. Kumar G, Steer RA. Factorial validity of the Conners’ Parent Rating Scale-Revised: Short Form with psychiatric outpatients. J Pers Assess. 2003;80(3):252–259.
  30. Wernicke JF, Faries D, Girod D, et al. Cardiovascular effects of atomoxetine in children, adolescents, and adults. Drug Saf. 2003;26(10):729–740.
  31. Zito JM, Safer DJ, dosReis S, et al. Methylphenidate patterns among Medicaid youths. Psychopharmacol Bull. 1997;33(1):143–147.
  32. Lipkin PH, Cozen MA, Thompson RE, et al. Stimulant dosage and age, race, and insurance type in a sample of children with attention-deficit/hyperactivity disorder. J Child Adolesc Psychopharm. 2005;15(2):240–248.
  33. Steele M, Jensen PS, Quinn DM. Remission versus response as the goal of therapy in ADHD: a new standard for the field? Clin Ther. 2006;28:1892–1908.

CORRESPONDENCE: Todd M. Durell, MD, Eli Lilly and Company, Lilly Corporate Center DC4135, Indianapolis, IN 46285 USA. E-MAIL: durellto@lilly.com