Black box warnings imposed by the
FDA to flag suicide risk in adolescents who take the
commonest class of antidepressants may have backfired
badly, suggests new research that finds suicides rose
as prescriptions fell.
Reporting in the September American
Journal of Psychiatry, researchers found that in
2004, a 22% decrease from the previous year in prescriptions
of selective serotonin reuptake inhibitors to patients
aged under 19 coincided with a 14% increase in youth
suicides. The research also looked at Holland, where
prescriptions fell by 22% in the period 2003-5. This
was accompanied by a startling 49% increase in suicides
among the young.
Even the smaller rise in US rates
is highly significant, representing the largest single-year
change since nationwide suicide reporting began in 1979.
It reverses a long-term decline in suicides among the
young, which had previously fallen 28% since 1990.
A total of 4,599 Americans aged under 24 took their
own lives in 2004, making it the third leading cause
of death in this age group. The overall US suicide rate
is about 32,000 in a typical year.
Professor Robert Gibbons, PhD,
of the University of Illinois at Chicago, lead author
of the study, says he's anxiously awaiting 2005 prescribing
and suicide data. The FDA only announced its decision
to put the warnings on SSRIs in March of 2004. He fears
an even bigger effect will be seen in the following
year's data. "The really major fall in prescribing came
just after the period we studied," he says.
He notes that in Holland, the rise
in suicides was proportionately much greater than the
decline in SSRI prescriptions. "My conjecture is that
patients deemed most at risk of suicide were those most
likely to be taken off the drug, leading to a selection
effect," he says.
The data appears to knock major holes in what had rapidly
become conventional wisdom on SSRIs, particularly since
it is based on actual suicides in the population. No
young person actually killed themselves in any of the
studies which led the FDA to link SSRIs and suicides.
Rather, the agency found a consistent pattern of increased
suicidal ideation, self-harming behaviour and other
surrogate markers for suicidality.
But the researcher who first proposed
a link between SSRIs and suicidality in the young, Professor
David Healy of the University of Cardiff, is not willing
to call it quits yet. He begins by challenging the accuracy
of the data, noting that suicides are only recorded
when a coroner says it's suicide.
"As suicide rates have fallen in
recent years, autopsy rates have also fallen and have
fallen so dramatically that it is quite plausible that
this is the reason for fewer diagnoses," he says. "Against
a background of recognition of SSRI adverse effects,
leading possibly to suicide, it's conceivable that deaths
formerly noted as uncertain or accidental have been
appropriately coded instead as suicides now, where they
Professor Healy has always maintained
that SSRI treatment involves two periods of sharply
elevated risk - the beginning of treatment, and withdrawal.
The new data leaves open the possibility, he notes,
"that there are in fact no fewer people on SSRIs. A
drop in the script rate may stem primarily from those
on treatment chronically 80% of those being treated
- and a constant or even increased rate of new patients
being put on the drugs." These new patients would be
at high suicide risk, he says.
Then there are withdrawals. It
is actually quite consistent with Professor Healy's
theories that a large number of people coming off SSRIs
would translate into a bump in suicides.
Finally, he says, "there has also
in the last two or three years been a huge switch from
antidepressants to antipsychotics. The antipsychotic
'mood stabilizers' have just as great an increase in
suicide risk as antidepressants - if not greater."
Professor Gibbons counters: "I understand where Professor
Healy is coming from but I don't think the data supports
it." He points to research published last month in the
American Journal of Psychiatry by his co-author
Dr Robert Valuck of the University of Colorado, which
found no corresponding increase in antipsychotic prescriptions
as SSRI prescriptions fell.
He adds that SSRI prescriptions
continued to rise in 2003-4 in one age group - the over-60s
- and this was the only age group in which suicides
Professor Gibbons also notes that
Dr Valuck's data shows "that the FDA warnings were followed
by a fall in the number of actual diagnoses of depression
among the young." This raises a third possibility not
yet discussed - that the rise in suicides may not be
a pharmacological effect at all. Could it be that the
FDA's questioning of the most obvious tool for treating
depression led doctors to undertreat their patients,
leaving the vulnerable young feeling more helpless and
abandoned than ever? That kind of effect can be very
hard to measure.
If there is a pharmacological effect, help may be on
the way. Thomas Insel, director of the National Institute
of Mental Health (NIMH), suspects that SSRIs are beneficial
in most patients, but are causing suicidality in a small
minority who may be genetically differentiated from
the rest of the population. Research sponsored by NIMH
has already found a gene test that can predict response
to the SSRI citalopram, and more results are on the
In the meantime, the FDA is watching
the situation closely, and plans a big review later
this year. Canada's Therapeutic Products Directorate,
which issued similar warnings, told NRM they
have no plans to change the monograph, which they consider
to be up to date.
For the record, Canada's youth
suicide rate actually fell slightly in 2004, the most
recent year for which statistics are available.