Antidepressants are effective to treat moderate to severe depression in adults. Five antidepressants appear more effective and better tolerated than others.
A major review of 522 antidepressant trials found that all of the 21 drugs studied performed better than placebo, in short-term trials measuring response to treatment. However, effectiveness varied widely.
Researchers ranked drugs by effectiveness and acceptability after eight weeks of treatment. Several drugs were more effective and were stopped by fewer people than others:
The review provides new evidence which may help people decide which antidepressant to choose first-line for moderate to severe depression. However, it did not assess antidepressants compared to other treatments such as cognitive behavioural therapy, or treatments in combination. Though there are some concerns over items not reported by individual trials, this review is likely to be reliable. It is extensive, included only placebo controlled double blind trials and searched successfully for unpublished trials.
Why was this study needed?
Depression is a common condition, affecting an estimated 1 in 10 adults at some point in their lives. Antidepressants are widely prescribed in primary and secondary care, along with psychological interventions such as cognitive behavioural or interpersonal therapy. There is conflicting evidence to guide which antidepressants should be prescribed first-line, although NICE recommends a selective serotonin reuptake inhibitor (SSRI).
There has been uncertainty in recent years about the effectiveness of antidepressants. Their mode of action is poorly understood, and improvement in mood tends to be modest. One 2008 meta-analysis suggested that antidepressants gave little benefit over placebo for mild to moderate depression.
This new analysis went to some lengths to find unpublished studies and additional data from published studies, to give us the best overview of the current state of research.
What did this study do?
This systematic review and network meta-analysis compared 21 antidepressants with placebo or each other, directly within trials and indirectly across trials. They included 522 double-blind randomised controlled trials of 116,477 adults with moderate to severe depression.
More than 100 trials were previously unpublished. As well as publication databases, international trial registers and drug approval websites, the researchers had contacted all pharmaceutical companies marketing antidepressants to ask for unpublished studies.
The antidepressants were compared for effectiveness (at least 50% improvement in symptoms) and acceptability (assessed as drop-out rate). They found 380 trials at possible risk of bias due mainly to lack of reporting of randomisation methods, and 46 at high risk. However, the trials were all placebo controlled.
What did it find?
- All 21 antidepressants were more likely to produce a treatment response after eight weeks treatment than a placebo. The most effective antidepressant compared to placebo was the tricyclic antidepressant amitriptyline, which increased the chances of treatment response more than two-fold (odds ratio [OR] 2.13, 95% credible interval [CrI] 1.89 to 2.41). The least effective was the serotonin and noradrenaline reuptake inhibitor reboxetine, which increased treatment response by 37% (OR 1.37, 95% CrI 1.16 to 1.63).
- Drop-out rates by eight weeks of treatment were similar to placebo for the majority of antidepressants. People were 30% more likely to stop taking the tricyclic clomipramine than placebo (OR 1.30, 95% CrI 1.01 to 1.68) and slightly less likely to stop taking agomelatine (an “atypical” antidepressant) or the SSRI fluoxetine (OR for agomelatine 0.84, 95% CrI 0.72 to 0.97; OR for fluoxetine 0.88, 95% CrI 0.8 to 0.96).
- In head-to-head comparisons between the drugs, five were identified as having a combination of better effectiveness and lower drop-out rates, compared to others: the SSRIs escitalopram, paroxetine and sertraline, and atypicals agomelatine and mirtazapine. Reboxetine (atypical), trazodone (similar to a tricyclic) and fluvoxamine (SSRI) were identified as having lower effectiveness and higher drop-out rates.
- Though absolute effect sizes were not reported in the results, the researchers described the effect sizes as “modest”. However, they also said that “non-response to treatment will occur.”
What does current guidance say on this issue?
The NICE 2009 guideline on depression advises that people with moderate to severe depression should be offered an antidepressant and psychological therapy such as cognitive behavioural therapy or interpersonal therapy. It says the antidepressant prescribed “should normally be an SSRI in a generic form because SSRIs are equally effective as other antidepressants and have a favourable risk-benefit ratio.”
The guideline warns that venlafaxine is more associated with risk of death from overdose than other SSRIs, while “tricyclic antidepressants, except for lofepramine, are associated with the greatest risk in overdose.”
The guideline, last updated in 2016, is under review.
What are the implications?
The findings are of interest to GPs and psychiatrists, who need to decide on the best initial treatment for adults with moderate to severe depression. The comparative data may help doctors select drugs with better efficacy and side-effects.
However, treatment choice will be guided by an individual patient's circumstances and preferences. The meta-analysis was unable to look at the potentially different effects of treatment on subgroups based on age, sex, the severity of symptoms or duration of illness.
The review did not consider combined drug and psychological treatments, as recommended by NICE for moderate to severe depression, or long-term effects which limit its applicability.
Citation and Funding
Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet. 2018. [Epub ahead of print].
This project was funded by the National Institute for Health Research Oxford Health Biomedical Research Centre (BRC-1215-20005) and the Japan Society for the Promotion of Science.
Parikh SV, Kennedy SH. More data, more answers: picking the optimal antidepressant. Lancet 2018.
NICE. Depression in adults: recognition and management. CG90. London: National Institute for Health and Care Excellence; 2009.
Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre