Original Article

Journal of Child Neurology

1-7
The Short-Term Placebo Response ª The Author(s) 2018
Reprints and permission:
sagepub.com/journalsPermissions.nav
in Children With Attention-Deficit DOI: 10.1177/0883073818756403
journals.sagepub.com/home/jcn
Hyperactivity Disorder (ADHD)

Ayala Cohen, PhD1, Moran Plonsky-Toder, MD2, and Emanuel Tirosh, MD3

Abstract
To assess short-term placebo response in 6- to 13-year-old children with ADHD, children who were administered a double-blind
placebo-methylphenidate trial, 1 week each, were included in the analysis. Conners’ parents and Teacher Rating scales, the
Aggregate Neurobehavioral Student Health and Educational Review inventory, and the Matching Familiar Figure Test were
employed. A reduction of 30% or more in one or more of the teachers report subscales was observed in 18.8% of the participants.
Attention test performance resulted in 58% of children exhibiting reduction in error rates and 36.2% exhibited longer latency
period. Significant correlations between placebo response and methylphenidate response in all of the teachers report subscales
were found. Base line severity, learning problem and emotional status were found associated with placebo response. Short-term
placebo response should be accounted for in children with ADHD.

Keywords
ADHD, anxiety, behavior, children, treatment

Received September 7, 2017. Received revised December 22, 2017. Accepted for publication January 4, 2018.

Attention-deficit hyperactivity disorder (ADHD) is the most emotional/behavioral disorders, and the factors associated with
prevalent neurobehavioral disorder among children.1,2 ADHD such response remain to be elucidated. The 2 main aims of the
has variable clinical presentations3-10 and consists of 3 types of present study were (1) to assess the short-term placebo
ADHD presentation: type 1: predominantly inattentive (PI); response prevalence and magnitude among children with
type 2: predominantly hyperactive and impulsive (PH/I); and ADHD and (2) to identify factors that possibly predicts such
type 3: a combined type (C).11 The empirical first-line treat- response. We hypothesized that short-term placebo response
ment for ADHD is based on stimulant medication, most com- rate would be as high as 30% in a substantial percentage
monly methylphenidate.12,13 Studies have shown that the of children. We expected that factors such as young age,
placebo effect contributes to positive outcomes in many phys- specific learning disability (SLD), ADHD predominant presen-
ical and mental disorders, including ADHD.14-22 We therefore tation, and emotional behavioral problems would be related
hypothesized that a sizeable placebo effect would be demon- to such response.
strated. The placebo effect is defined as “a genuine psycholo-
gical or physiological effect, in a human or another animal,
which is attributable to receiving a substance or undergoing a
procedure, but is not due to the inherent power of that sub-
stance.”23 It is known that some children with ADHD respond 1
Faculty of Industrial Engineering and Management, The Technion, Israel
well to administration of placebo. Past research suggested that Institute of Technology, Haifa, Israel
2
Department of Pediatrics, The Ruth Rappaport Children’s Hospital, Haifa,
20% to 30% of unselected children with ADHD are placebo
Israel
responders, with an average effect size of 0.32 compared to 3
The Hannah Khoushy Child Development Center, Bnai Zion Medical Center
baseline.21,24 An attempt to delineate characteristics of placebo and The Faculty of Medicine, Technion, Haifa, Israel
responders in pediatric clinical trials of ADHD identified inat-
tention, stimulant naı̈ve status, associated tic disorder, and non- Corresponding Author:
Emanuel Tirosh, MD, The Hannah Khoushy Child Development Center, Bnai
white ethnicity as predictive of a more enhanced placebo Zion Medical Center and The Faculty of Medicine, Technion, 33 Blitental St.,
response.25 However, the size of short-term placebo effect in Haifa, Israel 3479133.
children with diagnosed ADHD not associated with other Email: emi.tirosh@gmail.com
2 Journal of Child Neurology XX(X)

