Research Paper

Therapeutic effects of antidepressants for

global improvement and subdomain symptoms of
autism spectrum disorder: a systematic review
and meta-analysis
Shun-Chin Liang, MS*; Cheuk-Kwan Sun, MD, PhD*; Hsin-Yi Fan, MS;
Weilun Chung, MD; Ruu‐Fen Tzang, MD; Kuo-Chuan Hung, MD;
Hsien‐Jane Chiu, MD, PhD; Yu-Shian Cheng, MD†; Pin-Yang Yeh, PhD†

Background: No established pharmacological treatment is available for the core symptoms of autism spectrum disorder (ASD). This
study aimed at investigating the efficacy of antidepressants for the core and associated symptoms of ASD. Methods: We searched
PubMed, Embase, ClinicalKey, Cochrane CENTRAL, ScienceDirect, Web of Science and ClinicalTrials.gov using the keywords “ASD”
and “antidepressants.” We searched from database inception to June 2021 for randomized controlled trials of antidepressant use in
­patients with ASD. We calculated pooled effect sizes based on a random-effects model. Results: Analysis of 16 studies with 899 partici-
pants showed improvements in restricted and repetitive behaviours (effect size = 0.27) and global symptoms (effect size = 1.0) in pa-
tients with ASD taking antidepressants versus those taking placebos (p ≤ 0.01). We found no differences between the 2 groups (p ≥
0.36) in terms of dropout rate (odds ratio [OR] = 1.17) or rate of study discontinuation because of adverse events (OR = 1.05). We also
noted improvements in irritability and hyperactivity in the antidepressant group (Hedges g = 0.33 and 0.22, respectively, both p < 0.03).
Subgroup analyses showed significant effects of medication type (i.e., clomipramine was better than SSRIs) and age (antidepressants
were more effective in adults than in children or adolescents) on both restricted and repetitive behaviours and global improvement (p <
0.05). Meta-regression demonstrated that better therapeutic effects were associated with lower symptom severity and older age.
­Limitations: The small effect sizes and variations in treatment response that we found warrant further study. Conclusion: Our results
supported the effectiveness of antidepressants for global symptoms and symptom subdomains of ASD, with tolerable adverse effects.
Low symptom severity and adulthood were associated with better outcomes.

Introduction tion has approved only 2 agents, aripiprazole and risperi-

done, for the treatment of irritability in ASD;3 the use of most
Autism spectrum disorder (ASD) comprises a range of other medications is off-label, with uncertain therapeutic effi-
neuro­developmental conditions that have their onset in cacy. Identifying other pharmacological agents that are effec-
childhood; ASD affects about 1% of the population and is a tive against the core and associated symptoms of ASD
health burden for families and society.1 Although most ­remains a critical clinical issue.
­people with ASD experience a variety of disabling symptoms Antidepressants have been accepted as a standard treat-
— including irritability, hyperactivity, inappropriate speech, ment for obsessive–compulsive disorder (OCD), which is
social withdrawal and mood symptoms such as anxiety2 — characterized by compulsive and repetitive behaviours.4,5
no established pharmacological treatment is available for the Because OCD and ASD share some symptoms, 6–8 genetic
core symptoms of ASD. The US Food and Drug Administra- risks 9 and neurobiological pathologies,10 most previous

Correspondence to: Yu-Shian Cheng, Department of Psychiatry, Tsyr-Huey Mental Hospital, Kaohsiung Jen-Ai’s Home, No. 509, Fengping
1st Rd., Daliao Dist., Kaohsiung Daliao 831141, Taiwan, n043283@gmail.com; Pin-Yang Yeh, Department of Psychology, College of Medical
and Health Science, Asia University, Taichung, Taiwan; Clinical Psychological Center, Taichung, Taiwan; PsyYPY@asia.edu.tw
*Contributed equally as first authors.
†Contributed equally as corresponding authors.
Submitted Nov. 5, 2021; Revised Mar. 6, 2022; Revised May 12, 2022; Revised Jun. 18, 2022; Accepted Jun. 23, 2022
Cite as: J Psychiatry Neurosci 2022 August 10;47(4). doi: 10.1503/jpn.210191

© 2022 CMA Impact Inc. or its licensors

J Psychiatry Neurosci 2022;47(4) E299

Liang et al.

meta-analyses have focused on the effects of antidepressants or associated symptoms in patients with ASD. We excluded
on restricted and repetitive behaviours in patients with studies that were not clinical trials and studies that were un-
ASD.11–13 In contrast, only 1 meta-analysis has investigated related to antidepressants in ASD.
the effects of antidepressants on global improvement in ASD
symptoms.14 The numbers of trials were also limited in these Data extraction and management
meta-analyses (all less than 10).11–14 Moreover, although the
findings of 3 of the meta-analyses supported the use of anti- Two reviewers (Y.C. and P.Y.) were responsible for data
depressant treatment for restricted and repetitive behav- ­extraction and management; they determined the choice of
iours,11–13 the effect sizes were small (0.22 to 0.24) and none databases and keywords by discussion.
of the indicators of improvement achieved statistical signifi- After potentially eligible articles had been retrieved from
cance. Finally, although the influences of dosage and age on the databases, all references were imported into EndNote
the therapeutic effects of antidepressants in OCD have been (version X8) to identify and remove duplicates. Then, the
well studied in previous meta-analyses,5 the effect of dosage 2 reviewers examined the titles and abstracts of the remaining
on patients with ASD has not been adequately addressed. records to select articles that met the inclusion criteria. Finally,
The rationale for using antidepressants to treat ASD is based the 2 reviewers read the full manuscripts of the selected
on the fact that abnormal levels of serotonin have been found ­studies independently to determine each study’s appropriate-
in certain brain regions — particularly the frontal cortex — in ness for inclusion. We used the κ coefficient to evaluate inter-
patients with ASD.15,16 As well, serotonin-mediated neuropro- rater reliability.21 Disagreements about study eligibility were
tective effects17 and evidence of serotonin dysregulation such resolved via discussion. When data were missing, we con-
as the impaired availability of brain 5-hydroxytryptamine in tacted the corresponding authors for their original data. In
people with ASD15,16,18 further support a physiologic basis for cases of duplicate data, we selected the article with the largest
antidepressant treatment for patients with ASD in general — sample size or the latest publication.
not only for restricted and repetitive behaviours. However, the To assess possible sources of bias, we used the risk-of-bias
effects of antidepressants on other symptom subdomains of assessment tool developed by the Cochrane Collaboration.22
ASD (e.g., inappropriate speech) have not been addressed in
previous meta-analyses. Data synthesis and sensitivity analysis
The present meta-analysis was aimed at investigating the
efficacy of antidepressants for restricted and repetitive be- Changes in outcomes were expressed as effect sizes using
haviours in patients with ASD, as well as their effects on Hedges g and odds ratios. Primary outcomes were improve-
global impairment and other symptom subdomains. We also ment in restricted and repetitive behaviours, improvement in
evaluated the influence of different variables (e.g., age and global symptoms, and differences in dropout rates and inci-
dosage) on therapeutic outcomes with antidepressants and dence of adverse effects. Secondary outcomes were changes
analyzed their potential adverse effects, tolerability and ac- in the severity of anxiety and problem behaviours, including
ceptability in patients with ASD. irritability, social withdrawal, hyperactivity and inappropri-
ate speech. We used Comprehensive Meta-Analysis (version
Methods 2.2.064) to calculate effect sizes. Positive effect sizes indicated
improvements in outcomes among participants who received
Electronic searches and registration antidepressants.
Results were standardized and averaged to produce a sin-
Using the Preferred Reporting Items for Systematic Reviews gle effect size for studies with similar interventions (e.g.,
and Meta-analyses (PRISMA) guidelines,19,20 we systemat­ same medication with different doses). However, if a study
ically searched for randomized controlled trials in the fol- had 2 or more treatment arms with distinctly different phar-
lowing databases from inception to June 2021: PubMed, macologic actions (e.g., binding to different receptors that
Embase, ClinicalKey, Cochrane CENTRAL, ScienceDirect, might substantially affect treatment efficacy), we subdivided
Web of Science and ClinicalTrials.gov. The keywords we shared control (i.e., placebo) groups by the number of treat-
used in the various databases are listed in Appendix 1, ment arms.23 To prevent a reduction in statistical power from
Table S1, available at www.jpn.ca/lookup/doi/10.1503/jpn​.​ small sample sizes, we used a random-effects model to evalu-
210191​/tab-related-content. ate effect sizes, based on the assumption that true effect sizes
We registered the present study with the International were the same in all studies.24,25 In other words, this model
Prospective Register of Systematic Reviews (PROSPERO adjusted for sample-size bias by averaging the distribution of
CRD42021256770). effects across the eligible studies25 so that they carried similar
weights for comparison.24
Study eligibility and definitions We also conducted subgroup analyses using a random-
effects model.24 For continuous variables (e.g., age, percent
Study eligibility criteria were as follows: studies were ran- female and IQ), we used mixed-effects metaregression to
domized controlled trials in patients with ASD; interven- investigate their effects on outcomes in patients with ASD.
tions included antidepressants and comparators; and out- We also calculated Q statistics and used the corresponding
come assessment used behavioural rating scales for primary p values to assess the heterogeneity of effect sizes.

E300 J Psychiatry Neurosci 2022;47(4)

 Antidepressants for symptoms of autism spectrum disorder

We assessed publication bias by inspecting funnel plots in tricyclic antidepressants (e.g., clomipramine, desipramine).
cases of fewer than 10 data sets26 and by performing Egger Of the 16 studies, 15 used placebos as controls and 1 used
tests in cases of 10 or more data sets.27 When we encountered haloperidol.
funnel-plot asymmetry, we imputed the results of potentially Chugani and colleagues investigated the efficacy of
missing studies with Duval and Tweedie’s trim-and-fill 2 doses of a single agent (i.e., buspirone);32 for this study, we
method.28 We conducted sensitivity tests using the leave-one- standardized the data from the 2 interventional groups and
out approach (i.e., removing one study each time and repeat- averaged them to produce a single effect size for comparison
ing the step) to estimate the effect of each study on the over- with the placebo group. For the study by Gordon and col-
all effect size.26 leagues,31 we split the placebo group in half to compare it
with 2 different pharmacological agents (clomipramine and
Results desipramine) because of the reportedly more potent sero­
tonergic effect of clomipramine.45
Study characteristics More than half of the included studies targeted children
and adolescents; 4 (25%) also included an adult popula-
Figure 1 summarizes our identification of eligible studies for tion. Seven data sets from 6 studies provided IQ data. The
the meta-analysis. From the 1103 articles initially retrieved duration of treatment varied from 2 to 48 weeks. Most
from the databases, 1047 were excluded after reviewing their studies were conducted in the United States; 1 was con-
titles and abstracts. Of the 56 full-text articles assessed for eli- ducted in Japan.
gibility, 40 were excluded because they did not meet the in- In the included studies, 2 categories of behavioural rating
clusion criteria (Appendix 1, Table S2). A total of 16 studies scales were adopted for outcome assessment. Some studies
were included in the meta-analysis,29–44 of which 2 were un- used domain-specific questionnaires mainly for restricted
published studies from ClinicalTrials.gov (Table 1).43,44 The and repetitive behaviours, such as the Children’s Yale–Brown
κ coefficient for interrater reliability was 1.0. Obsessive–Compulsive Scale or the National Institute of
The included studies involved a total of 899 participants Mental Health Global Obsessive–Compulsive Scale. Others
with a mean age of 12.57 years (range 2.5–60 years); 18.62% used more general behavioural rating scales that contain dif-
(range 0%−33.3%) were female. Antidepressants were pre- ferent symptom subdomains, including stereotyped behav-
scribed in 457 patients (50.8%), including selective serotonin iours, irritability, social withdrawal, hyperactivity and inap-
reuptake inhibitors (SSRIs; e.g., fluoxetine, citalopram or propriate speech, such as the Aberrant Behavior Checklist,
sertraline), selective serotonin and norepinephrine reuptake the Behavioural Assessment Scale, the Children’s Psychiatric
inhibitors (e.g., venlafaxine), buspirone, mirtazapine and Rating Scale and visual analogue scales.

Records identified through Additional records identified

Identification

database searching through other sources

n = 1088 n = 15
Screening

Records identified
n = 1103
Excluded by title and abstract
n = 1047
Eligibility

Full-text articles assessed

for eligibility
n = 56 Articles excluded n = 40
• No antidepressant treatment n = 17
• Not an RCT n = 7
• Lack of adequate data for analysis n = 5
Included

• Included diagnoses other than ASD n = 8

Studies included in current
• Duplicated sample source n = 3
meta-analysis
n = 16

Figure 1: PRISMA flow diagram for study selection. ASD = autism spectrum disorder; PRISMA = Preferred Reporting Items for Systematic
Reviews and Meta-analyses; RCT = randomized controlled trial.

J Psychiatry Neurosci 2022;47(4) E301

Liang et al.

Table 1: Characteristics of studies in the meta-analysis (part 1 of 3)

Intervention and comparator Mean
Diagnosis Exclusion Duration, age, yr Female,
Study (criteria) criteria Design Medication DDD n wk Outcome IQ CGI-S (range) % Country

