Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 May 5; 334(7600): 911–912.
PMCID: PMC1865459

Onset of action of antidepressants

Andre Tylee, professor of primary care mental health and Paul Walters, Medical Research Council fellow

Most benefit is evident in the first two weeks, not six, as conventional wisdom says

Recent guidance from the National Institute for Health and Clinical Excellence (NICE) says that antidepressant drugs should be offered routinely to all patients with depression of at least moderate severity and recommends a selective serotonin reuptake inhibitor as first line treatment.1 The NICE guidance goes on to state that “Patients started on antidepressants should be informed about the delay in onset of effect.” This reflects conventional wisdom, but is it time to revisit this idea?

Speed of onset of the actions of antidepressants is clinically important for several reasons. Delayed onset means that depression, its associated disability, and for some patients the potential risk of suicide continue. Early onset of effects may improve future compliance and thus outcomes.

When tricyclic antidepressants were first introduced in the 1950s delays in antidepressant effects were not reported. Indeed, researchers on early tricyclic antidepressants asserted that they usually started to work within the first few days of treatment.2 3 Later clinical experience suggested, however, that the drugs did not act immediately. The ensuing debate continued into the 1970s. By the mid-1970s, animal models suggested that the dissociation of acute biochemical changes induced by antidepressant treatment and the therapeutic action were due to the development of subsensitivity in the postsynaptic monoamine receptor.4 5 In animal models, these changes became apparent only after dosing with antidepressants over a similar period to that taken for clinical efficacy to develop. In the 1980s, a series of pooled clinical studies seemed to confirm this delayed onset of action.6 Since then, increasingly refined neurobiological theories of the action of antidepressants have incorporated this delay.7

This message is largely unchanged, despite the development of newer antidepressants, such as the selective serotonin reuptake inhibitors, and even though newer antidepressants can often be started at a therapeutic dose, rather than titrated upwards over two to three weeks, as is necessary with the older tricyclic antidepressants to minimise adverse effects.

Research on this question is hampered by lack of an agreed definition of onset of action.8 This is particularly true in clinical practice, where it may be difficult to distinguish between signs of response and side effects. For example, sedation may relieve symptoms but it is not directly related to the medicine's antidepressant properties. Recently, however, several studies have challenged the assumption of a delay in the onset of antidepressant action.9 10 11

A meta-analysis of 76 double blind placebo controlled trials of antidepressant treatment for depression in 2005 found that 60% of overall improvement occurred during the first two weeks and that half of all patients who respond to a six week trial respond in the same period.10 More recently, a meta-analysis of placebo controlled trials of selective serotonin reuptake inhibitors suggested that therapeutic response is greatest in the first week, with a gradual decline in the size of benefit over successive weeks of treatment.11 One third of the total effect seen at six weeks was apparent in the first week.11 As the studies were placebo controlled trials, this improvement was unlikely to be a placebo effect.

These recent findings raise further questions. Is speed of therapeutic benefit with antidepressants a class effect, and do differences occur within classes? Do some symptoms respond quicker than others? Does early response predict future response and, if so, should we routinely review response earlier and change treatment if no response occurs in the first week or two? Does this phenomenon apply only to a subset of the population and is it genetically determined? Should we be encouraging patients to anticipate early relief from symptoms and, if so, is there a risk of disappointment if benefit is delayed?

Until studies are specifically designed to measure the onset of action of antidepressants, results from meta-analyses of studies not designed for this purpose should be treated with caution. We suggest that future studies should look for subsets of symptoms that may be ameliorated earlier than others and seek to discover how this is mediated. In the meantime, if these findings are correct it is good news for many patients with depression treated with antidepressants. But these results are unlikely to alter clinical practice until these additional questions are answered.


Competing interests: AT has received fees for speaking, consulting, and attending advisory boards of several companies that manufacture antidepressant drugs (Eli Lilly, Lundbeck, Wyeth, GSK, Servier, Pfizer, Organon, etc). He is currently receiving research funding from Servier Pharmaceuticals. PW has also received fees for speaking and attending advisory boards for several companies that manufacture antidepressants.

Provenance and peer review: Non-commissioned; externally peer reviewed.


1. National Collaborating Centre for Mental Health. Depression: management of depression in primary and secondary care Clinical guideline 23. 2004. London, National Institute for Health and Clinical Excellence.
2. Kuhn R. The treatment of depressive states with G22355 (imipramine hydrochloride). Am J Psychiatry 1958;115:459-64. [PubMed]
3. Pollack B. Clinical findings in the use of Tofranil in depressive and other psychiatric states. Am J Psychiatry 1959;116:317
4. Vetulani J, Sundell K. Action of various antidepressant treatments reduces reactivity of noradrenergic cyclic AMP-generating system in the limbic forebrain. Nature 1975;257:495-6. [PubMed]
5. Banerjee SP, Kung LS. Development of B-adrenergic receptor subsensitivity by antidepressants. Nature 1977;268:455-6. [PubMed]
6. Quitkin FM, Rabkin JG, Ross D, Stewart JW. Identification of true drug response to antidepressants. Use of pattern analysis. Arch Gen Psychiatry 1984;41:782-6. [PubMed]
7. Reid IC, Stewart CA. How antidepressants work: new perspectives on the pathophysiology of depressive disorder. Br J Psychiatry 2001;178:299-303. [PubMed]
8. Leon AC. Measuring onset of antidepressant action in clinical trials: an overview of definitions and methodology. J Clin Psychiatry 2001;62(suppl 4):12-6; discussion 37-40.
9. Nierenberg AA, Farabaugh AH, Alpert JE, Gordon J, Worthington JJ, Rosenbaum JF, et al. Timing of onset of antidepressant response with fluoxetine treatment. Am J Psychiatry 2000;157:1423-8. [PubMed]
10. Posternak MA, Zimmerman M. Is there a delay in the antidepressant effect? A meta-analysis. J Clin Psychiatry 2005;66:148-58. [PubMed]
11. Taylor MJ, Freemantle N, Geddes JR, Bhagwagar Z. Early onset of selective serotonin reuptake inhibitor antidepressant action: systematic review and meta-analysis. Arch Gen Psychiatry 2006;63:1217-23. [PMC free article] [PubMed]

Articles from The BMJ are provided here courtesy of BMJ Publishing Group