Methods The Aggregate Neurobehavioral Student Health and Educational

Review (ANSER) inventory28 has been used in previous clinical stud-
Study design and procedure: a retrospective study, based on data ies and reliability and validity have been previously reported.31,32 The
collected from medical files of children referred by community pedia- following subscales were included in the present study: social,
tricians and family practitioners for suspected ADHD. anxiety, and strengths. The items related to mood were used to score
dysthymia (not a disorder). The total score obtained using a 0 to 2
Participants Likert-type scale for each item was used as a continuum rather than a
categorical diagnostic classification. In addition, attention was also
Between the years 2001 and 2012, a total of 236 children underwent a assessed using the Matching Familiar Figures Test.33 The psycho-
crossover study of methylphenidate and placebo. The data were part of metric properties of this test have been reported previously.34,35 The
an n-of-1 trial (see procedure) routinely employed in the Hannah Conners’ Teacher Rating Scale as well as the Aggregate Neurobeha-
Khoushy child development center of the Bnai-Zion medical center, vioral Students Health and Learning Review behavioral section were
catering for the greater Haifa city area. Exclusion criteria were previ- blindly reviewed following placebo and methylphenidate treatment.
ous or current medical treatment (stimulant, or any other drugs with
possible effects on the central nervous system), general cognitive
deficit as assessed by the board of education, or enrollment into a Statistical Analysis
special education service, and comorbid neurologic, cardiac, or psy- Because the validity of the separate category of the hyperactive-
chiatric disorder as revealed by the Diagnostic and Statistical Manual impulsive ADHD subtype has been questioned,36 the predominantly
of Mental Disorders (Fourth Edition)26–based interview. hyperactive-impulsive and combined ADHD subtypes (HI/C) were
collapsed into 1 group. This approach is also in line with the Diag-
nostic and Statistical Manual of Mental Disorders (Fifth Edition)
Procedure terminology of presentation rather than categorical subtypes.11
The assessment included interviews with the child and parents, phys- All statistical analyses were done with the SAS Statistical Anal-
ical and neurologic examinations, attention and sensory testing, and ysis System software.37 For comparative analysis, at each stage and
questionnaires filled by the parents and teachers. The final diagnosis for each group, logistic regressions were fitted. The SAS procedure
of ADHD with no associated emotional or behavioral disorder was for Generalized Linear Mixed Models (GLIMMIX) was applied to
based on Diagnostic and Statistical Manual of Mental Disorders allow for the repeated measures structure of the data. Bivariate rela-
(Fourth Edition) criteria26 and the Conners’ parents’ and teacher’s tions were assessed using Fisher exact test and standard t tests for
questionnaires.27 The diagnosis of dysthymic mood was defined as categorical and continuous variables, respectively. For missing data
at least 2 months of either 2 of the following complaints since the last (15% of the participants’ files had some missing data), multiple
year: appetite or sleep problems, fatigue, low self-esteem and hope- imputation was used using the MI procedure of SAS. The imputed
lessness; anxiety and psychosomatic traits that were based on parents’ values were the average of 40 imputations. In pairwise comparisons
interviews and the parent and teacher reports using the Aggregate of the least squares (LS) means, the P values were adjusted to
Neurobehavioral Students Health and Learning Review (ANSER).28 account for the multiple comparisons. Effect sizes were calculated
Specific learning deficits (SLDs) were determined using the school using Glass’s delta model.
report. The data were collected using an n-of-1 trial as a routine The relative difference between the various stages was assessed,
clinical procedure accepted by all families, candidates for methylphe- using the following equation: [CTRS (placebo/methylphenidate) –
nidate treatment, using a randomized double-blind crossover study, CTRS (baseline)] / [CTRS (baseline)] accounting for the initial value.
for the purpose of comparing placebo and stimulant medication.29 The As the binary dependent variable in the logistic regressions, we con-
cohort has been managed by one clinician responsible for the ADHD sidered a 30% change in the Conners’ Teacher Rating Scale baseline
clinic. Each stage of the trial lasted 1 week, in which regular scores. Baseline variables were included as independent variables in
immediate-release methylphenidate (0.3 mg/kg/d PO) or lookalike addition to the individual characteristics including age, gender, type of
placebo capsules were administered once daily at 7.30 AM, with a ADHD, learning difficulties, dysthymia, anxiety, psychosomatic com-
washout interval of 44 hours between the 2 periods. Both methylphe- plaints, social skills, and associated strengths as well as response to
nidate and placebo capsules packages were prepared by the hospital methylphenidate. Pearson correlation was used to examine the relation
pharmacy using a table of random numbers with a balanced assign- between baseline disorder severity and placebo response.
ment. Parents and teachers were unaware of the drug/placebo periods
and teachers were not aware of the specific purpose of the weekly
questionnaires. The Conners’ Teacher Rating Scale (CTRS),27 visual
attention test, and child and parents’ interviews were recorded at the Results
clinic at the end of each week. The final cohort included 133 drug-naı̈ve children (76.7%
boys, 23.3% girls), between 6 and 13 years of age (mean
Instruments 9.06 years, SD 1.8), who were diagnosed with ADHD and were
candidates for treatment with methylphenidate.
The Conners’ Teacher Rating Scale27 includes 28 items, scored on a 0
Eighty-four children (63%) presented with ADHD type 1
to 3 Likert-type scale relating to the following subscales: conduct,
hyperactivity-impulsivity, inattentive-passive, and hyperactivity (predominantly inattentive), 9 (7%) presented with type 2 (pre-
index. The Conners’ Parent Rating Scale (CPRS)27 includes 48 items dominantly hyperactive and impulsive) and 39 (30%) presented
and the following subscales: conduct, learning, psychosomatic, impul- with type 3 (combined).26 ADHD type of 1 child was missing.
sive- hyperactive, anxiety, and hyperactivity index. The psychometric Specific learning disability diagnosed as a delay of 1.5 SD or
properties of the questionnaires have been reported previously.30 more in at least 1 learning field such as arithmetic or reading
Cohen et al 3