SSRI versus placebo

Herscu ASD AS, PDD-NOS, RCT Fluoxetine NA 78 14 CY-BOCS- NA 4.87 9 14.55 United
et al.30 (DSM- Rett syndrome, 2–18 mg/d PDD (5–17) States
(2020) IV-TR) childhood Placebo None 80
disintegrative
disorder, active
seizure disorder;
taking
psychotropic
medication; high
levels of
aggression or
self-injurious
behaviour
Potter ASD Fragile X RCT Sertraline NA 32 24 VAS 49.6 NA 4.03 20.7 United
et al.29 (DSM-5) syndrome full 2.5–5 mg/d (anxiety/ (2–6) States
(2019) mutation; Placebo None 26 obsessive–
serious compulsive
comorbid behaviour,
medical disorder aggression/
affecting brain hyperarousal/
function hyperactivity,
language/
commun-
ication);
CGI-I; SRS;
PARS-R
Reddihough ASD Rett syndrome, RCT Fluoxetine NA 75 16 CY-BOCS- NA NA 11.2 15 Australia
et al.41 (DSM- childhood 4–30 mg/d PDD; CGI-I; (7.5–18)
(2019) IV-TR) disintegrative Placebo None 71 ABC (irritability,
disorder, hyperactivity,
schizophrenia social
or major withdrawal,
depression; inappropriate
taking speech);
psychotropic Spence
medications; children anxiety
comorbid
medical
conditions
Sikich Autism AS, PDD-NOS, RCT Fluoxetine NA 8 48 ABC (irritability) NA NA 3.62 0 United
et al.44 (DSM-III-R) Rett syndrome, 2–20 mg/d (2.5–4.8) States
(2014)* childhood Placebo None 10
disintegrative
disorder
Hollander ASD None RCT Fluoxetine 3.24 22 12 Y-BOCS; CGI-I 103 4.39 34.31 31 United
et al.39 (DSM-IV); 20–80 mg/d (18–60) States
(2012) CGI-S ≥ 4 Placebo None 15
King et al.38 AD, AS, Rett syndrome RCT Citalopram NA 73 12 CY-BOCS- NA 4.94 9.36 14.1 United
(2009) PDD-NOS or childhood 10–20 mg/d PDD; ABC (5–17) States
(DSM- disintegrative Placebo None 76 (irritability,
IV-TR); disorder; hyperactivity,
CGI-S ≥ 4 seizure within social
the past 6 mo, withdrawal,
weight < 15 kg; inappropriate
bipolar disorder speech)
or manic episode
Hollander ASD Psychotic RCT/ Fluoxetine NA 19 8 CY-BOCS; 63.7 4.61 8.18 23.1 United
et al.36 (DSM- disorders, cross- 4.8–20 mg/d CGI-I (5–17) States
(2005) IV-TR) seizures; over Placebo None 20
clinically
significant
medical illness
Sugie Autism Underlying RCT/ Fluvoxamine NA 18 12 BAS (emotional NA NA 5.33 26.7 Japan
et al.37 (DSM-IV) diseases, such cross- 1–3 mg/kg/d instability, (3–8.5)
(2005) as chromosomal over Placebo None 18 hyperactivity,
aberration; social
congenital rubella withdrawal,
syndrome, inappropriate
apparent speech)
neurologic
deficits

E302 J Psychiatry Neurosci 2022;47(4)

 Antidepressants for symptoms of autism spectrum disorder

Table 1: Characteristics of studies in the meta-analysis (part 2 of 3)

Intervention and comparator Mean
Diagnosis Exclusion Duration, age, yr Female,
Study (criteria) criteria Design Medication DDD n wk Outcome IQ CGI-S (range) % Country

McDougle AD Schizophrenia, RCT Fluvoxamine 2.77 15 12 Y-BOCS; 79.9 4.12 30.1 10 United
et al.34 (DSM-III-R); psychotic 50–300 mg/d CGI-I; Brown (18–53) States
(1996) CGI-S ≥ 4 symptoms, Placebo None 15 Aggression
illicit substances Scale
within the
previous 6 mo;
notable medical
condition,
including
seizure disorder;
pregnancy
SNRI versus placebo
Carminati PDD Epilepsy or any RCT Venlafaxine 0.185 6 8 ABC NA NA 22 15.4 Switzer-
et al.40 (ICD-10); indication 18.75 mg/d (stereotype); (18–30) land
(2016) mild to against CGI-I;
Placebo None 7
profound ID somatic– ABC (irritability, Median:
psychotropic hyperactivity, 19
treatments; social (19–31)
pregnancy withdrawal,
inappropriate
speech)
Tricyclic antidepressant versus placebo
Remington Autism NA RCT/ Clomipramine NA 36 7 ABC (repetitive NA NA 16.3 16.7 Canada
et al.35 (DSM-IV) cross- 100–150 mg/d behaviours, (10–36)
(2001) over Placebo None 36 irritability,
hyperactivity,
social
withdrawal,
inappropriate
speech);
CARS
Gordon AD Significant RCT/ Clomipramine NA 12 10 Modified 57.1 NA 11.7 33.3 United
et al.33 (DSM-III-R) problems, cross- 25–250 mg/d CPRS OCD (6–23) States
(1993) including over Placebo None 12 subscale;
seizures CPRS autism-
relevant
subscale
Gordon AD Significant RCT/ Clomipramine NA 7 5 NIMH-GOCS; NA NA 9.6 28.6 United
et al.31 (DSM-III-R) medical cross- 25–250 mg/d CPRS (6–18) States
(1992) problems, over Desipramine NA 7
including 25-250 mg/d
seizures
Placebo None 7
Other antidepressant versus placebo
McDougle AD, AS, Rett syndrome, RCT Mirtazapine NA 20 10 PARS NA NA 11 20 United
et al.43 PDD-NOS childhood 7.5–45 mg/d (5–17) States
(2018)* (DSM-IV) integrative Placebo 10
disorder,
OCD,
post-traumatic
stress disorder,
major mood
disorder,
psychotic
disorder,
substance use
disorder
Chugani ASD Neurologic RCT Buspirone NA 54 24 CY-BOCS; 64.1 NA 3.62 17.5 United
et al.32 (DSM- disorders, 5 mg/d ADOS-CTS; (2–6) States
(2016) IV-TR) phenylketonuria, ABC (social
Buspirone NA 55
tuberous withdrawal,
10 mg/d
sclerosis inappropriate
complex, Placebo None 57 speech);
Rett syndrome, anxiety
Fragile X composite
syndrome; score (ABC
Down syndrome; irritability +
traumatic Leiter emotion
brain injury or regulation)
other medical
or behavioural
problems

J Psychiatry Neurosci 2022;47(4) E303

Liang et al.

Table 1: Characteristics of studies in the meta-analysis (part 3 of 3)

Intervention and comparator Mean
Diagnosis Exclusion Duration, age, yr Female,
Study (criteria) criteria Design Medication DDD n wk Outcome IQ CGI-S (range) % Country

Antidepressant versus antipsychotic

Sanchez AD Identifiable RCT/ Clomipramine NA 8 4.5 CGI-S; CPRS NA NA 5.6 12.5 United
et al.42 (DSM-III-R; causes of cross- 2.8–4.4 (hyperactivity, (2.3–7.8) States
(1995) infantile autism; seizures over mg/kg/d speech
autism or other deviance)
Haloperidol NA 8
(DSM-III) systemic
0.02–0.05
disease
mg/kg/d

ABC = Aberrant Behavior Checklist; AD = autistic disorder; ADOS-CTS = Autism Diagnostic Observation Schedule Composite Total Score; AS = Asperger syndrome; ASD = autism
spectrum disorder; BAS = Behavioural Assessment Scale; CARS = Childhood Autism Rating Scale; CGI-I = Clinical Global Impression–Improvement; CGI-S = Clinical Global Impression–
Severity; CPRS = Children’s Psychiatric Rating Scale; CY-BOCS = Children’s Yale–Brown Obsessive–Compulsive Scale; CY-BOCS-PDD = Children’s Yale–Brown Obsessive–
Compulsive Scale modified for pervasive developmental disorder; DDD = defined daily dose; DSM-III = Diagnostic and Statistical Manual of Mental Disorders, 3rd edition; DSM-III-R =
Diagnostic and Statistical Manual of Mental Disorders, 3rd edition, revised; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, 4th edition; DSM-IV-TR = Diagnostic and
Statistical Manual of Mental Disorders, 4th edition, text revision; DSM-5 = Diagnostic and Statistical Manual of Mental Disorders, 5th edition; ICD-10 = International Statistical
Classification of Diseases and Related Health Problems, 10th revision; ID = intellectual disability; NA = not available; NIMH-GOCS = National Institute of Mental Health Global Obsessive–
Compulsive Scale; OCD = obsessive–compulsive disorder; PARS = Pediatric Anxiety Rating Scale; PARS-R = Pediatric Anxiety Rating Scale revised; PDD = pervasive developmental
disorder; PDD-NOS = pervasive developmental disorder, not otherwise specified; RCT = randomized controlled trial; SNRI = serotonin and norepinephrine reuptake inhibitor; SRS =
Social Responsiveness Scale; SSRI = selective serotonin reuptake inhibitor; VAS = visual analogue scale; Y-BOCS = Yale–Brown Obsessive–Compulsive Scale.
*Study identified from ClinicalTrials.gov.

In terms of risk of bias, most studies were of fair quality. suicidal ideation, sedation (or lethargy), self-injury or in-
We found that 75.8% (81/112), 18.8% (21/112) and 5.4% somnia (all p > 0.22).
(6/112) of the included studies had an overall low, unclear With respect to secondary outcomes, 7 of the included
and high risk of bias, respectively (16 studies × 7 categories ­studies provided data on irritability, 7 on social withdrawal, 7
for the risk of bias tool = 112). Two studies received financial on hyperactivity, 7 on inappropriate speech and 4 on anxiety.
support from pharmaceutical companies (Figure 2). Among patients with ASD who received antidepressants, we
found significantly better improvement in irritability (Hedges
Quantitative data synthesis g = 0.33, p < 0.03) and hyperactivity (Hedges g = 0.22, p < 0.03),
but not in inappropriate speech (Hedges g = 0.24, p = 0.1), social
For restricted and repetitive behaviours, one of our primary withdrawal (Hedges g = 0.03, p = 0.88) or anxiety (Hedges g =
outcomes, a meta-analysis of 13 data sets from 12 studies 0.44, p = 0.31). Leave-one-out sensitivity analysis of the effect
found significant improvement in patients with ASD who re- sizes for the secondary outcomes found that the main effects
ceived antidepressants compared to controls (p < 0.01; were not driven by any single study. The risks of publication
Figure 3A). The effect size was strong in the leave-one-out bias reflected by funnel plot asymmetry for these outcomes are
sensitivity analysis (p < 0.01), indicating that no single study shown in Appendix 1, Figure S1. With the random-effects
had a substantial influence on the overall effect size. The re- model, the trim-and-fill method revealed potentially missing
sult of the Egger test was not significant (p = 0.053), suggest- studies on the left side of the plot: 2 for irritability, yielding an
ing a low risk of publication bias. adjusted effect size of 0.27 (0.03 to 0.50); 2 for social withdrawal,
For global improvement, 11 data sets from 11 studies yielding an adjusted effect size of −0.1 (−0.42 to 0.22); 2 for
showed a significant improvement in global symptoms in pa- ­hyperactivity, yielding an adjusted effect size of 0.19 (0.02 to
tients with ASD who received antidepressants compared to 0.36); 0 for inappropriate speech, yielding an adjusted effect
controls (p = 0.01; Figure 3B). The effect size was strong in the size of 0.24 (−0.05 to 0.54); and 0 for anxiety, yielding an ad-
leave-one-out sensitivity analysis (p < 0.01), indicating that no justed effect size of 0.44 (−0.41 to 1.29).
single study had a substantial influence on the overall effect
size. The result of the Egger test was not significant (p = 0.10), Subgroup analysis and meta-regression
suggesting a low risk of publication bias.
Based on 14 data sets from 13 studies, we found no signifi- Subgroup analyses showed significant effects of type of medi-
cant differences between patients with ASD who received anti- cation (i.e., SSRI v. clomipramine) and age group (i.e., chil-
depressants and controls with respect to dropout rate or rate of dren and adolescents v. adults) on restricted and repetitive
study discontinuation because of adverse events (Figure 4). The behaviours, as well as on global improvement (all p < 0.04;
leave-one-out sensitivity analysis of effect size for both dropout Table 2). Using a mixed-effects model, meta-regression dem-
rate and study discontinuation demonstrated nonsignificant ef- onstrated significant associations between the therapeutic
fects of individual studies on the overall outcomes. The results ­effects of antidepressants for the symptoms of ASD and the
of the Egger tests were nonsignificant for dropout rate and ­Severity aspect of the Clinical Global Impression scale (regres-
study discontinuation, suggesting a low risk of publication bias. sion coefficient −0.95 for restricted and repetitive behaviours,
We also found no significant differences between the anti- p = 0.02). We also detected a positive correlation between the
depressant and control groups in the incidence of overall and effects of antidepressants and age for global improvement (re-
specific adverse events, including irritability (or activation), gression coefficient 0.08, p = 0.02; Appendix 1, Table S3).

E304 J Psychiatry Neurosci 2022;47(4)

 Antidepressants for symptoms of autism spectrum disorder

minimal effectiveness of antidepressants (effect size 0.22, p <

0.05), and this finding became nonsignificant after adjustment
Low risk of bias for publication bias.12

Allocation concealment
Unclear risk of bias Our study demonstrated that antidepressants had signifi-

Random sequence
cant effectiveness for the treatment of restricted and repeti-

Performance bias
High risk of bias
tive behaviours in patients with ASD compared to placebo,

Reporting bias
Detection bias

Attrition bias
without evidence of publication bias. Still, our small effect

Other bias
size (0.27) was similar to that reported in previous meta-
analyses (0.22 to 0.24). One reason that our findings achieved
significance may have been that we included more studies
†
Herscu et al.30 (2020) (n = 16), together with narrower confidence intervals com-
‡ pared to those previously reported.12 As well, although a pre-
Potter et al.29 (2019) vious meta-analysis46 found a superior treatment effect with
§
clomipramine compared to SSRIs for pediatric OCD, none of
Reddihough et al.41 (2019)
§
the previous meta-analyses in patients with ASD included
Carminati et al.40 (2016) trials for clomipramine.11–13 One of the interesting findings of
§ our subgroup analysis was the higher efficacy of clomip-
Chugani et al.32 (2016) ramine compared to SSRIs, highlighting a possible contribu-
§
Hollander et al.39 (2012) tion of clomipramine to the higher efficacy of antidepressants
§ in the present study compared to previous meta-analyses.
King et al.38 (2009) Despite our positive findings for antidepressants to treat
§
restricted and repetitive behaviours in patients with ASD, the
Hollander et al.36 (2005)
§
effect size was still very small. Previous studies have re-
Sugie et al.37 (2005) ported that antidepressant treatments were less effective in
§ adolescents or children with ASD than in adults.3,47 This find-
Remington et al.35 (2001) ing is further supported by the results of our subgroup analy-
§
McDougle et al.34 (1996) ses, which showed that adults had a better response to anti-
† depressants than adolescents and children. Together with
Sanchez et al.42 (1995) evidence demonstrating that drug susceptibility in the central
§
nervous system varies with age,48 our results support the idea
Gordon et al.33 (1993)
§
that age might be an important factor influencing the efficacy
Gordon et al.31 (1992) of antidepressants for restricted and repetitive behaviours in
§ patients with ASD.
McDougle et al.43 (2018)*
In addition to age, our finding of a negative association be-
§
Sikich et al.44 (2014)* tween symptom severity and antidepressant efficacy implies
that antidepressants may be more effective for patients with
less severe ASD. Experts have consistently suggested that
Figure 2: Risk of bias in included studies. *Study identified from more severe forms of ASD may be less responsive to SSRI
ClinicalTrials.gov. †Study received financial support from pharma- treatment and more susceptible to the adverse effects of
ceutical companies. ‡Authors received financial support from medi­cations.49 Our study is the first meta-analysis to support
­pharmaceutical companies (e.g., consultation). §Neither study nor this concept. This idea is also reinforced by the findings of a
authors received financial support from pharmaceutical companies. study that showed a good response to SSRIs in a subgroup of
patients with high-functioning ASD and familial major affec-
tive disorders who demonstrated a relatively high level of
Discussion anxiety.50 Together with evidence supporting the use of anti-
psychotics for severe behavioural symptoms in patients with
To the best of our knowledge, this was the first meta-analysis ASD,51–53 our finding might suggest the need to tailor medica-
to investigate the efficacy of antidepressants in patients with tion treatments for different subgroups of patients with
ASD to identify important variables that might influence ASD.48,49 Further studies are needed to investigate the correla-
therapeutic outcomes. tion between severity of ASD and response to SSRIs.
Only a few meta-analyses with limited numbers of trials For the therapeutic effect of antidepressants on global im-
(all less than 10) have studied the effects of antidepressants provement of the symptoms of autism, the present study
for symptoms of ASD,11–14 and they did not address important found that antidepressant treatment was significantly better
factors that might affect treatment outcomes. As well, most than placebo, with a large effect size (1.0) in 10 trials, similar to
previous meta-analyses have focused on the effects of anti­ a study published by the Cochrane Collaboration (Hedges g =
depressants on restricted and repetitive behaviours in 0.89, p < 0.01).14 In the studies included in our meta-analysis,
­patients with ASD, and have failed to show significant thera- the efficacy of antidepressants for global improvement cor-
peutic benefits in this setting.11,13 One meta-analysis showed related well with improvements in restricted and repetitive

J Psychiatry Neurosci 2022;47(4) E305

Liang et al.

A Stereotyped behaviours
Study Hedges g 95% CI p value Hedges g (95% CI)
SSRI
Herscu et al.30 (2020), fluoxetine 0.17 (–0.15 to 0.48) 0.30
Reddihough et al.41 (2019), fluoxetine 0.35 (–0.02 to 0.73) 0.07
Hollander et al.39 (2012), fluoxetine 0.38 (–0.31 to 1.06) 0.28
Hollander et al.36 (2005), fluoxetine 0.22 (–0.40 to 0.83) 0.49
Potter et al.29 (2019), sertraline 0.11 (–0.47 to 0.68) 0.72
King et al.38 (2009), citalopram 0.03 (–0.29 to 0.35) 0.84
McDougle et al.34 (1996), fluvoxamine 1.04 (0.29 to 1.78) 0.01
Subtotal 0.24 (0.04 to 0.44) 0.02
SNRI
Carminati et al. (2016), venlafaxine
40
0.19 (–0.83 to 1.21) 0.72
Subtotal 0.19 (–0.86 to 1.24) 0.72
Tricyclic antidepressant
Remington et al.35 (2001), clomipramine 0.08 (–0.38 to 0.54) 0.73
Gordon et al.33 (1993), clomipramine 0.90 (0.08 to 1.71) 0.03
Gordon et al.31 (1992), clomipramine 1.06 (–0.15 to 2.27) 0.09
Gordon et al.31 (1992), desipramine 0.17 (–1.05 to 1.40) 0.78
Subtotal 0.37 (–0.03 to 0.77) 0.07
Other
Chugani et al.32 (2016), buspirone 0.33 (–0.12 to 0.78) 0.15
Subtotal 0.33 (–0.20 to 0.86) 0.22
Total 0.27 (0.10 to 0.44) 0.002

–3.00 –1.50 0.00 1.50 3.00

Placebo Antidepressant

B Global improvement
Study Hedges g 95% CI p value Hedges g (95% CI)
SSRI
Reddihough et al.41 (2019), fluoxetine 0.14 (–0.23 to 0.52) 0.46
Hollander et al.39 (2012), fluoxetine 0.79 (0.08 to 1.50) 0.03
Hollander et al.36 (2005), fluoxetine –0.31 (–0.93 to 0.31) 0.33
Potter et al.29 (2019), sertraline –0.09 (–0.65 to 0.48) 0.77
McDougle et al.34 (1996), fluvoxamine 6.80 (4.95 to 8.66) 0.00
Subtotal 1.00 (–0.01 to 2.02) 0.05
SNRI
Carminati et al. (2016), venlafaxine
40
0.07 (–0.94 to 1.09) 0.89
Subtotal 0.07 (–2.26 to 2.40) 0.95
Tricyclic antidepressant
Remington et al.35 (2001), clomipramine 0.21 (–0.25 to 0.67) 0.37
Gordon et al.33 (1993), clomipramine 1.02 (0.20 to 1.85) 0.02
Gordon et al.31 (1992), clomipramine 3.67 (1.77 to 5.57) 0.00
Gordon et al.31 (1992), desipramine 0.71 (–0.55 to 1.98) 0.27
Subtotal 1.17 (–0.02 to 2.36) 0.06
Other
Chugani et al.32 (2016), buspirone 1.17 (0.69 to 1.66) 0.00
Subtotal 1.17 (–0.98 to 3.33) 0.29

Total 1.00 (0.30 to 1.69) 0.01

–9.00 –4.50 0.00 4.50 9.00
Placebo Antidepressant

Figure 3: Forest plots comparing differences in effect sizes of (A) stereotyped behaviours and (B) global improvement between antidepres-
sant and control groups. CI = confidence interval; SNRI = serotonin and norepinephrine reuptake inhibitor; SSRI = selective serotonin reup-
take inhibitor.

E306 J Psychiatry Neurosci 2022;47(4)

 Antidepressants for symptoms of autism spectrum disorder

A Dropout rate
Study OR 95% CI p value OR (95% CI)
SSRI
Herscu et al.30 (2020), fluoxetine 1.50 (0.73 to 3.11) 0.27
Reddihough et al.41 (2019), fluoxetine 1.40 (0.74 to 2.66) 0.31
Sikich et al.44 (2014), fluoxetine 0.83 (0.17 to 4.01) 0.82
Potter et al.29 (2019), sertraline 1.30 (0.38 to 4.45) 0.68
King et al.38 (2009), citalopram 0.96 (0.42 to 2.21) 0.93
Sugie et al.37 (2005), fluvoxamine 1.00 (0.06 to 17.25) 1.00

Subtotal 1.26 (0.87 to 1.85) 0.23

SNRI
Carminati et al.40 (2016), venlafaxine 1.17 (0.06 to 22.94) 0.92
Subtotal 1.17 (0.06 to 22.94) 0.92
Tricyclic antidepressant
Remington et al.35 (2001), clomipramine 0.83 (0.20 to 3.44) 0.80
Gordon et al.33 (1993), clomipramine 1.00 (0.06 to 17.62) 1.00
Gordon et al.31 (1992), desipramine 2.33 (0.09 to 62.68) 0.61
Subtotal 0.98 (0.30 to 3.21) 0.98
Other
McDougle et al.43 (2018), mirtazapine 1.54 (0.06 to 41.08) 0.80
Chugani et al.32 (2016), buspirone 0.73 (0.26 to 2.05) 0.54
Subtotal 0.78 (0.29 to 2.09) 0.62
Total 1.17 (0.83 to 1.64) 0.36

