August 18, 2009
This extensive research by the FDA examined reports of suicidal thoughts or behaviour in 372 placebo-controlled trials of antidepressants in all ages for various reasons. It found that under-25s on the drugs tended to have an increased risk of suicidal thoughts or behaviour (preparatory actions for suicide or attempted or completed suicide).
There are several important limitations to these findings, but they are likely to lead to further research and may lead to changes in drug regulatory information. The results highlight the need for warnings on medications and prescribing guides to alert practitioners to the potential for increased suicide risk in this age group.
Where did the story come from?The research was carried out by Dr Marc Stone and colleagues from the Center for Drug Evaluation and Research at the US Food and Drug Administration (FDA). The study received no specific grants from any external agencies other than the FDA. The study was published in the (peer-reviewed) British Medical Journal.
What kind of scientific study was this?This review examined the risk of suicidal behaviour in adults taking part in clinical trials of antidepressants. It involved a systematic review of placebo-controlled trials with meta-analysis. The researchers aimed to test the theory that suicidal ideation (suicidal thoughts) or preparatory actions for suicide or worse (attempted or completed suicide), would be increased among adults using antidepressants compared to placebo.
The FDA commissioned the review in 2005-06, when it asked the industry sponsors (such as pharmaceutical companies) of 12 marketed antidepressant drugs to submit data on trials of antidepressants in adults for any indication. Information was requested for all completed double blind, randomised placebo controlled trials. If the sponsors had excluded any trials, they were asked to provide reasons for this.
The sponsors were asked to search their databases for adverse events reported during clinical trials. The various search terms that were used related to suicidality, and may have included examples such as ‘attempt’, ‘burn’, ‘cut’ and ‘jump’. False positives, where these terms had been used but did not relate to suicidality, were also identified.
Sponsors prepared a narrative report of all adverse events, which were classified by a board of expert reviewers into one of several categories:
completed suicide, suicide attempt, preparatory acts towards imminent suicidal behaviour, suicidal ideation, self-harm, intent unknown, not enough information (fatal), and not enough information (non-fatal). For participants who had multiple events, only the most severe event was coded.
The researchers did not include trials that had fewer than 20 participants in each treatment arm, those without sufficient patient data and those that had used active comparison drugs rather than placebo.
The primary outcome was defined as definite suicidal ideation or behaviour, while the secondary outcome was preparatory actions or worse (also called suicidal behaviour).
What were the results of the study?Following exclusions, 372 trials were included in the review, with a total of 99,231 participants. Of these, 295 trials had used antidepressants for psychiatric indications, while the other 77 trials had looked at their use for non-psychiatric reasons. Most of the studies were unpublished and had not been included in previous reviews of antidepressant trials.
The average (mean) age of participants was 43.1 years, 63.1% were female and 86.9% were white. The trials investigated selective serotonin reuptake inhibitors (SSRIs, eight different drugs), tricyclics (five different drugs) and other antidepressants (five different drugs).
Across the trials there were reportedly eight completed suicides, 134 suicide attempts, 10 reports of preparations without attempted suicide and 378 reported suicidal ideation alone, without action.
The researchers first carried out an analysis by medical indication. This showed that suicidality rates were higher in those treated for major depression (341 reports of suicidality) compared to other depressive disorders (22 reports), psychiatric disorders (148 reports) and non-psychiatric behavioural disorders (nine reports).
When the researchers carried out an analysis of suicidality by age group they found a non-significant increased risk of suicidality (either ideation or actual behaviour) in those under the age of 25 (OR 1.62, 95% CI 0.97 to 2.71). However, when they looked at the subcategory of suicidal behaviour alone, the increase in risk for those under 25 became significant (OR 2.30, 95% CI 1.04 to 5.09).
There was a trend for a decreased risk of suicidality in all age groups above the age of 25, but the association was non-significant for the majority of age brackets. When age brackets were combined (25 to 64 year olds), antidepressants had a decreased risk of ideation (OR 0.79, 95% CI 0.64 to 0.98) but had no effect on actual suicidal behaviour. For those aged 65 years and over, antidepressants reduced both ideation (OR 0.37, 95% CI 0.18 to 0.76) and behaviour (0.06, 95% CI 0.01 to 0.58).
When the researchers looked at individual antidepressant drugs, most associations with suicidality were not significant (neither increased nor decreased risk). Across all age groups, the only significant observations were a decreased risk of suicidality with the SSRIs fluoxetine and sertraline.
Combining all age groups, active treatment for psychiatric disorders with any antidepressant drug decreased suicidality with only borderline significance (OR 0.83, 95% CI 0.69 to 1.00). This result was calculated from a total of 314 suicidal eve
nts in 50,043 people treated with an active drug compared to 197 suicidal events in 27,164 treated with placebo (rate 0.63% versus 0.73%). However, in the under-25 group there were 64 events in 4,780 people treated with an active drug compared to 21 events in 2,621 treated with placebo (1.3% versus 0.80%).
When the researchers modelled age as a continuous variable, they observed that the risk of suicidality associated with taking antidepressants decreased at a rate of 2.6% per year of age, and actual suicidal behaviour by 4.6% per year of age.
What interpretations did the researchers draw from these results?The researchers conclude that risk of suicidality associated with antidepressants is strongly age dependent. There is an increased risk for suicidality and suicidal behaviour in adults under 25 treated with active treatment compared to placebo.
They say that antidepressants seem to protective against suicidal ideation in adults aged between 25 and 64, but have no effect on suicidal behaviour, and they reduce the risk of both suicidality and suicidal behaviour in those aged 65 or over.
What does the NHS Knowledge Service make of this study?This reliable and extensive research found an overall trend for any antidepressant treatment to reduce the risk of suicidality in people aged 25 years or above.
In the under 25s, however, there was a non-significant increased risk of suicidal thoughts or behaviour (preparatory actions for suicide or attempted or completed suicide) with antidepressant treatment. When limited to suicidal behaviour alone this increased risk became significant.
These findings are likely to lead to further research and may lead to changes in drug regulatory information. The research highlights the need for warnings on medications and prescribing guides to alert practitioners to the potential for increased suicidality risk among this young age group.
As the authors say, the possibility of separate therapeutic and adverse effects from antidepressant drugs on suicidal thoughts or behaviour requires further investigation, particularly in terms of possible mechanisms for age-related differences.
A few points to note:
The review included any suicidal events that were reported during the treatment phase of trials. However, it is difficult to determine whether this behaviour represented a change in condition or reflected the pre-treatment condition, i.e. suicidal thoughts may have been present before the person commenced treatment and persisted unchanged with treatment, rather than being new onset suicidal thoughts in a person who did not have any prior to treatment. The data were obtained from drug development programmes by drug sponsors. Most of the trials were unpublished. These unpublished trials are valuable as they are unlikely to have been included in previous reviews; however, their methods are not available for critique and as such it is not possible to comment on the quality of these trials. The trial setting may also only provide information from a select population group. As the researchers say, people with severe depression which clearly warranted treatment are unlikely to have been entered into a randomised trial where they could have been allocated to inactive placebo. In general, the trials were relatively short and treatment was given for weeks rather than months or years. Longer term trials may have given different results. Individual trials may have reported adverse events differently. In particular, the reported rate of suicidality may be an underestimate of the true number of suicidal thoughts, as the rate at which people reported these thoughts to researchers may have varied between trials