Table 1. Average CTRS Scores.a

Baseline Placebo Methylphenidate

CTRS subscale M SD M SD Effect size M SD Effect size

Conduct 68. 7 14. 7 63. 4 15. 3 –0. 36 59. 4 15. 4 –0. 63

Hyperactive- impulsive 72. 8 12. 8 67. 4 13. 9 –0. 42 60. 4 13. 8 –0. 97
Inattentive-passive 64. 1 9. 3 60. 2 9. 4 –0. 42 56 9. 2 –0. 87
Hyperactivity- index 72. 2 9. 6 67. 4 12. 8 –0.5 60. 2 12. 1 –1. 25

Abbreviation: CTRS, Conners’ Teacher Rating Scale.

a
In each row, all means are significantly different between placebo and baseline and methylphenidate and placebo according to the least squares–means
comparisons; P < .0002.

Table 2. Placebo Responders in CTRS Scores as Compared to BL respectively. A similar analysis employed for methylphenidate
(>30%) and MPH (Similar or Better Placebo Response). response revealed a similar effect related to baseline Conners’
Teacher Rating Scale subscale scores. In addition, unlike the
>30% relative Zero or positivea relative
results obtained for placebo response, a significant effect of
CTRS subscale difference: BL-P difference: MPH-P
younger age for the subscales of inattentive-passive and
Conduct 9.8 45.9 hyperactivity index and low anxiety as reported by the parents
Hyperactivity 8.3 34.1 for the hyperactivity index Conners’ Teacher Rating Scale
Inattentive-passive 3.8 36.8 subscale were demonstrated. Additionally, we found signifi-
Hyperactivity index 6.8 33.8 cant correlations between placebo response and the response
Abbreviations: BL, baseline; CTRS, Conners’ Teacher Rating Scale; MPH, to methylphenidate in all Conners’ Teacher Rating Scale
methylphenidate; P, Placebo. subscales (Table 4). Analyzing the ADHD predominant
a
Zero or positive: no difference or placebo superiority. subtypes separately reveals consistent correlations between
the subscale scores at baseline and their respective placebo
response (Table 5).
with a normal verbal and general intelligence quotient (IQ) was
diagnosed in 48% of the children.
Based on Conners’ Teacher Rating Scale ratings, both pla- Discussion
cebo and methylphenidate showed a significant reduction in The aim of our study was to assess placebo response in children
ADHD manifestations. However, methylphenidate was associ- with ADHD and identify possible child’s variables related to
ated with a significantly greater effect (Table 1). Of the total such response. Of the total cohort, 76.7% were males, a pro-
cohort, 18.8% exhibited a reduction of 30% or more in at least 1 portion similar to the known proportion diagnosed with ADHD
of the Conners’ Teacher Rating Scale subscales following pla- in the population.38 Of the children, 63% presented with the
cebo administration. Similar or better performance being predominantly inattentive type ADHD. This prevalence is dif-
treated with placebo compared with methylphenidate was ferent from that found in a recent meta-analysis with 45% of
observed in 61.6% of children in at least 1 of Conners’ Teacher children with predominantly inattentive type.38 This difference
Rating Scale subscales (Table 2). Using Matching Familiar can be the result of a small sample size, strict exclusion criteria,
Figures Test scores, a significant improvement with placebo and referral bias. Our results reveal that 18.8% of the partici-
treatment compared to baseline was shown, 58% of children pants showed placebo response of 30% or more in at least one
exhibited reduction in error rates, and 36.2% exhibited longer of the Conners’ Teacher Rating Scale subscales, with effect
response latency while on placebo. size between 0.36 and 0.5. Additionally, the attention/impul-
Using bivariate analyses, we did not find a statistically sig- sivity test scores showed placebo response in both latency
nificant relationship between ADHD type and placebo period and error rate. The size of the placebo responses in the
response. A regression analysis revealed a significant positive different Conners’ Teacher Rating Scale subscales found in
relationship between placebo response of 30% or more and our study are smaller than those obtained in previous reports
their baseline respective scores in all the Conners’ Teacher focusing on other emotional behavioral disorders and partic-
Rating Scale subscales (Table 3). Emotional and specific learn- ularly in depression21,39 and anxiety39-41 as well as in children
ing problems diagnosed using the Aggregate Neurobehavioral with intellectual disabilities.42 Newcorn et al,25 in their study
Students Health and Learning Review system and the school pertaining to unselected children with ADHD including chil-
report, respectively, appear to be differentially related to pla- dren with psychiatric comorbidities, also found a similar
cebo response in the different Conners’ Teacher Rating Scale response rate.
subscale scores (Table 3). No predictive variables for a 30% Methylphenidate is the first line of treatment for
response other than the baseline scores were found for the ADHD.1,5,12 Therefore, it can be considered as a reference for
subscales of hyperactivity-impulsivity and hyperactivity index, expected intervention effect. Most previous studies examined
4 Journal of Child Neurology XX(X)