0.01 0.1 1 10 100

Placebo Antidepressant

B Rate of study discontinuation because of adverse events

Study OR 95% CI p value OR (95% CI)

SSRI
Herscu et al.30 (2020), fluoxetine 1.40 (0.43 to 4.60) 0.58
Reddihough et al. (2019), fluoxetine
41
1.18 (0.31 to 4.58) 0.81
King et al.38 (2009), citalopram 0.75 (0.26 to 2.11) 0.58

Subtotal 1.03 (0.52 to 2.02) 0.94

Tricyclic antidepressant
Remington et al.35 (2001), clomipramine 7.54 (0.38 to 151.26) 0.19

Subtotal 7.54 (0.38 to 151.26) 0.19

Other
McDougle et al.43 (2018), mirtazapine 1.54 (0.06 to 41.08) 0.80
Chugani et al.32 (2016), buspirone 0.52 (0.13 to 2.20) 0.38
Subtotal 0.62 (0.17 to 2.32) 0.48

Total 1.05 (0.40 to 2.76) 0.92

0.01 0.1 1 10 100

Placebo Antidepressant

Figure 4: Forest plots comparing odds ratios for (A) dropout rate and (B) rate of study discontinuation because of adverse events between
­antidepressant and control groups. CI = confidence interval; OR = odds ratio; SNRI = serotonin and norepinephrine reuptake inhibitor; SSRI =
selective serotonin reuptake inhibitor.

J Psychiatry Neurosci 2022;47(4) E307

Liang et al.

Table 2: Difference in repeated behaviours and global improvement among participants with autism spectrum disorder
taking antidepressants versus placebo — subgroup analysis

Characteristic No. of trials Hedges g (95% CI)* Z score Cochran Q† p value‡

Stereotyped behaviours
Drug intervention
SSRI 7 0.25 (0.03 to 0.48) 2.21§ 0.55 0.005
Clomipramine (TCA) 3 0.45 (–0.01 to 0.90) 1.93
Age group
Children and adolescents 8 0.21 (0.09 to 0.38) 2.64§ 1.94 0.001
Adults 3 0.58 (0.11 to 1.04) 2.29§
Global improvement
Drug intervention
SSRI 5 1.04 (–0.04 to 2.12) 1.90§ 0.20 0.007
Clomipramine (TCA) 3 1.45 (0.02 to 2.89) 1.98§
Age group
Children and adolescents 6 0.59 (–0.09 to 1.27) 1.69 1.49 0.040
Adults 3 2.41 (–0.44 to 5.25) 1.66
CI = confidence interval; SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant.
*Random-effects model.
†For heterogeneity assessment based on random-effects analysis.
‡Significance of the difference between effect sizes in the subgroups.
§p < 0.05.

behaviours in patients with ASD, highlighting an overall im- was more commonly associated with fluoxetine, citalopram,
provement attributable to the alleviation of ASD symptom sertraline and buspirone; sedation occurred more frequently
subdomains. Moreover, our secondary analyses showed that in patients taking mirtazapine (Appendix 1, Table S4).
antidepressants were effective against irritability and hyperac- Never­theless, we found no significant differences between
tivity in patients with ASD, suggesting that the overall thera- the antidepressant and placebo groups in terms of percent-
peutic benefits of antidepressants might come from collective age of overall adverse effects or incidence of irritability, in-
improvements in different symptom subdomains. Similar to somnia, lethargy, suicidal ideation or self-injurious behav-
findings for restricted and repetitive behaviours, the results of iours. We also found no significant differences between the
our subgroup analysis of global improvement of ASD symp- 2 groups in terms of dropout rate or rate of study discon­
toms demonstrated positive associations with older age (v. tinuation because of adverse events. One study using clo-
younger age) and clomipramine (v. SSRIs). Moreover, our mipramine (a tricyclic antidepressant) reported severe
metaregression analysis also supported the finding that older tachycardia and grand mal seizure in a 7-year-old girl.33
age was correlated with better global effects of antidepressants Furthermore, a previous review highlighted the possibility
in patients with ASD. of sudden death related to the use of tricyclic antidepres-
Previous evidence supporting the effectiveness of anti­ sants in children,56 raising serious concerns about the safety
depressants to treat irritability and hyperactivity remains of prescribing them for children. Although our study
controversial, because adverse effects (e.g., agitation) have showed satisfactory tolerability of adverse effects related to
been reported more frequently in trials targeting adolescents antidepressant use compared to placebo, treatment strat­
and children with ASD.54,55 The results of our secondary egies should be individualized, taking into account the
analy­ses showed equal effectiveness for antidepressants in all ­potential adverse effects of different agents.
3 behavioural subdomains of ASD (i.e., irritability, hyper­
activity and restricted and repetitive behaviours). Although Limitations
antidepressants were originally designed to improve mood,
none of the studies included in the present meta-analysis in- The present meta-analysis had some limitations. First, al-
vestigated their therapeutic effects on depressive symptoms, though we included more trials than previous meta-analyses,
and only 4 included symptoms of anxiety. Our results our total sample size (n = 899) and effect sizes were still too
showed that antidepressants might be more effective than small to reach robust conclusions for most of our outcome
placebos for improving anxiety, despite their failure to reach measures. In particular, compared to our primary out-
statistical significance. Given the small sample size in our comes, the sample sizes were much smaller for our second-
analysis for anxiety, further studies are needed to address ary outcomes, and we found evidence of publication bias;
this issue in patients with ASD. these findings should be interpreted with caution. Second,
The findings of the present study demonstrated that the the heterogeneity of the included studies — including dif-
most common adverse effects on the central nervous system ferences in choice of antidepressants, dosage and age
were lethargy, irritability, activation and insomnia. Insomnia groups — might limit the generalizability of our results.

E308 J Psychiatry Neurosci 2022;47(4)

