Risk of suicidal behaviour in adults taking antidepressantsBMJ 2009; 339 doi: https://doi.org/10.1136/bmj.b3066 (Published 11 August 2009) Cite this as: BMJ 2009;339:b3066
All rapid responses
Despite the cynicism of others, the FDA and the BMJ should be
congratulated on brining to the attention of a wider medical community the
age related relationship between suicidality and SSRI antidepressants.
Despite the disadvantages of this type of study, as pointed out in
both the paper and in the accompanying editorial, the fact that younger
people are clearly at greater risk than adults, is an important message.
If for no other reason it should act as a warning that SSRIs should be
used with a great deal of caution in adolescents and that younger adults
should have increased surveillance. All doctors who wish to prescribe
SSRIs to these patients should clearly follow, in the interests of safety,
at least the recommendations of the manufacturers concerning their use by
suitably qualified professionals, their use at the lowest possible dose,
frequent assessment in the early stages of treatment and a consultation
with someone other than the patients, especially in adolescents, to warn
them to look for any signs of increasing suicidality. In adolescents,
they should probably be used after a trial of psychotherapy and in
conjunction with it.
If this publication does no more than take this message to one
physician, good will have been done.
I have provided support to parents of children admitted to the Capio
Hospital’s Adolescent Mental Health Unit while there was early debate on
the risks to children and to have such clear evidence of an age
relationship is helpful to them in deciding if they wish their children to
be prescribed such drugs. Reinforcement by clinicians to parents and
other carers of the risks and the need to closely observe their loved ones
has to be encouraged.
As many have pointed out, this is only the start of answering many
questions in relationship to this matter. These of course include ‘Are
there differences between the various SSRIs?’ and ‘How do we distinguish
prospectively those patients who will have decreased suicidal ideation and
those who may have increased suicidal ideation when prescribed these
As important is further work to ascertain the risk benefit ratio when
SSRIs are prescribed clinically to those whom are actively excluded from
GCP conducted registration quality studies. They are clearly a different
sub-population due to inclusion bias by investigators, the protocoled
exclusion of those with severe depression and suicidality and their age
Of equal importance to suicidality, is the question as to whether
SSRI drugs induce patients to behave aggressively, violently or to commit
homicide. I wonder whether there were enough cases of violence in the FDA
database to begin to answer this question. There are unlikely to be
anywhere near the number of homicides required.
The accompanying editorial by Geddes et al contains some important
observations; however, it is disappointing that they used their status and
the opportunity to write an editorial on an important subject to quote
their own work and to promote the use of Sertraline over and above other
It is to be regretted that they were allowed to state without
reference to other SSRIs that ‘the odds of suicidal behaviour on
Sertraline for example is around half that on placebo’ without giving any
reference or even noting the populations studied, or without looking for
an age relationship as was clearly shown by the FDA. Such statements
cannot move forward such a complex area which has needed the study of over
100,000 patients by the FDA. Whether or not Sertraline is better
tolerated was not the point of the FDA review and thus comments on this
were completely inappropriate in the editorial, even if this is true,
which many would dispute.
Dr Malcolm VandenBurg BSc MBBS MISMA FCP FFPM FRCP
26 Christchurch Hill, London NW3 1LG
Tel: 020 7435 9386/07850 049 134
Dr Malcolm VandenBurg has given many medico legal opinions to UK courts in legal cases in which SSRIs were implicated, instructed by both defence and prosecution.
Competing interests: No competing interests