Powered by the Evidence-based Practice Centers
SHARE:
FacebookTwitterFacebookPrintShare

Treatment of Depression in Children and Adolescents

Systematic Review Apr 10, 2020
Treatment of Depression in Children and Adolescents

Documents

This report is available in PDF only (Final Report [12.5 MB]; Evidence Summary [1.1 MB];  Disposition of Comments Report [585 KB]); and report data available in a Systematic Review Data Repository file. For additional assistance, please contact us.

Purpose of Review

The purpose of the review is to examine the benefits and harms of pharmacological and nonpharmacological treatments for child and adolescent depressive disorders.

Key Messages

  • Cognitive behavioral therapy (CBT), fluoxetine, escitalopram, and combined fluoxetine plus CBT may reduce depressive symptoms in the short term; clinical significance is unclear.
  • CBT may improve symptoms and functional status. CBT plus medications may help prevent relapse.
  • Selective serotonin reuptake inhibitors (SSRIs) as a class may improve response and functional status.
  • However, SSRIs may be associated with a higher risk of serious adverse events and with a higher risk of withdrawal. Paroxetine may be associated with a higher risk of suicidal ideation or behaviors. Evidence to judge the risk of suicidal ideation or behavior for SSRIs other than paroxetine is insufficient for major depressive disorder. However, this report excluded data on inpatients and those without depressive disorders whom the Food and Drug Administration included in finding an increased risk of suicidality for all antidepressants across all indications.

Structured Abstract

Background. Depressive disorders can affect long-term mental and physical health functioning among children and adolescents, including increased risk of suicide. Despite access to several nonpharmacological, pharmacological, and combined treatment options for childhood depression, clinicians contend with sparse evidence and are concerned about harms associated with treatment.

Methods. We conducted a systematic review to evaluate the efficacy, comparative effectiveness, and moderators of benefits and harms of available nonpharmacological and pharmacological treatments for children and adolescents with a confirmed diagnosis of a depressive disorder (DD)—major depressive disorder (MDD), persistent depressive disorder (previously termed dysthymia) or DD not otherwise specified. We searched five databases and other sources for evidence available from inception to May 29, 2019, dually screened the results, and analyzed eligible studies.

Results. We included in our analyses data from 60 studies (94 articles) that met our review eligibility criteria. For adolescents (study participants’ ages range from 12 to 18 years) with MDD, cognitive behavioral therapy (CBT), fluoxetine, escitalopram, and combined fluoxetine and CBT may improve depressive symptoms (1 randomized controlled trial [RCT] each, n ranges from 212 to 311); whether the magnitude of improvement is clinically significant is unclear. Among adolescents or children with MDD, CBT plus medications (8–17 years) may be associated with lower rates of relapse (1 RCT [n = 121]). In the same population (6–17 years), selective serotonin reuptake inhibitors (SSRIs) may be associated with improved response (7 RCTs [n = 1,525]; risk difference [RD], 72/1,000 [95% confidence interval (CI), 2 to 24], I2 = 9%) and functional status (5 RCTs [n = 941]; standardized mean difference, 0.16 [95% CI, 0.03 to 0.29]; I2 = 0%). For adolescents or children with any DD (7–18 years), CBT or family therapy may be associated with improvements in symptoms, response, or functional status (1 RCT each, n ranges from 64 to 99). Among children with any DD (7–12 years), family-based interpersonal therapy may be associated with improved symptoms (1 RCT, n = 38). Psychotherapy trials did not report harms. SSRIs may be associated with a higher risk of serious adverse events among adolescents or children with MDD (7–18 years; 9 RCTs [n = 2,206]; RD, 20/1,000 [95% CI, 1 to 440]; I2, 4%) and with a higher risk of withdrawal due to adverse events among adolescents with MDD (12–18 years; 4 RCTs [n = 1,296], RD, 26/1,000 [95% CI, 6 to 45]; I2, 0%). Paroxetine (1 RCT [n = 180]) may be associated with a higher risk of suicidal ideation or behaviors among adolescents with MDD (12–18 years). Evidence was insufficient to judge the risk of suicidal ideation or behavior for other SSRIs for adolescents and children with MDD or other DD (7–18 years) (10 RCTs [n = 2,368]; relative risk, 1.14 [95% CI, 0.89 to 1.45]; I2, 8%). However, this report excluded data on inpatients and those without depressive disorders, whom the Food and Drug Administration included in finding an increased risk of suicidality for all antidepressants across all indications.