Table 3. The Variables Related to Placebo and Methylphenidate (MPH) Response.

CTRS subscale

Conduct Hyperactive Impulsive Inattentive Passive Hyperactivity Index

OR (CI); P OR (CI); P OR (CI); P OR (CI); P

Placebo
Variable
Baseline 1.05 (1.00, 1.10); .04 1.15 (1.02, 1.20); .03 1.10 (1.02, 1.16); .08 1.10 (1.02, 1.29); .02
No SLD 0.08 (0.01, 0.49); .006
No dysthymia 6.76 (1.30, 34.50); .02 9.52 (1.13, 80); .03
Methylphenidate
Baseline 1.06 (1.00, 1.10); .04 1.11 (1.00, 1.2); .001 1.10 (1.01, 1.12); .05 1.11 (1.02, 1.21); .03
Young age 1.06 (1.02, 1.10); .03 1.05 (1.00, 1.10); .05
Anxiety 0.84 (0.71, 0.99); .05

Abbreviations: CI, confidence interval; CTRS, Conners’ Teacher Rating Scale; OR, odds ratio; SLD, specific learning disability.

Table 4. Correlations Between Methylphenidate and Placebo technique and inconsistent account for initial values in the
Response. analysis.21,23,25,42
We believe that these results point to the added value of a
CTRS subscale r P value
double-blind “n-of-1 trial” prior to the beginning of more long-
Conduct 0.58 <.0001 term treatment. This approach was also found to be related
Hyperactivity 0.48 <.0001 to parents’ and children’s attitudes and maintenance on
Inattentive-passive 0.59 <.0001 drug.29,43-48 Mechanisms proposed to underlie the placebo
Hyperactivity index 0.58 <.0001 effect include expectancy effect, conditioning, learning by
Abbreviation: CTRS, Conners’ Teacher Rating Scale. imitation, and changes in the caregiver perception and
action.21,23-25 All these mechanisms are potentially responsible
for the different response-related variables discussed below:
Table 5. Correlations Between Placebo Response and Baseline CTRS
Scores as Related to the 2 ADHD Subtypes. 1. Type of ADHD: The lack of difference in placebo
response as related to the type of ADHD, the similar
ADHD subtype
correlation of the response as related to baseline in the 2
CTRS subscale PI, r HI/C, r subtypes and the significant relationship with clinical
a
symptoms. All point to the possible validity of a dimen-
Conduct 0.66 0.82a
sional approach to the disorder.
Hyperactive-impulsive 0.59a 0.63a
Inattentive-passive 0.53a 0.37b 2. Learning disability: A previous study showed a reduced
Hyperactivity Index 0.56a 0.65a response to methylphenidate in children with ADHD
and learning disability compared to ADHD alone.49
Abbreviations: ADHD, attention-deficit hyperactivity disorder; CTRS, We did not find previous studies that explored the rela-
Conners’ Teacher Rating Scale; HI/C, hyperactive impulsive/combined; PI,
predominantly inattentive. tionship between learning disability and placebo
a
P < .0001. response. It does appear, however, that associated learn-
b
P ¼ .008. ing disability is predictive of a placebo response in
children with ADHD. This finding possibly points to
a different underlying mechanism, contributing to
the placebo response by comparing the outcome scores ADHD in these children.49 Alternatively, it is concei-
obtained following placebo administration with that recorded vable that the perceptions associated with placebo
at baseline. One meta-analysis42 compared placebo and drug administration are different in children with and with-
treatment (including, but not limited to methylphenidate) and out specific learning disability, and consequently a
found an effect size of 0.3. Comparing placebo and methylphe- change in the conduct subscale scores but not in other
nidate response, albeit a general superiority of methylphenidate aspects of ADHD are observed.
was evident, we found that 61.6% of the children showed sim- 3. Dysthymia: We showed statistically significant associ-
ilar short-term response to placebo and methylphenidate in at ation between no dysthymia and placebo response, in
least one of the Conners’ Teacher Rating Scale subscales. The the conduct and hyperactivity subscales of the Conners’
effect size ranged between 0.26 and 0.6. The between-studies Teacher Rating Scale. The possible relationship
differences in the observed placebo responses although rather between emotional state and placebo response remains
small are attributable to a lack of standardized measurement elusive. 50 We found no studies that reported the
Cohen et al 5

relationship between dysthymia (not depression) and the proportion of children with hyperactive impulsive and com-
placebo response. It is possible that dysthymic state in bined ADHD in the present study. The exclusion criteria we
children with ADHD is emotionally consuming and employed may limit our results to children with no associated
thus may restrict a positive placebo response. disorders. The common practice in n-of-1 trial drug/placebo
cycles ranges from 2 days to few weeks of different
The strongest and most persistent predicting factor associ- dosing.25,54,55 In this study, we examined the short-term
ated with placebo response is that of the severity of the Con- placebo effect, using a 1-week trial of placebo and 1 week of
ners’ Teacher Rating Scale baseline scores in the respective methylphenidate. Therefore, we cannot extrapolate the long-
subscales. Unlike our findings, Newcorn and colleagues25 term placebo response. The baseline results of the Conners’
showed no correlation between baseline disorder severity and Teacher Rating Scale were used for the calculated effect of
placebo response in children with ADHD. However, their study both trial periods as employed in our clinical routine n-1-trial.29
used only physicians’ assessments, without parents’ or teach- One should also take into account a possible change in baseline
ers’ reports and included children with major associated dis- following the washout period of the first week as compared to
orders.25 In adults with ADHD, there were reports supporting the first baseline report.
our results, with positive correlation between disease severity
and placebo response.51 Similarly, such correlation was evident
among children with other emotional behavioral disorders.39 Conclusions
The effect of baseline severity of the disorder is possibly Our study demonstrates that among children with ADHD, the
related to the suggested mechanisms of placebo response short-term placebo effect is significant, and present in a mean-
including in particular expectation, in which a parent and a ingful percentage of children. We did not find a significant
teacher of a child with severe disorder will experience any correlation between ADHD type and placebo response, but
change as significant, and their evaluation would be inflated. we did find several factors that presumably predict significant
Previous studies showed positive correlation between more placebo response, most importantly severity of baseline disor-
severe ADHD and better treatment response as reported by a der, specific learning disability and better emotional state. Our
caregiver, even when objective parameters have not shown the results lend support to the value of an n-of-1 trial in tailoring
same improvement.52,53 the medical and nonmedical treatment for a child with
The correlation between placebo and methylphenidate ADHD.56 However, clinicians should exercise caution in light
response: in line with our results, Weimer et al21 suggested that of the significant placebo effect as compared not only to base-
among children who are “drug responders” in drug trials, there line but also to methylphenidate. The discussion of such effect
are many who are in fact “placebo responders.” These children with the child and the parents could lead to a more conscious
would have been identified if “n-of-1” trials were performed decision related to the possible intervention. A future long-term
routinely. In other words, drug response is at least partly due to study can help to understand the long-term placebo effect. The
placebo response as related to child’s behavior. Therefore, the significant correlation between placebo and methylphenidate
response magnitude is a personal attribute not necessarily response emphasizes the placebo influence, in the short term at
related to the drug per se. We conclude that children who tend very least, during the empirical treatment with methylpheni-
to be placebo responders may respond better to methylpheni- date. Longer n-of-1 trials and the inclusion of children with
date as well, partly because of inherent placebo response and associated disorders may better demonstrate sustained placebo
partly because of their unique characteristics. Whichever be the effects and possibly generalize our results to a broader popu-
case, our results point to placebo effect probably being a non- lation with ADHD.
negligible part of a short-term drug response in a proportion of
children with ADHD. Therefore, it appears that offering chil- Author Contributions
dren and parents an n-1-trial is of value in the management of
children with ADHD. Children, foremost, should make an The authors shared responsibility in the writing of this paper and each
informed decision as to whether to use the procedure. took responsibility in performing the research.

Declaration of Conflicting Interests

Limitations The authors declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Our study included a relatively small group of children who
were referred to one service. However, we are not aware of any
Funding
bias related to the referral of children to our service or a bias
related to randomization of ADHD subtypes in the placebo/ The authors received no financial support for the research, authorship,
drug trials.45,54 Furthermore our results are valid for children and/or publication of this article.
with no associated disorders as formally diagnosed using the
Diagnostic and Statistical Manual of Mental Disorders criteria. Ethical Approval
It is plausible that the exclusion of children with associated The study was approved by the Helsinki committee of the Bnai-Zion
conduct and oppositional defiant disorders (ODD) has affected medical center (0095-12-BNZ).
6 Journal of Child Neurology XX(X)

References 21. Weimer K, Gulewitsch MD, Schlarb AA, Schwille-Kiuntke J,

1. Biederman J, Faraone SV. Attention-deficit hyperactivity disor- Klosterhalfen S, Enck P. Placebo effects in children: a review.
der. Lancet. 2005;3:237-248. Pediatr Res. 2013;74:96-102.
2. Feldman HM, Reiff MI. Clinical practice. Attention deficit- 22. Stein MA, Sikirica V, Weiss MD, Robertson B, Lyne A, Newcorn
hyperactivity disorder in children and adolescents. N Engl J Med. JH. Does guanfacine extended release impact functional impair-
ment in children with attention-deficit/hyperactivity disorder?
2014;370:838-846.
Results from a randomized controlled trial. CNS Drugs. 2015;
3. Biederman J. Attention-deficit/hyperactivity disorder: a selective
29:953-962.
overview. Biol Psychiatry. 2005;57:1215-1220.
23. Waschbusch DA, Pelham WE Jr, Waxmonsky J, Johnston C. Are
4. Daley D. Attention deficit hyperactivity disorder: a review of the
there placebo effects in the medication treatment of children with
essential facts. Child Care Health Dev. 2006;32:193-204.
attention-deficit hyperactivity disorder? J Dev Behav Pediatr.
5. Pliszka S; AACAP Work Group on Quality Issues. Practice para-
2009;30:158-168.
meter for the assessment and treatment of children and adoles-
24. Sandler AD, Glesne CE, Bodfish JW. Conditioned placebo dose
cents with attention-deficit/hyperactivity disorder. J Psychiatry.
reduction: a new treatment in attention-deficit hyperactivity dis-
2007;46:894-921.
order? J Dev Behav Pediatr. 2010;31:369.
6. Spencer TJ, Biederman J, Mick E. Attention-deficit/hyperactivity
25. Newcorn JH, Sutton VK, Zhang S, et al. Characteristics of pla-
disorder: diagnosis, lifespan, comorbidities, and neurobiology. J
cebo responders in pediatric clinical trials of attention-deficit/
Pediatr Psychol. 2007;32:631-642.
hyperactivity disorder. J Am Acad Child Adolesc Psychiatry.
7. Floet AM, Scheiner C, Grossman L. Attention-deficit/hyperactiv-
2009;48:1165-1172.
ity disorder. Pediatr Rev. 2010;3:56-66.
26. American Psychiatric Association. Diagnostic and Statistical
8. Wehmeier PM, Schacht A, Barkley RA. Social and emotional
Manual of Mental Disorders: DSM-IV-TR. Washington, DC:
impairment in children and adolescents with ADHD and the
American Psychiatric Association; 2000.
impact on quality of life. J Adolesc Health. 2010;46:209-217. 27. Goyette CH, Conners CK, Ulrich RF. Normative data on revised
9. Dalsgaard S. Attention-deficit/hyperactivity disorder (ADHD). Conners parent and teacher rating scales. J Abnorm Child Psy-
Eur Child Adolesc Psychiatry. 2013;22(suppl 1):S43-S48. chol. 1978;62:221-236.
10. Tarver J, Daley D, Sayal K. Attention-deficit hyperactivity dis- 28. Levine MD. The ANSER system (Aggregate Neurobehavioral Stu-
order (ADHD): an updated review of the essential facts. Child dent Health and Educational Review). Cambridge, MA: Educa-
Care Health Dev. 2014;40:762-774. tors Publishing Service; 1984.
11. American Psychiatric Association. Diagnostic and Statistical 29. Tirosh E, Elhasid R, Kamah SC, Cohen A. Predictive value of
Manual of Mental Disorders. 5th ed. Arlington, VA: American placebo methylphenidate. Pediatr Neurol. 1993;99:131-133.
Psychiatric Publishing; 2013. 30. Gianarris WJ, Golden CJ, Greene L. The Conners’ Parent Rating
12. Atkinson M, Hollis C. NICE guideline: attention deficit hyperac- Scales: a critical review of the literature. Clin Psychol Rev. 2001;
tivity disorder. Arch Dis Child Educ Pract Ed. 2010;95:24-27. 21:1061-1093.
13. Wolraich M, Brown L, Brown RT, et al; Subcommittee on Atten- 31. Schmidt SL, Snyder TJ, Rouget AC, Gray E. Empirical analysis
tion Deficit/Hyperactivity Disorder; Steering Committee on Qual- of the selective attention and associated behavior checklists of the
ity Improvement and Management. ADHD: clinical practice Aggregate Neurobehavioral Student Health and Educational
guideline for the diagnosis, evaluation, and treatment of attention- Review. J Dev Behav Pediatr. 2000;21:165-171.
deficit/hyperactivity disorder in children and adolescents. 32. Or D, Cohen A, Tirosh E. The reliability and validity of the
Pediatrics. 2011;128:1007-1022. Aggregate Neurobehavioral Student Health and Educational
14. Koshi EB, Short CA. Placebo theory and its implications for Review parent’s questionnaire (ANSER-PQ). J Child Neurol.
research and clinical practice: a review of the recent literature. 2010;25:157-164.
Pain Pract. 2007;7:4-20. 33. Kagan J, Rosman BL, Day D. Information processing in children,
15. Price DD, Finniss DG, Benedetti F. A comprehensive review of significance of analytic and reflective attitude. Psychol Monogr.
the placebo effect: recent advances and current thought. Annu Rev 1964;78:1-37.
Psychol. 2008;59:565-590. 34. Kirchner-Nebot T, Amador-Campos JA. Internal consistency of
16. Brody H, Miller FG. Lessons from recent research about the scores on Matching Familiar Figures Test-20 and correlation of
placebo effect—from art to science. JAMA. 2011;306:2612-2613. scores with age. Percept Mot Skills. 1998;86:803-807.
17. Annoni M. Highlights from the 2013 Science of Placebo thematic 35. Amador-Campos JA, Kirchner-Nebot T. Children’s embedded
workshop. Ecancermedicalscience. 2013;7:346. figures test and matching familiar figures test-20: factorial struc-
18. Gupta U, Verma M. Placebo in clinical trials. Perspect Clin Res. ture for boys and girls from 6 to 11 years old. Percept Mot Skills.
2013;4:49-52. 2001;93:709-912.
19. Jutte R. The early history of the placebo. Complement Ther Med. 36. Willcutt EG, Nigg JT, Pennington BF, et al. Validity of DSM-IV
2013;21:94-97. attention deficit/hyperactivity disorder symptom dimensions and
20. van Haselen R. The placebo revisited: what we do, don’t and subtypes. J Abnorm Psychol. 2012;121:991-1010.
maybe won’t know. Complement Ther Med. 2013;21:83-84. 37. SAS Institute Inc. SAS/STAT User’s Guide, Version 9.2. Cary,
NC: SAS Institute Inc; 2007.
Cohen et al 7

