Re: Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study
Dr. Lu et al. recently published study[1] makes an important contribution to our understanding of the complex relationship between antidepressant prescription and risk for suicide.
The reported trend in increased psychotropic drug poisonings following the suicidality black box warning for adolescents is an interesting observation. The authors interpreted their finding as indicative of increased risk for suicidal behavior due to decreased antidepressant use[1].
However, the study has a few important limitations that need to be carefully considered.
There are a number of concerns about the reliability of psychotropic overdoses as proxy measure for all suicidal behavior [2,3]; but even if we accept this measure the study data remains difficult to interpret as it does not answer the question about who were the individuals who ended taking psychotropic overdoses.
First, the authors imply that under-diagnosing depression might have lead to increased overdoses on psychotropic medication. However it is unlikely that under-diagnosis is at issue here. If the black bow warning somehow resulted in increasing physicians’ reluctance to diagnose depression this will likely decrease the likelihood to receive not only an antidepressant but other psychotropic medications too. In other words there is a low likelihood that patients not receiving antidepressants somehow ended up receiving prescription for other psychotropics on which they later on overdosed.
Secondly, the individuals who ended up overdosing might be patients who discontinued antidepressants due
to the black box warning. The interpretation of increased suicidality in a patient discontinuing an antidepressant is complex. If increased suicidality ensues in this population one needs to consider the following:
1. psychopharmacological mechanisms that might result in an relapse in suicidality previously well controlled presumably due to an effective antidepressant regimen
2. psychopharmacological mechanisms that might result in an increase in new onset suicidality due to withdrawal
3. psychological mechanisms that might result in an relapse in suicidality previously well controlled presumably due to fear of depression/suicidality relapse
Thirdly, the individuals who ended up overdosing might be patients who continued antidepressants despite the black box warning. The interpretation of increased suicidality in a patient continuing an antidepressant is similarly complex. If increased suicidality ensues in this population one needs to consider the following:
1. psychological mechanisms that might result in an increase in suicidality previously well controlled presumably due to fear of depression/suicidality relapse
2. psychological mechanisms that might result in an increase in new onset suicidality presumably due to a self-fulfilled prophecy of increased suicidality in light of the new black box announcement
3. from an epidemiological standpoint an increase in the reporting of “suicidal” overdoses in patients taking antidepressants (overdoses that prior to the warning might have been consider accidental and be under-reported)
Last but not least it is possible that the rate of depression with suicidality has simply been on the raise over the study period in a way that does not correlate with the rate of antidepressant prescription. Does depression in the United Stated of America culture, characterized by media’s attention to sensational news commonly including violent themes among other things, independently decrease the threshold for acting on suicidal ideation once one is depressed? Time will tell; for now this is yet another alternative explanation for Lu’s at al. US based epidemiological findings.
Lu and colleagues findings can reverberate from the first line of treating physicians who might be reconsidering the way they prescribe antidepressants to the level of policy makers who might be revisiting guidelines and polices about antidepressant recommendations.
Due to its importance, Lu and colleagues far reaching conclusions need to be considered in the light of the many limitations we discussed. While interesting, this study’s results raise many questions, which only further prospective controlled trials of antidepressant specifically targeting suicidality can answer.
Adrian Preda, M.D.
University of California Irvine
References:
1 Lu CY, Zhang F, Lakoma MD, et al. Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study. BMJ 2014;348:g3596–g3596. doi:10.1136/bmj.g3596
2 Gotzsche PC. Re: Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study | BMJ. BMJ. 2014.http://www.bmj.com/content/348/bmj.g3596/rr/702751 (accessed 20 Jun2014).
3 Olfson M, Schoenbaum M. Uncoupling antidepressant use and psychotropic drug poisoning. BMJ. 2014.http://www.bmj.com/content/348/bmj.g3596/rr/702742 (accessed 20 Jun2014).
Rapid Response:
Re: Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study
Dr. Lu et al. recently published study[1] makes an important contribution to our understanding of the complex relationship between antidepressant prescription and risk for suicide.
The reported trend in increased psychotropic drug poisonings following the suicidality black box warning for adolescents is an interesting observation. The authors interpreted their finding as indicative of increased risk for suicidal behavior due to decreased antidepressant use[1].
However, the study has a few important limitations that need to be carefully considered.
There are a number of concerns about the reliability of psychotropic overdoses as proxy measure for all suicidal behavior [2,3]; but even if we accept this measure the study data remains difficult to interpret as it does not answer the question about who were the individuals who ended taking psychotropic overdoses.
First, the authors imply that under-diagnosing depression might have lead to increased overdoses on psychotropic medication. However it is unlikely that under-diagnosis is at issue here. If the black bow warning somehow resulted in increasing physicians’ reluctance to diagnose depression this will likely decrease the likelihood to receive not only an antidepressant but other psychotropic medications too. In other words there is a low likelihood that patients not receiving antidepressants somehow ended up receiving prescription for other psychotropics on which they later on overdosed.
Secondly, the individuals who ended up overdosing might be patients who discontinued antidepressants due
to the black box warning. The interpretation of increased suicidality in a patient discontinuing an antidepressant is complex. If increased suicidality ensues in this population one needs to consider the following:
1. psychopharmacological mechanisms that might result in an relapse in suicidality previously well controlled presumably due to an effective antidepressant regimen
2. psychopharmacological mechanisms that might result in an increase in new onset suicidality due to withdrawal
3. psychological mechanisms that might result in an relapse in suicidality previously well controlled presumably due to fear of depression/suicidality relapse
Thirdly, the individuals who ended up overdosing might be patients who continued antidepressants despite the black box warning. The interpretation of increased suicidality in a patient continuing an antidepressant is similarly complex. If increased suicidality ensues in this population one needs to consider the following:
1. psychological mechanisms that might result in an increase in suicidality previously well controlled presumably due to fear of depression/suicidality relapse
2. psychological mechanisms that might result in an increase in new onset suicidality presumably due to a self-fulfilled prophecy of increased suicidality in light of the new black box announcement
3. from an epidemiological standpoint an increase in the reporting of “suicidal” overdoses in patients taking antidepressants (overdoses that prior to the warning might have been consider accidental and be under-reported)
Last but not least it is possible that the rate of depression with suicidality has simply been on the raise over the study period in a way that does not correlate with the rate of antidepressant prescription. Does depression in the United Stated of America culture, characterized by media’s attention to sensational news commonly including violent themes among other things, independently decrease the threshold for acting on suicidal ideation once one is depressed? Time will tell; for now this is yet another alternative explanation for Lu’s at al. US based epidemiological findings.
Lu and colleagues findings can reverberate from the first line of treating physicians who might be reconsidering the way they prescribe antidepressants to the level of policy makers who might be revisiting guidelines and polices about antidepressant recommendations.
Due to its importance, Lu and colleagues far reaching conclusions need to be considered in the light of the many limitations we discussed. While interesting, this study’s results raise many questions, which only further prospective controlled trials of antidepressant specifically targeting suicidality can answer.
Adrian Preda, M.D.
University of California Irvine
References:
1 Lu CY, Zhang F, Lakoma MD, et al. Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study. BMJ 2014;348:g3596–g3596. doi:10.1136/bmj.g3596
2 Gotzsche PC. Re: Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study | BMJ. BMJ. 2014.http://www.bmj.com/content/348/bmj.g3596/rr/702751 (accessed 20 Jun2014).
3 Olfson M, Schoenbaum M. Uncoupling antidepressant use and psychotropic drug poisoning. BMJ. 2014.http://www.bmj.com/content/348/bmj.g3596/rr/702742 (accessed 20 Jun2014).
Competing interests: No competing interests