Conclusion. Efficacious treatments exist for adolescents with MDD. SSRIs may be associated with increased withdrawal and serious adverse events. No evidence on harms of psychotherapy were identified.

Visual Dashboard

 

Report Snapshot

Summary of
Findings
Key
Questions
Related Findings
and Clinical Guidelines
Clinical and Policy
Implications
Caveats, Applicability,
and Limitations
 

Combined — Any combined treatment that includes two or more types of nonpharmacological, pharmacological, and/or collaborative care interventions, either started together or given as augments to initial treatment types.

Head to Head comparison — Any nonpharmacological, pharmacological, or collaborative care intervention alone or in combination compared to any other nonpharmacological, pharmacological, or collaborative care intervention alone.

Collaborative Care — Any treatment that includes collaborative care, integrated care, integrative care, stepped care, coordinated care, comanaged care, or colocated care. We found no studies of collaborative care interventions that met our inclusion and exclusion criteria.

List of Abbreviations — CBT = Cognitive Behavioral Therapy; DD = Depressive Disorder; DD NOS = Depressive Disorder not otherwise specified; IPT = Interpersonal psychotherapy; MD = Mean difference; MDD = Major Depressive Disorder; NA = Not applicable; OR = Odds Ratio; PCIT = Parent-Child Interaction Therapy; RR = Risk ratio; SMD = Standard mean difference; STS = Selegiline Transdermal System; TAU = Treatment as usual; UC = Usual Care

Within any selected outcome, click on any highlighted table row to open the relevant studies in PubMed (some included studies are not presented in PubMed and the table row will not highlight; see the full report for all references).

Nonpharmacologic
 
Pharmacologic
 
Combined
 
Head-to-head Comparison
 
Collaborative Care (no data)
 
Selected population:
Depressive symptoms
open
Functional impairment
open
Mortality
open
Recovery
open
Relapse
open
Remission
open
Response
open
Severe Adverse Events (SAEs)
open
Suicidality
open
Withdrawal due to adverse events
open
Where links are available within the Report Snapshot tables, clicking the link will take you to the PubMed listing for the studies available within PubMed. Not all studies in all findings are available in PubMed.

We sought to address five Key Questions:

  • Key Question 1: Nonpharmacological interventions
    • Key Question 1a: In adolescents and children, what are the benefits and harms of nonpharmacological interventions for depressive disorders (DDs) (e.g., major depressive disorder [MDD], persistent depressive disorder [PDD]/dysthymic disorders, or DD not otherwise specified [DD NOS])?
    • Key Question 1b: How do the benefits and harms vary by subpopulation (e.g., patient characteristics, parent/caregiver characteristics, disorder characteristics, history of previous treatment, comorbid condition, exposure to a traumatic life event)?
  • Key Question 2: Pharmacological treatment
    • Key Question 2a: In adolescents and children, what are the benefits and harms of pharmacological interventions for DDs (MDD, PDD/dysthymic disorders, or DD NOS)?
    • Key Question 2b: How do the benefits and harms vary by subpopulation (e.g., patient characteristics, disorder characteristics, history of previous treatment, comorbid condition, exposure to a traumatic life event)?
  • Key Question 3: Combination treatment
    • Key Question 3a: In adolescents and children, what are the benefits and harms of combination interventions for DDs (MDD, PDD/dysthymic disorders, or DD NOS)?
    • Key Question 3b: How do the benefits and harms vary by subpopulation (e.g., patient characteristics, disorder characteristics, history of previous treatment, comorbid condition, exposure to a traumatic life event)?
  • Key Question 4: Collaborative care interventions
    • Key Question 4a: In adolescents and children, what are the benefits and harms of collaborative care interventions for DDs (MDD, PDD/dysthymic disorders, or DD NOS)?
    • Key Question 4b: How do the benefits and harms vary by subpopulation (e.g., patient characteristics, disorder characteristics, history of previous treatment, comorbid condition, exposure to a traumatic life event)?
  • Key Question 5: Head-to-head comparisons of treatments
    • Key Question 5a: In adolescents and children, what are the comparative benefits and harms of treatments (pharmacological, nonpharmacological, combined, collaborative care interventions) for DDs (MDD, PDD/dysthymic disorders, or DD NOS)?
    • Key Question 5b: How do the benefits and harms vary by subpopulation (e.g., patient characteristics, disorder characteristics, history of previous treatment, comorbid condition, exposure to a traumatic life event)?