38. Willcutt EG. The prevalence of DSM-IV attention-deficit/hyper- the Adolescent Placebo Impact Composite Score (APICS). J Psy-
activity disorder: a meta-analytic review. Neurotherapeutics. chiatr Res. 2015;68:346-653.
2012;9:490-499. 49. Grizenko N, Bhat M, Schwartz G, Ter-Stepanian M, Joober R.
39. Parellada M, Moreno C, Moreno M, Espliego A, de Portugal E, Efficacy of methylphenidate in children with attention-deficit
Arango C. Placebo effect in child and adolescent psychiatric hyperactivity disorder and learning disabilities: a randomized
trials. Eur Neuropsychopharmacol. 2012;22:787-799. crossover trial. J Psychiatry Neurosci. 2006;31:46-51.
40. Cohen D, Deniau E, Maturana A, et al. Are child and adolescent 50. Flaten MA, Aslaksen PM, Lyby PS, Bjorkedal E. The relation of
responses to placebo higher in major depression than in anxiety emotions to placebo responses. Philos Trans R Soc Lond Biol Sci.
disorders? A systematic review of placebo-controlled trials. PLoS 2011;366:1818-1827.
One. 2008;3:e2632. 51. Buitelaar JK, Sobanski E, Stieglitz RD, Dejonckheere J, Waechter
41. Dobson ET, Strawn JR. Placebo response in pediatric anxiety S, Schauble B. Predictors of placebo response in adults with
disorders: implications for clinical trial design and interpretation. attention-deficit/hyperactivity disorder: data from 2 randomized
J Child Adolesc Psychopharmacol. 2016;26:686-693. trials of osmotic-release oral system methylphenidate. J Clin Psy-
42. Curie A, Yang K, Kirsch I, et al. Placebo responses in genetically chiatry. 2012;73:1097-1102.
determined intellectual disability: a meta- analysis. PLoS One. 52. Sumner CR, Haynes VS, Teicher MH, Newcorn JH. Does placebo
2015;10:e0133316. response differ between objective and subjective measures in
43. Ottenbacher KJ, Cooper HM. Drug treatment of hyperactivity in children with attention-deficit/hyperactivity disorder? Post Grad
children. Dev Med Child Neurol. 1983;25:358-366. Med. 2010;122:52-61.
44. Nikles CJ, Mitchell GK, Del Mar CB, Clavarino A, McNairn N. 53. Park S, Kim BN, Cho SC, et al. Baseline severity of parent-
An n-of-1 trial service in clinical practice: testing the effective- perceived inattentiveness is predictive of the difference between
ness of stimulants for attention-deficit/hyperactivity disorder. subjective and objective methylphenidate responses in children
Pediatrics. 2006;117:2040-2046. with attention-deficit/hyperactivity disorder. J Child Adolesc Psy-
45. Taragin D, Berman S, Zelnik N, Karni A, Tirosh E. Parents’ chopharmacol. 2013;23:410-414.
attitudes toward methylphenidate using n-of-1 trial: a pilot study. 54. Moshe K, Karni A, Tirosh E. Anxiety and methylphenidate in
Atten Defic Hyperact Disord. 2013;5:105-109. attention deficit hyperactivity disorder: a double-blind placebo-
46. Finniss DG, Kaptchuk TJ, Miller F, Benedetti F. Biological, clinical, drug trial. Atten Defic Hyperact Disord. 2012;4:153-158.
and ethical advances of placebo effects. Lancet. 2010;375:686-695. 55. Faber A, Keizer RJ, van den Berg PB, de Jong-van den Berg LT,
47. Cohen D, Consoli A, Bodeau N, et al. Predictors of placebo Tobi H Use of double-blind placebo-controlled N-of-1 trials
response in randomized controlled trials of psychotropic drugs among stimulant-treated youths in the Netherlands: a descriptive
for children and adolescents with internalizing disorders. J Child study. Eur J Clin Pharmacol. 2007;63:57-63.
Adolesc Psychopharmacol. 2010;20:39-47. 56. Scuffham PA, Nikles J, Mitchell GK, et al. Using N-of-1 trials to
48. Nakonezny PA, Mayes TL, Byerly MJ, Emslie GJ. Predicting improve patient management and save costs. J Gen Intern Med.
placebo response in adolescents with major depressive disorder: 2010;25:906-913.