washingtonpost.com  > Live Discussions > Health

Drugs and Risk of Suicide

Analysis of Data Adds to Concerns On Antidepressants

Dean Fergusson
Friday, February 18, 2005; 11:00 AM

Adults taking popular antidepressants such as Prozac, Paxil and Zoloft are more than twice as likely to attempt suicide as patients given sugar pills, according to an analysis released Thursday of hundreds of clinical trials involving tens of thousands of patients.

The new study was conducted by epidemiologist Dean Fergusson and his colleagues at the Ottawa Health Research Institute and included scientists from McGill University.

Fergusson was online Friday, Feb. 18, at 11 a.m. ET to discuss the findings of the study and explain the significance of the tests.

A transcript follows.

Editor's Note: Washingtonpost.com moderators retain editorial control over Live Online discussions and choose the most relevant questions for guests and hosts; guests and hosts can decline to answer questions.


washingtonpost.com: Dean Fergusson, welcome to washingtonpost.com. How does one balance the results of this analysis with the need certain individuals have for taking an antidepressant drug?

Dean Fergusson: That's a good question. Undoubtedly, these drugs bring benefit to many people with debilitating conditions. But, like any drug, it has its risks. And although the risks are rare these are commonly prescribed drugs which make it a public health issue. The study certainly does not conclude that people currently taking these drugs stop taking them. Instead we need to reinforce close monitoring for those who take them and make treating physicians aware of the uncommon risks.


Washington, D.C.: How does your study account for the idea that if someone is being treated for depression they might be pre-disposed to suicide in the first place? Is there a way to predict which patients might commit suicide while on these drugs? I have been taking an SSRI for depression and it's really helped me a lot. It did just what it was supposed to -- keeps me on an even keel and I am able to face the day better and am more willing to do the work of getting over my depression. So I'd hate to see these drugs get an even stronger stigma than they already have when so many people benefit from them.

Dean Fergusson: We evaluated published randomized controlled studies the control groups consisted of either placebo, tricyclic antidepressants, or other therapies (e.g. exercise, psychotherapy). The vast majority of studies excluded patients with suicidal thoughts when assessing eligibility for the trials. As for your last question, I agree, the risks need to be put in the right perspective and balanced with the benefits.


Michigan: Could you tell us what medications were studied in the research? Is it just Prozac, Zoloft and Paxil, or does it include newer drugs too such as Celexa and Wellbutrin? Thanks.

Dean Fergusson: The five drugs were Prozac, Zoloft, Paxil, Luvox, and Celexa. Wellbutrin was not assessed.


New York, N.Y.: Most reports of suicidal ideation on SSRI therapy have historically occurred in severely depressed patients who appear energized enough to make an attempt they were not capable of before therapy. Shouldn't these patients then be more intensely managed on SSRI therapy rather than the mild-to-moderates as you suggest.

Dean Fergusson: In our study we found consistent rates of increased odds of suicide attempts across three categories of patients: major depression, mild to moderate depression, and "other." The "other" category includes a host of different conditions such as panic disorder, anxiety disorders, sexual dysfuntion etc. This suggests that all patients be monitored by treating physicians.


New York, N.Y.: Why should this particular meta analysis be given more attention than the dozens of other reviews that have been undertaken in the U.S. and Europe, with different conclusions?

Don't these publications just increase the risk that fragile patients will independently terminate their therapy and become unprotected from suicidality at the rates we experienced before SSRIs?

Dean Fergusson: The difference is that our study includes many more patients and studies. When looking for uncommon risks, we need to evaluate many many patients. I believe the earlier meta-analyses found the same trend yet lacked "statistical" power to rule out chance. Our study presents the most extensive evidence to date, uses observations in published trials, and uses a conservative definition of suicide attempt (i.e the authors of the study had to report it as a suicide attempt). As for obseravtional studies assessing trends, they are prone to many biases.

As for your last question, the message should be strongly stated that patients do not stop therapy. However,if they have concerns, speak to their physician.


Anonymous: I'm curious as to how long the individuals taking SSRIs were on the medication before being suicidal. Is it possible that going on those medications just for a short while will have an effect? Or were the study participants only surveyed after they had been on the medication for more than, say, a month? Thank you.

Dean Fergusson: Unfortunately, the studies do not provide this data. Of the 702 trials, the average study period was about 10-11 weeks. We did not look at studies with study duration less than a week. There were very few studies with long-term follow up periods and this should be an avenue for future research.


Pittsburgh, Pa.: Was there any analysis of suicides among those suffering from bi-polar disorder?

Dean Fergusson: We did not analyse those with bi-polar disorder separately. We do have those trials identified but the number of patients studied is too low to garner any meaningful estimates.


Washington, D.C.: Are the SSRIs more likely to be associated with suicidal thoughts and actions than the older TCAs? And did any of the studies you looked at specifically screen the participants for suicidal thoughts or actions vs. waiting for spontaneous reports?

Dean Fergusson: We found no difference in suicide attempts between SSRIs and tricyclic agents. We did not evaluate suicidal ideation or thoughts. As for screening, the vast majority of studies did not enroll patients with suicidal thoughts or risk. In other words, they were not eligible for the study.


Bethesda, Md.: Wouldn't those suffering from depression have a much higher rate of suicide whether they were on drugs or not? Isn't there also a HUGE amount of evidence that these drugs have helped many more people come out of depression and thus prevented many more suicides? Sorry to sound harsh, but this study sounds like bunk to me ...

Dean Fergusson: You raise a very good point. It is very hard to tease out the effects of the disease form the effects of the drug. Depression is linked to suicide and we tried to ascertain whether SSRIs were as well. To control for this, we only examined randomized controlled trials which are the most valid method for determining risk or benefit of a drug. When evaluating the best quality evidence, we found an increased risk in attempts on SSRIs compared to placebo across different types of patients (those with major depression, mild to moderate depression, and those with conditions other than depression). The effect was consistent across all groups.


washingtonpost.com: The American Psychiatric Association says that drug-induced suicide fears are vastly exaggerated. Would you please comment.

Dean Fergusson: We need to keep in mind that the risk is small but the risk is shared by millions of people prescribed SSRIs. Thus, on an individual level, the risks are rare but across the population taking SSRIs this results in quite a few events. As for the results, we are confident in our conclusions. By no means is this definitive evidence. We need large randomized controlled trials with long follow-up periods to end the debate.


Arlington, Va.: Do you have any thoughts on the recently failed "Zoloft Defense"?

Dean Fergusson: No, I do not. This is out of my area of expertise.


Fairfax, Va.: Did you find any correlation between subject age and likelihood of attempting suicide, or did your data not permit this analysis?

Dean Fergusson: We did not specifically look for age. When mean trial age was divided between those less than 60 years of age and those greater than 60 years of age, the results were consistent.


Dean Fergusson: I've enjoy the opportunity to answer questions. In closing, the results need to be put in proper perspective. We have highlighted that SSRIs may be associated with an increase in suicide attempts. The event is uncommon but it is remains an issue as prescriptions are common. These drugs do bring benefit to many people but, like all drugs, there may be risks. All benefits and risks of therapy need to be considered to make informed decisions.


© 2005 Washingtonpost.Newsweek Interactive
Viewpoint: Paid Programming

Sponsored Discussion Archive
This forum offers sponsors a platform to discuss issues, new products, company information and other topics.

Read the Transcripts
Viewpoint: Paid Programming