For each Key Question, we first considered evidence about individuals ≤ 18 years old with depressive disorders and then considered the following subpopulations, outcomes, and settings:

Population Children and adolescents (≤ 18 years old) with a DD (MDD or PDD/Dysthymia)

Subgroups of interest:

  • Patient characteristics (e.g., developmental age—child or adolescent, gender, race/ethnicity)
  • Parent/caregiver characteristics
  • Disorder characteristics (e.g., type, severity)
  • History of previous treatment
  • Comorbid conditions
  • Exposure to a traumatic life event
Outcomes Benefits
  • Remission
  • Response
  • Relapse
  • Recovery
  • Depressive symptoms
  • Suicidality
  • Mortality
  • Functional impairment
Harms
  • Any adverse effects of intervention (e.g., death, serious adverse effects)
Settings Outpatient care in countries with a very high Human Development Index

This section provides information on how this AHRQ review relates to other systematic reviews of and clinical guidelines for treatment of childhood depression.

Related Review Findings

In 2003, the Food and Drug Administration (FDA) issued a public health advisory for paroxetine because of concerns about suicidal ideation. FDA extended the warning to all antidepressants in 2004, based on a systematic review new tab of 24 trials of participants with depression. The FDA review found an increased risk of suicidality as an adverse event (AE) for all antidepressants and all indications; however, these findings also noted the lack of statistically significant differences in Major Depressive Disorder (MDD) populations by drug, other than for venlafaxine extended release.

Results from our review also suggest an increased but nonstatistically significant effect for populations with depressive disorders (DDs). There are, however, several noteworthy differences between our analyses and those conducted by FDA:

  1. The scope of FDA's analysis differed from that of the current review:
    1. FDA relied on a meta-analysis of 24 trials, 16 of which included depressed patients. Our review required all participants in all studies to have diagnosed DDs; we excluded studies of patients with other disorders. Suicidality as an AE may be easier to identify in patients with disorders other than depression, where the mitigating benefit of clinical improvement in depressive symptoms (including suicidal ideation) is not observed.
    2. We also excluded studies with inpatient populations. All these exclusions resulted in a smaller yield than FDA's analysis for all antidepressants and all indications for the corresponding time period; the reduced power of the remaining trials lowered the precision of the estimate.
  2. The time period covered also differed:
    1. Our review included studies published after the issuance of the FDA warnings more than a decade ago. These studies did not find evidence of increased suicidal ideation for the Selective Serotonin Reuptake Inhibitor (SSRI) drug class as a whole, in part because some studies were not powered to detect differences.
  3. Methodological differences may account for other variations in findings between our review and FDA's analysis:
    1. Our review relied on reported suicide ideation or behavior data, whereas FDA's review included analyses of text string data of AE reporting systems collected by drug manufacturers. In fact, reanalysis of the data from text strings that used a standardized approach to categorizing suicidality revealed high rates of discrepancies in suicidality determinations.
      1. Of note, although none of the studies new tab included in FDA's review reported any completed suicides in either the intervention or control group throughout the study period, this fact points to the rarity of suicides and the low power of existing trials to adequately address this outcome.
      2. An added problem relates to documented instances of selective outcome reporting. As noted previously, the "restoring invisible and abandoned trials" [RIAT] analysis of a paroxetine trial found increased suicidal ideation for paroxetine. Similar analyses for older trials may further alter the evidence base.

