What is the evidence for Antidepressants for Insomnia? Results of a Cochrane Systematic Review

Talk Code: 
Hazel Everitt
Hazel Everitt, David Baldwin, Beth Stuart, Gosia Lipinska, Andrew Mayers, Andrea L Malizia, Christopher Manson, Sue Wilson
Author institutions: 
University of Southampton, University of Cape Town, Bournemouth University, University of Bristol, Imperial College London


Insomnia is common, 6-15% of adults report significant sleep problems. Hypnotics have potential problems with addiction and tolerance and guidelines recommend only short term use. Psychological therapies can help but availability is very limited. Alternatives are sought. Antidepressants are prescribed despite being unlicensed for this use. This systematic review aims to clarify the evidence.


Cochrane systematic review procedures were used. OVID MEDLINE, EMBASE, PsycINFO and the Cochrane Register (CENTRAL) were searched. Randomised controlled trials in adults with a primary diagnosis of insomnia were included that compared any antidepressant as monotherapy at any dose to: placebo; other insomnia medications (e.g. benzodiazepines, 'Z' drugs); different antidepressants or treatment as usual.The primary outcome measure was subjective improvement in sleep (quality or quantity); secondary outcomes were objective sleep measures (eg EEG), tolerability, daytime symptoms or functioning.


4245 studies were identified, 220 screened in full text and 20 included. 3 studies (n=135) compared Selective Serotonin Reuptake Inhibitors (SSRIs) with placebo. Combining results was not possible due to reporting differences. The 2 paroxetine studies showed significant improvement in subjective sleep measures at 6 (p=0.03) and 12 weeks (p<0.001) respectively. No difference was seen in the fluoxetine study.One study compared paroxetine with alprazolam (n= 60) with greater improvement in subjective sleep measures in the paroxetine group (p <0.05) Three studies compared SSRIs with another antidepressant (agomelatine, nefazodone) (n= 489) revealing no significant differences in sleep quality (SMD 0.97, 95% CI -0.91 to 2.85; I2 = 99%) though heterogeneity was high. In SSRI studies, adverse events were either not reported, or showed similar low rates between drug and comparator.6 studies (812 participants) compared Tricyclic Antidepressants (TCA) with placebo, 5 doxepin and one trimipramine. 4 studies (n=518) could be pooled, showing significant improvement in subjective sleep quality over placebo (SMD -0.39, 95% CI -0.56 to -0.21; I2 = 0%). No significant difference in reported side effects or adverse events were found between TCAs and placebo. One study compared doxepin with lormetazepam and one doxepin with imipramine; neither revealed significant differences.Six studies compared other antidepressants with placebo (1 mianserin; 5 trazodone). Three trazodone studies (n=370) provided extractable data indicating improvement in subjective sleep outcomes for trazodone over placebo (SMD -0.34, 95% CI -0.66 to -0.02). Three trazodone studies reported more side effects in trazodone than placebo groups (ie ‘morning grogginess’, increased dry mouth and thirst).


Relatively few, mostly small studies with short term follow up and design limitations were identified. Analysis produced some data supporting short-term use for some antidepressants. No evidence was found for amitriptyline which is commonly used in clinical practice, or to support long-term antidepressant use for insomnia. Current evidence does not support current practice. High quality trials of antidepressants for insomnia are needed.

Submitted by: 
Hazel Everitt
Funding acknowledgement: 
Partially funded by a NIHR SCPR grant