PsychEducation.org (home)                                                                                                                                

Antidepressants and Suicidal Thinking: A Biologic Marker? 
(Written 11/2007)

The United States Food and Drug Administration (FDA) raised a huge controversy when it required a "black box warning" for antidepressants, indicating that these medications might increase the risk of suicidal thinking.  Initially this warning applied to children and adolescents; later it was extended to young adults.  (For more details, see my summary page on the actions of the FDA).

Since the warning was required, many psychiatrists have openly criticized the FDA for raising this concern.  Antidepressant prescriptions for children and adolescents have decreased about 30% since the warning was placed.  Substantial evidence, though not conclusive, suggests that since the warning, suicide rates may actually be higher -- perhaps because of this decreased antidepressant use (e.g. Gibbons). Many psychiatrists have called for a repeal of the warning, fearing that it may have done more harm than good.

However, in October 2007 an important study was published that suggests the fears of the FDA are warranted but also raises hope for being able to identify those who are at risk for this low-frequency but high-danger consequence of antidepressant use. This study looked at a very large number of patients who were enrolled in a study of antidepressants in patients with depression, conducted by the National Institutes of Mental Health (NIMH). Although their results need some form of replication before too much weight is placed on this, the researchers seem to have found two genes that may be associated with antidepressant-induced suicidal thinking.  First, the results; then some thoughts about the implications thereof --

Laje and colleagues at the NIMH: study results

This research team looked at genes associated with five major signaling pathways of the brain: dopamine, serotonin, norepinephrine, glutamate, and neurotrophins -- 68 particular genes in all. (I find it a hopeful sign that such searches do not have to look blindly at the DNA anymore, but instead can be directed specifically at relevant targets -- this shows how far this research has already come). They found that two versions of genes which code for glutamate receptors were dramatically more common in the few subjects from this study who had the new onset of suicidal thinking within the first month of antidepressant use.  This is such an important advance, such a big step forward in how we can think about problems like this, that I think you should see the experimental result itself:

Two different gene markers stick out of the noisy jumble of results (nice demonstration of a good "signal-to-noise-ratio", if you know that term). One of them lies in the gene (GRIK2)  which codes for a glutamate receptor (GluR6).  The other (GRIA3) lies in a gene which codes for a different glutamate receptor (AMPA3).

Patients with the risky version of the GRIK2 gene were 8 times more likely to develop suicidal thinking when given the antidepressant citalopram. Those with the risky version of the GluR6 gene were twice as likely as those without it to have this response. Patients unlucky enough to inherit both were 15 times more likely to develop suicidal thinking than those with neither of these risky gene versions. Note that none of these patients had a history of suicidal thinking:  these gene variations seem to be associated with developing such thinking on antidepressants, not with suicidal thinking itself.

A second important findings from this study is the frequency of suicidal thinking: 120 out of 1862 participants (6%) developed such thoughts.  Note that these were adults, not children.  For complex reasons, it is thought that children are at greater risk for developing this problem than adults.  A study of similar design would theoretically find more suicidality developing, if there is indeed a causal relationship involved here. Unfortunately, this study was designed to examine other questions, so there was no placebo control group against which to compare. Nevertheless, several experimental features make it more likely that this suicidal thinking was related directly to the medication: for example, over 90% who developed suicidal thoughts did so within one month of starting the antidepressant. None of the patients had suicidal thinking previously. Suicidal thinking was carefully defined and examined.

Therefore, in my opinion, these results strengthen the concern that antidepressants can indeed cause suicidal thinking in susceptible individuals.  We now even have a clue as to who those individuals might be.  Indeed, another study of similar design found additional  risk markers in a different set of genes, and a similar timing in the onset of suicidal thinking in those at risk.Perlis

Implications

Is the day coming, then, when we will do gene testing before we give someone an antidepressant? Several reasons come to mind to worry about this approach, but here's one that I hope would get plenty of attention:  we do not have unlimited resources to spend on healthcare.  Already there are thousands of people in the United States who do not have access to basic medical treatment. Do we want to spend our money doing gene analysis before using antidepressants?  Or can we save that money and do something else with it, and simply warn patients to watch out for emerging suicidal thinking when taking these medications?  True, using the gene method,we might be able to prevent a rare case in which such thinking led to an actual suicide (there were no suicides in the study described above. There were two attempts.  Only one had gene data.  He was positive for both risky genes.).  But might we prevent more suicides by making health care available to more individuals?  Remember, we now have data suggesting that use antidepressants can actually lower suicide rates.  Maybe we can save more lives by making sure that all those who need to can see a doctor, and find the few among them who are at risk of suicide (with or without an antidepressant) by providing a healthcare system that allows them to be seen again within a few weeks rather than a few months.

 

(One of the authors on the study was my hero at the NIMH, Husseini Manji. Check out his work on the biologic basis of depression; here's a start)