Clinical Guidelines

  1. Three recent guidelines—from the American Psychological Association new tab, the National Institute for Health and Care Excellence (NICE) new tab, and the Guidelines for Adolescent Depression in Primary Care (GLAD-PC parts oneand two)—continue to have uncertainty related to treating children, disorders other than MDD, and partial or no response to initial therapy. They do support Cognitive Behavioral Therapy (CBT), fluoxetine, and combined therapies.
  2. In our review, few comparative effectiveness trials demonstrated a clear advantage of one type of treatment over another:
    1. The exceptions generally arose from single studies, which showed some benefit for SSRIs in combination with CBT when compared CBT or SSRIs alone.
    2. The finding of insufficient evidence for most comparators is consistent with the NICE guidelines new tab that found no clear evidence of superiority of one treatment over another.
    3. Our review demonstrated some initial subgroup differences in efficacy and comparative effectiveness. Differences in efficacy and effectiveness based on depression severity and comorbid conditions support current guidelines (GLAD-PC part one and two new tab, NICE guidance new tab, American Academy of Adolescent Psychiatry new tab) that recommend different treatments based on such characteristics.
  3. Clinical practice guidelines (GLAD-PC part one and two new tab, NICE guidance new tab) for providers treating depressed children and adolescents have included language careful to acknowledge that some increased suicide risk may occur but that the benefits of SSRIs outweigh potential harms.
    1. Balance between risks and potential harms might differ by age group (i.e., because both benefits and harms might differ by age group), so these guidelines typically separate out adolescent and younger children populations.
      1. Research in adults new tab has also shown age-dependent effects: antidepressant use is associated with increased suicide risk among young adults, has no significant associations among middle-aged adults, and is inversely associated with suicide risk among older adults. There is a paucity of data available to study associations with antidepressant use among younger children.
      2. Careful study of differences in suicidality associated with antidepressant use in adolescents versus younger children is difficult because the sample sizes to conduct these analyses are further compromised in these studies.
        1. These recommendations offer support for treating adolescent depression, especially when the proximal time to the initial prescription includes vigilant monitoring for increases in suicidality to mitigate the excess risk.
        2. Our findings are consistent with current clinical practice guidelines by the GLAD-PC part one and two new tab that recommend the use of psychotherapies (specifically CBT), SSRI medications, or both for those age 10 or older.
        3. They are also consistent with the U.S. Preventive Services Task Force recommendation statement new tab based on found efficacy of CBT, fluoxetine, escitalopram, and combined treatments among adolescents.

Clinical Implications

Across populations and disorders

  • The findings of this review indicate that several interventions may be associated with low strength of evidence of benefits such as cognitive behavioral therapy (CBT), fluoxetine, escitalopram, and combined fluoxetine and CBT in the short term; we found insufficient evidence on harms for these individual interventions.

  • Except for paroxetine having a higher risk of suicidal ideation or behaviors in adolescents with major depressive disorder (MDD), evidence was insufficient for other Select Serotonin Reuptake Inhibitors (SSRIs) as a drug class across populations and disorders, likely as a result of low power to detect these events.




  •  
  •  
  •  

When parsed by population and disorder our results suggest

  • For adolescents with MDD
    • CBT, fluoxetine, escitalopram, and combined fluoxetine plus CBT may reduce depressive symptoms in the short term. However, the clinical implication of improvement in continuous measures of depressive symptoms is unclear.
    • Paroxetine may be associated with a higher risk of suicidal ideation or behaviors.
  • Among adolescents and children with MDD
    • As a class, SSRIs may improve response and functional status. However, they may be associated with a higher risk of serious adverse events (AEs) among adolescents and children with MDD. SSRIs may also be associated with a higher risk of withdrawal due to AEs among adolescents with MDD.
    • CBT plus medications may help prevent relapse.
  • For adolescents and children with MDD, persistent depressive disorder (PDD), or depressive disorder not otherwise specified (DD NOS), CBT and family therapy may improve symptoms, response, and functional status.
  • For children with MDD alone or with a wider range of depressive disorders (MDD, PDD, or DD NOS) evidence is sparse.

This review does not provide cost information.

