Article Text
PDF
Statistics from Altmetric.com
Question
Question: In children and adolescents with mood or anxiety disorders, is antidepressant use associated with symptoms of excessive emotional arousal or behavioural activation?
Outcomes: Observed symptoms of excessive emotional arousal or behavioural activation, in addition to mania-hypomania (details of scales and methods not provided).
Methods
Design: Systematic review with meta-analysis.
Data sources: MEDLINE, PsycINFO, CINAHL, PubMed, Cochrane Library and Web of Science were searched from inception to March 2012, supplemented by hand search of reference lists.
Study selection and analysis: Randomised controlled trials (RCTs), retrospective studies, reviews and meta-analyses of antidepressant use in children or adolescents (aged less than 18 years) with diagnosed depression or anxiety disorders. Studies were reviewed to identify reports of the new-onset behavioural symptoms. Continuous measures were compared using analysis of variance methods or paired t tests and random-effects meta-analysis.
Main results
Forty-two studies (n=6767) met inclusion criteria. Depression: 17 studies (six RCTs) were in children/adolescents (mean age 12.5 years; 54.3% boys) with depression disorders (n=2637; 2083 taking antidepressants and 554 placebo). Exposure to antidepressants ranged from 6 to 40 weeks. In children/adolescents with depression, antidepressant use was associated with an increased rate of behavioural symptoms compared with placebo (observed mean crude rates: 16.8%±14.6% vs 0.82%±0.86%, p=0.027). With adjustment for exposure times rates were 1.16% (±1.25%) per week with antidepressants versus 0.084% (±0.091%)/week with placebo. When only data from controlled studies were used, antidepressant use more than tripled the rate of behavioural symptoms compared with placebo (3.95% vs 1.08%, p=0.0018). Antidepressants were also observed to increase rates of mania-hypomania compared with placebo (10.4% vs 0.45%, p<0.0001). Anxiety: 25 studies (16 RCTs) were in children/adolescents (mean age 12 years; 57.5% boys) with anxiety disorders (n=4130; 3211 taking antidepressants and 919 placebo). Mean exposure to antidepressants was 17.6 weeks (range 8–104 weeks). In children/adolescents with anxiety disorders, antidepressants were associated with an increased rate of behavioural symptoms compared with placebo (observed mean crude rates: 22.6%±20.3% vs 7.23%±7.45%, p=0.006). With adjustment for exposure times rates were 2.10% (±2.22%)/week with antidepressants vs 0.714% (±0.820%)/week with placebo. When only data from controlled studies were used, antidepressant use more than doubled the rate of behavioural symptoms compared with placebo (11.7% vs 5.22%, p<0.0001). Antidepressants also increased rates of mania-hypomania compared with placebo (1.98% vs 0.00%, p=0.0001). Meta-analysis (19 controlled trials) of antidepressant use for any indication revealed that overall antidepressants increased risk of behavioural symptoms compared with placebo (RR 1.66, 95% CI 1.23 to 2.25, p=0.001). Separate meta-analyses for both depression and anxiety also found that antidepressants increased risk of behavioural symptoms compared with placebo (depression: RR 3.61, 95% CI 1.60 to 8.10; anxiety: RR 1.49, 95% CI 1.08 to 2.06).
Conclusions
Antidepressant use increases risk of excessive emotional arousal, behavioural activation and mania-hypomania in children or adolescents with depression or anxiety disorders.
Commentary
This study examined the risk of ‘activation’ and ‘mania-hypomania’ in antidepressant trials of depressed or anxious youth. ‘Activation’ included insomnia, arousal, irritability and anger. Within drug-placebo pairs, the rates of ‘activation’ for drug versus placebo in depression and anxiety were 3.95% vs 1.08% and 11.7% vs 5.22%, respectively, with a mean onset time of 5 weeks. For mania-hypomania, the drug-placebo differences for depression and anxiety were 10.4% vs 0.45% and 1.98% vs 0%, respectively. Weaknesses acknowledged by the authors include lack of a standard, reliable assessment of either outcome. Tests for heterogeneity and biases from unpublished data were not reported, and participants in these studies were treated with antidepressants, stimulants or combinations of antidepressants and mood stabilisers.
This study is useful as it gives some estimate for the prevalence of pharmacologically induced arousal-activation in depressed and anxious youth, as well as an expected time to onset. It also suggests that agitation is more common in the treatment of anxiety, whereas mania-hypomania is more common in the treatment of depression. This study complements another study which examined the risk of mania-hypomania in a large, representative database of youth treated with antidepressants and found an incidence rate of 5.4% in an unselected sample.1 In a study of depressed youth with a bipolar parent, 57% of those treated with an antidepressant developed irritability, aggression or hyperactivity.2 These two studies found the risk to be higher in younger patients.1 2 Taken together, these findings support the need for more reliable and consistent measures of activation and mania-hypomania, the need for biomarkers and pharmacokinetic studies (ie, to identify slow metabolisers) that predict activation or mania, and consideration of the use of psychotherapeutic alternatives, when clinically appropriate, in anxious youth and those at risk for mania.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online references
Footnotes
-
Sources of funding Bruce J Anderson Foundation; McClean Private Donors Psychopharmacology Research Fund.