 Antidepressants for symptoms of autism spectrum disorder

Still, we performed subgroup analyses and meta-regression Contributors: S. Lian, H. Fan, W. Chung, R. Tzang, K. Hung, H. Chiu
and Y. Cheng designed the study. C. Sun and P. Yeh analyzed the
analyses to help identify important factors that might have
data. S. Liang, C. Sun, Y. Cheng and P. Yeh wrote the article, which
influenced treatment response. Third, we could provide H. Fan, W. Chung, R. Tzang, K. Hung and H. Chiu reviewed. All
only effect sizes because of limited availability of data; we authors approved the final version to be published, agree to be ac-
were unable to provide odds ratios or numbers needed to countable for all aspects of the work and can certify that no other
treat, which would have offered clearer information for ref- individuals not listed as authors have made substantial contribu-
tions to the paper.
erence in clinical practice. Fourth, 1 study allowed the use
of other psychotropic medications (e.g., zuclopenthixol and Disclaimer: The contents of this article are the authors’ sole responsi-
clonazepam) in the study and control groups.40 Because the bility and do not necessarily represent the views of their institutions.
doses of those psychotropics in the placebo arm were Content licence: This is an Open Access article distributed in accord­
higher than those in the study arm by the end of the 56-day ance with the terms of the Creative Commons Attribution (CC BY-NC-
trial, it was difficult to determine the effectiveness of the ND 4.0) licence, which permits use, distribution and reproduction in
any medium, provided that the original publication is properly cited,
antidepressants versus placebo based on the results of that the use is noncommercial (i.e., research or educational use), and no
study. Nevertheless, our sensitivity test did not reveal a modifications or adaptations are made. See: https://creativecommons.
substantial influence of any single study, including that org/licenses/by-nc-nd/4.0/
particular trial,40 on the overall effect size. Fifth, because in-
formation was limited about defined daily doses for chil- References
dren and adolescents, we were unable to investigate the ef-
fects of dosage on the efficacy of antidepressants for the 1. Lyall K, Croen L, Daniels J, et al. The changing epidemiology of
autism spectrum disorders. Annu Rev Public Health 2017;38:81-102.
symptoms of ASD in this age group. Finally, the pooling of
2. Maneeton N, Maneeton B, Putthisri S, et al. Aripiprazole in acute
results from different assessment tools for behavioural out- treatment of children and adolescents with autism spectrum disorder:
comes is an inherent down side of meta-analyses in the a systematic review and meta-analysis. Neuropsychiatr Dis Treat
psychiatric field. Nevertheless, most of our included 2018;14:3063-72.
­studies used similar tools for outcome measurements (e.g., 3. Goel R, Hong JS, Findling RL, et al. An update on pharmacotherapy
of autism spectrum disorder in children and adolescents. Int Rev
Children’s Yale–Brown Obsessive–Compulsive Scale or Psychiatry 2018;30:78-95.
Aberrant Behavior Checklist), so that the confounding 4. Soomro GM, Altman D, Rajagopal S, et al. Selective serotonin re-
­effect was relatively minor. uptake inhibitors (SSRIs) versus placebo for obsessive compulsive
disorder (OCD). Cochrane Database Syst Rev 2008;2008:CD001765.
Conclusion 5. Bloch MH, McGuire J, Landeros-Weisenberger A, et al. Meta-analysis
of the dose-response relationship of SSRI in obsessive-compulsive
disorder. Mol Psychiatry 2010;15:850-5.
Our results support the effectiveness of antidepressants in 6. Zandt F, Prior M, Kyrios M. Similarities and differences between
treating the global symptoms of ASD, as well as symptom children and adolescents with autism spectrum disorder and those
subdomains, including irritability and hyperactivity, with with obsessive compulsive disorder: executive functioning and re-
petitive behaviour. Autism 2009;13:43-57.
tolerable adverse effect profiles. Treatment response might be
7. Zandt F, Prior M, Kyrios M. Repetitive behaviour in children with
better achieved in patients with less severe overall ASD high functioning autism and obsessive compulsive disorder. J Autism
symptoms and in the adult population. However, given the Dev Disord 2007;37:251-9.
small effect sizes we found for each symptom subdomain, 8. McDougle CJ, Kresch LE, Goodman WK, et al. A case-controlled
wide variations in treatment response in different subgroups study of repetitive thoughts and behavior in adults with autistic
disorder and obsessive-compulsive disorder. Am J Psychiatry 1995;​
and publication bias associated with outcomes for irritability
152:772-7.
and hyperactivity, we suggest judicious and individually tai- 9. Micali N, Chakrabarti S, Fombonne E. The broad autism phenotype:
lored use of antidepressants for patients with ASD. findings from an epidemiological survey. Autism 2004;8:21-37.
10. Hollander E, Kim S, Braun A, et al. Cross-cutting issues and future
Affiliations: From the Department of Management Centre, Jianan directions for the OCD spectrum. Psychiatry Res 2009;170:3-6.
Psychiatric Centre, Ministry Of Health and Welfare, Tainan, Taiwan
11. Zhou MS, Nasir M, Farhat LC, et al. Meta-analysis: pharmacologic
(Liang); the Centre for General Education, University of Kun Shan, treatment of restricted and repetitive behaviors in autism spec-
Tainan, Taiwan (Liang); the Department of Emergency Medicine, trum disorders. J Am Acad Child Adolesc Psychiatry 2021;60:35-45.
E-Da Hospital, Kaohsiung, Taiwan (Sun); the School of Medicine for
12. Carrasco M, Volkmar FR, Bloch MH. Pharmacologic treatment of
International Students, College of Medicine, I-Shou University,
repetitive behaviors in autism spectrum disorders: evidence of
Kaohsiung, Taiwan (Sun); the Department of Psychiatry, Tsyr-Huey
publication bias. Pediatrics 2012;129:e1301-10.
Mental Hospital, Kaohsiung Jen-Ai’s Home, Taiwan (Fan, Chung,
Cheng); the Department of Psychiatry, Mackay Memorial Hospital, 13. Yu Y, Chaulagain A, Pedersen SA, et al. Pharmacotherapy of
restricted/repetitive behavior in autism spectrum disorder: a
Taipei, Taiwan (Tzang); the Department of Anesthesiology, Chi Mei
systematic review and meta-analysis. BMC Psychiatry 2020;20:121.
Medical Centre, Tainan, Taiwan (Hung); the Taoyuan Psychiatric Cen-
tre, Ministry of Health and Welfare, Taoyuan City, Taiwan (Chiu); the 14. Williams K, Brignell A, Randall M, et al. Selective serotonin
Institute of Hospital and Health Care Administration, National Yang- reuptake inhibitors (SSRIs) for autism spectrum disorders (ASD).
Cochrane Database Syst Rev 2013;CD004677.
Ming University, Taipei City, Taiwan (Chiu); the Institute of Biomedical
Sciences, National Sun Yat-Sen University, Kaohsiung City, Taiwan 15. Boccuto L, Chen CF, Pittman AR, et al. Decreased tryptophan
(Cheng); the Department of Psychology, College of Medical and Health ­m etabolism in patients with autism spectrum disorders. Mol
Science, Asia University, Taichung, Taiwan (Yeh); Clinical Psychology ­Autism 2013;4:16.
Center, Asia University Hospital, Tai­chung, Taiwan (Yeh). 16. Chandana SR, Behen ME, Juhász C, et al. Significance of abnormal-
ities in developmental trajectory and asymmetry of cortical sero-
Competing interests: None declared. tonin synthesis in autism. Int J Dev Neurosci 2005;23:171-82.

J Psychiatry Neurosci 2022;47(4) E309

 Liang et al.