Applicability

  • The results of this review are generally applicable to a population of adolescents and children with Major Depressive Disorder (MDD) with limited psychiatric comorbidity with access to (1) mental health professionals who can provide Cognitive Behavioral Therapy (CBT) or (2) psychiatrists or pediatricians who can prescribe Selective Serotonin Reuptake Inhibitors (SSRIs), Serotonin-Norepinephrine Reuptake Inhibitors (SNRIs), tricyclic antidepressants, or monoamine oxidase inhibitors.
  • Studies were more likely to include adolescents than children.
  • Patients with MDD are vastly more represented than those with other forms of depression such as dysthymia.
  • When comorbidity was described, the most common comorbidities were anxiety disorders, attention deficit hyperactivity disorder (ADHD), or disruptive disorders.
  • Most studies were conducted in majority-white populations in North America or Europe.
  • Baseline depression severity was generally in the mild to moderate range.
  • It is difficult to determine how the results would apply to a more diverse, complicated, or impaired population.

Limitations

  • Selective outcome reporting and the potential for publication bias, particularly with regard to harms of pharmacotherapy.
  • Small body of evidence examining effective treatments for children and adolescents.
    • We restricted inclusion to randomized controlled trials when examining benefits; this synthesis does not include data from observational studies and pooled analyses.
    • High risk of bias in many trials; this rating limited our confidence in the study conclusions regarding the effectiveness of the intervention.
    • Limited eligibility to studies of outpatients with depressive disorders (DDs), thereby reducing the available power of the analysis and possibly resulting in understated harms of antidepressants.
  • Limited treatment modalities.
    • Several psychotherapy trials examined effectiveness of CBT, fewer examined interpersonal psychotherapy (IPT) or family therapy, and still fewer examined effectiveness of other interventions for depression in children.
  • Limited evidence on children.
    • Our inclusion criteria required a diagnosed DD; the evidence base in this review is not representative of interventions for children who have clinically elevated symptoms but do not have mood disorders.
  • No studies examined effectiveness of newer approaches to treatment (i.e., motivational interviewing or acceptance and commitment therapy).
  • Very limited findings on harms in treatment and placebo arms of studies.
    • Studies that reported harms were generally not powered to do so, further limiting any conclusions on harms.
    • Information on the rate of harms of untreated depression is particularly important in the context of rising suicidality following the boxed warning among depressed children and adolescents.
    • In interpreting the available data on harms from treatment, clinicians also need to account for the profound harms of untreated depression.
  • Heterogeneity and lack of specification regarding:
    • Level of training of individuals providing the intervention
    • Frequency or dose of the intervention
    • Duration of treatment
  • Specific to medication trials include:
    • Short duration of intervention and follow-up
    • Lack of generalizability of the study population
    • Heterogeneity of assessment tools
    • Susceptibility to risk of bias
  • Lack of uniformity of measures for assessment of depression.
    • Heterogeneity of assessment instruments used to diagnose depression likely influenced variability in results.
    • Outcomes for some studies were measured using self-reports of depressive symptoms; other studies used clinician or parent reports of depressive symptoms.
  • Many trials had a small number of participants.
    • This limitation influenced our judgement on the precision of the evidence and the consequent certainty.

Citation

Suggested citation: Viswanathan M, Kennedy SM, McKeeman J, Christian R, Coker-Schwimmer M, Cook Middleton J, Bann C, Lux L, Randolph C, Forman-Hoffman V. Treatment of Depression in Children and Adolescents: A Systematic Review. Comparative Effectiveness Review No. 224. (Prepared by the RTI International–University of North Carolina at Chapel Hill Evidence-based Practice Center under Contract No. 290-2015-00011-I.) AHRQ Publication No. 20-EHC005-EF. Rockville, MD: Agency for Healthcare Research and Quality; April 2020. DOI: https://doi.org/10.23970/AHRQEPCCER224. Posted final reports are located on the Effective Health Care Program search page.

Project Timeline

Treatment of Childhood Depression

Jan 31, 2018
Topic Initiated
Aug 24, 2018
Apr 10, 2020
Systematic Review
Page last reviewed March 2022
Page originally created April 2020

Internet Citation: Systematic Review: Treatment of Depression in Children and Adolescents. Content last reviewed March 2022. Effective Health Care Program, Agency for Healthcare Research and Quality, Rockville, MD.
https://effectivehealthcare.ahrq.gov/products/childhood-depression/research

Select to copy citation