17. Liu MT, Kuan Y-H, Wang J, et al. 5-HT4 receptor-mediated neuro- 38. King BH, Hollander E, Sikich L, et al. Lack of efficacy of citalo-
protection and neurogenesis in the enteric nervous system of adult pram in children with autism spectrum disorders and high levels
mice. J Neurosci 2009;29:9683-99. of repetitive behavior: citalopram ineffective in children with au-
18. Makkonen I, Riikonen R, Kokki H, et al. Serotonin and dopamine tism. Arch Gen Psychiatry 2009;66:583-90.
transporter binding in children with autism determined by SPECT. 39. Hollander E, Soorya L, Chaplin W, et al. A double-blind
Dev Med Child Neurol 2008;50:593-7. placebo-controlled trial of fluoxetine for repetitive behaviors and
19. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for global severity in adult autism spectrum disorders. Am J Psychiatry
systematic reviews and meta-analyses: the PRISMA statement. 2012;169:292-9.
J Clin Epidemiol 2009;62:1006-12. 40. Carminati GG, Gerber F, Darbellay B, et al. Using venlafaxine to
20. Moher D, Shamseer L, Clarke M, et al. Preferred reporting items treat behavioral disorders in patients with autism spectrum disor-
for systematic review and meta-analysis protocols (PRISMA-P) der. Prog Neuropsychopharmacol Biol Psychiatry 2016;65:85-95.
2015 statement. Syst Rev 2015;4:1. 41. Reddihough DS, Marraffa C, Mouti A, et al. Effect of fluoxetine on
21. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med obsessive-compulsive behaviors in children and adolescents with
(Zagreb) 2012;22:276-82. autism spectrum disorders: a randomized clinical trial. JAMA
2019;322:1561-9.
22. Higgins JP, Thomas J, Chandler J, et al. Cochrane handbook for systematic
reviews of interventions. Hoboken (NJ): John Wiley & Sons; 2019. 42. Sanchez LE, Adams PB, Uysal S, et al. A comparison of live and video-
tape ratings: clomipramine and haloperidol in autism. Psychopharmacol
23. Higgins JPT, Eldridge S, Li T. Including variants on randomized Bull 1995;31:371-8.
trials. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T,
Page MJ, Welch VA, editors. Cochrane handbook for systematic reviews 43. Mirtazapine treatment of anxiety in children and adolescents with
of interventions version 6.3 (updated February 2022). London: pervasive developmental disorders. ClinicalTrials.gov: NCT01302964;
Cochrane Collaboration; 2022: chapter 23. Available: www.training. 2018. Available: www.clinicaltrials.gov/show/NCT01302964 (ac-
cochrane.org/handbook (accessed 2022 May 28). cessed 2021 Jul. 23).
44. Early intervention with fluoxetine in autism. ClinicalTrials.gov:
24. Borenstein M, Hedges LV, Higgins JP, et al. A basic introduction to
NCT00183339; 2014. Available: www.ClinicalTrials.gov/show/
fixed-effect and random-effects models for meta-analysis. Res Synth
NCT00183339 (accessed 2021 Jul. 23).
Methods 2010;1:97-111.
45. Leonard HL, Swedo SE, Rapoport JL, et al. Treatment of obsessive-
25. Riley RD, Higgins JP, Deeks JJ. Interpretation of random effects
compulsive disorder with clomipramine and desipramine in chil-
meta-analyses. BMJ 2011;342:d549.
dren and adolescents. A double-blind crossover comparison. Arch
26. Higgins J, Green S, editors. Cochrane handbook for systematic reviews Gen Psychiatry 1989;46:1088-92.
of interventions. London: Cochrane Collaboration; 2009.
46. Geller DA, Biederman J, Stewart SE, et al. Which SSRI? A meta-
27. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis analysis of pharmacotherapy trials in pediatric obsessive-compulsive
detected by a simple, graphical test. BMJ 1997;315:629-34. disorder. Am J Psychiatry 2003;160:1919-28.
28. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based 47. Doyle CA, McDougle CJ. Pharmacologic treatments for the behav-
method of testing and adjusting for publication bias in meta-analysis. ioral symptoms associated with autism spectrum disorders across
Biometrics 2000;56:455-63. the lifespan. Dialogues Clin Neurosci 2012;14:263-79.
29. Potter LA, Scholze DA, Biag HMB, et al. A randomized controlled 48. Brandenburg C, Blatt GJ. Differential serotonin transporter (5-HTT)
trial of sertraline in young children with autism spectrum disor- and 5-HT(2) receptor density in limbic and neocortical areas of
der. Front Psychiatry 2019;10:810. adults and children with autism spectrum disorders: implications
30. Herscu P, Handen BL, Arnold LE, et al. The SOFIA study: negative for selective serotonin reuptake inhibitor efficacy. J Neurochem
multi-center study of low dose fluoxetine on repetitive behaviors in 2019;151:642-55.
children and adolescents with autistic disorder. J Autism Dev Disord 49. Reiersen AM, Handen B. Commentary on “Selective serotonin
2020;50:3233-44. reuptake inhibitors (SSRIs) for autism spectrum disorders (ASD).”
31. Gordon CT, Rapoport JL, Hamburger SD, et al. Differential re- Evid Based Child Health 2011;6:1082-5.
sponse of seven subjects with autistic disorder to clomipramine 50. DeLong GR, Ritch CR, Burch S. Fluoxetine response in children
and desipramine. Am J Psychiatry 1992;149:363-6. with autistic spectrum disorders: correlation with familial major af-
32. Chugani DC, Chugani HT, Wiznitzer M, et al. Efficacy of low-dose fective disorder and intellectual achievement. Dev Med Child Neurol
buspirone for restricted and repetitive behavior in young children 2002;44:652-9.
with autism spectrum disorder: a randomized trial. J Pediatr 2016;​170:​ 51. Downs J, Hotopf M, Ford T, et al. Clinical predictors of antipsy-
45-53.e1-4. chotic use in children and adolescents with autism spectrum disor-
33. Gordon CT, State RC, Nelson JE, et al. A double-blind comparison ders: a historical open cohort study using electronic health re-
of clomipramine, desipramine, and placebo in the treatment of au- cords. Eur Child Adolesc Psychiatry 2016;25:649-58.
tistic disorder. Arch Gen Psychiatry 1993;50:441-7. 52. Posey DJ, Stigler KA, Erickson CA, et al. Antipsychotics in the
34. McDougle CJ, Naylor ST, Cohen DJ, et al. A double-blind, placebo- treatment of autism. J Clin Invest 2008;118:6-14.
controlled study of fluvoxamine in adults with autistic disorder. 53. Perry R, Campbell M, Adams P, et al. Long-term efficacy of halo-
Arch Gen Psychiatry 1996;53:1001-8. peridol in autistic children: continuous versus discontinuous drug
35. Remington G, Sloman L, Konstantareas M, et al. Clomipramine ver- administration. J Am Acad Child Adolesc Psychiatry 1989;28:87-92.
sus haloperidol in the treatment of autistic disorder: a double-blind, 54. McDougle CJ, Kresch LE, Posey DJ. Repetitive thoughts and be-
placebo-controlled, crossover study. J Clin Psychopharmacol 2001;21​:​ havior in pervasive developmental disorders: treatment with sero-
440-4. tonin reuptake inhibitors. J Autism Dev Disord 2000;30:427-35.
36. Hollander E, Phillips A, Chaplin W, et al. A placebo controlled 55. Reinblatt SP, DosReis S, Walkup JT, et al. Activation adverse events
crossover trial of liquid fluoxetine on repetitive behaviors in induced by the selective serotonin reuptake inhibitor fluvoxamine
childhood and adolescent autism. Neuropsychopharmacology 2005;​ in children and adolescents. J Child Adolesc Psychopharmacol
30:582-9. 2009;19:119-26.
37. Sugie Y, Sugie H, Fukuda T, et al. Clinical efficacy of fluvoxamine 56. Varley CK. Sudden death related to selected tricyclic antidepres-
and functional polymorphism in a serotonin transporter gene on sants in children: epidemiology, mechanisms and clinical implica-
childhood autism. J Autism Dev Disord 2005;35:377-85. tions. Paediatr Drugs 2001;3:613-27.

E310 J Psychiatry Neurosci 2022;47(4)