Furious Seasons

February 04, 2009

Study On Suicide And SSRI Use: Bad For Kids, Good For Adults

There's a new study out in the Canadian Medical Association Journal delving once again into the question of whether there is a connection between SSRI use and suicide and suicidality, the latter only measured as a suicide attempt not suicidal ideation. It is a meta-study of eight large observational studies--so these aren't clinical trials.

The short answer to what the study found is that SSRI use increases the risk of suicide in children, is allegedly "protective" in adults aged 18 to 64, and decreases the risk of suicide in people 65 and older. This study has not received much press yet--probably because there aren't many fulltime journalists remaining who specialize in medicine--and the press that it has gotten stressed the good news on adults pretty uncritically and didn't make much hay about the troubling news on children. The study was accompanied by a disingenuous editorial, co-authored by Robert Gibbons. He authored a paper in 2007 claiming that a decrease in SSRI use earlier this decade had led to an increase in suicides when, in fact, that wasn't the case (and he was making his case based upon one year's worth of data). Gibbons has testified on behalf of Pfizer, makers of Zoloft, on the matter of SSRI use and suicide. I'll come back to his editorial in a moment.

The main study is by Italian researchers affiliated with the World Health Organization. I'm not shocked that a WHO study would come back with at least partially-positive results for anti-depressants. You can read the study for yourself to see where the data came from and decide for yourself whether it's good or bad. (Interestingly, one the studies the WHO authors pulled data from was a Gibbons-led study.)

Conclusion:

"Although exposure to SSRIs increased the risk of completed or attempted suicide among adolescents (odds ratio [OR] 1.92, 95% confidence interval [CI] 1.51–2.44), the risk was decreased among adults (OR 0.57, 95% CI 0.47–0.70). Among people aged 65 or more years, exposure to SSRIs had a protective effect (OR 0.46, 95% CI 0.27–0.79). Sensitivity analyses did not change these findings. In particular, for studies that used completed suicide as an outcome, exposure to SSRIs was associated with increased risk among adolescents (OR 5.81, 95% CI 1.57–21.51) and decreased risk among adults (OR 0.66, 95% CI 0.52–0.83) and older people (OR 0.53, 95% CI 0.26–1.06).
"Interpretation: Based on data from observational studies, use of SSRIs may be associated with a reduced risk of suicide in adults with depression. Among adolescents, use of SSRIs may increase suicidality."

So the basic takeaway from the study is SSRIs are bad for kids and teens but are good for adults and seniors, as measured in terms of suicide or suicide attempts. I know that may run counter to what some readers know about these issues, but, like I've said for a long time, the evidence for suicide and SSRI use is mixed in adults. I've seen evidence in both directions. A lot of the studies use competing patient populations so there's a certain amount of apples-to-oranges going on. As for suicidality, the study's use of an actual suicide attempt as a proxy for suicidality doesn't really tell us much, given the broad range of behaviors and thoughts that comprise suicidality.

I think that there is enough evidence--anecdotal, scientific and clinical--to where only a fool wouldn't acknowledge that there is a significant link of some kind (however strong or weak) to suicide among adults using SSRIs, and that the evidence is very strong for a connection between suicide and SSRI use and kids and teens up to 18 years of age. Need evidence? Check out the FDA black box warning on anti-depressants in kids 0 to 14 from 2004 (and its extension to young adults in 2006) and the British ban on the use of anti-depressants (except Prozac) in kids 0 to 17 from 2003. Check out David Healy's Let Them Eat Prozac. And so on.

Which brings me to a fool.

The editorial accompanying the study is authored by the aforementioned Gibbons of the University of Illinois and John Mann of Columbia University. They use the WHO study to justify a call for more study of the question of kids and SSRI-induced suicide and suicidality. I'd say at this point that this issue really doesn't merit a lot of further study, especially given that the British ban forced down anti-depressant use while Britain's suicide rate was also dropping. It appears that Gibbons in particular will never let this sleeping dog lie. Why it appears to be his central life purpose to defend anti-depressants to the death is beyond me.

Of course, there is the tricky question of what the heck might be driving suicide risk and suicidality for kids and teens who take SSRIs while there's less of signal in adults. I don't even have a guess, but something is surely going on.

The editorial gets even more fun:

"Alarmingly, concerns about the risk of suicide in youth have led not only to fewer SSRI prescriptions without substitution of alternative medications or psychotherapies, but also to a decrease in predicted rates of diagnosis of mood disorders."

Using a shortfall in diagnoses to claim alarm is a bit much. Perhaps, the predictions by some in psychiatry of rates of mood disorders (meaning depression and bipolar disorder) in youth were wrong. What with arguments about the overdiagnosis of depression in adults and the misdiagnosis of bipolar disorder in adults (at almost a 50 percent rate), you have to wonder about the judgment of someone who would cry about rates of diagnosis in children and teens as not being high enough. What makes him think that misdiagnosis might not be a gigantic issue in children and teens? One of the peer reviewers didn't challenge the authors on this point?

One other fun bit:

"The conclusion by Barbui and colleagues that SSRIs are protective against suicidality among young adults challenges the justification of a 'black box' warning about the risk of suicide with the use of SSRIs."

The authors refer to the 2006 black box decision by the FDA which extended warnings on anti-depressants to 15 to 24 year-olds. I don't think it's fair for them to claim that the WHO study's findings challenge the warning. It's a different result from a different set of data and adds to a mixed evidence base. What's more, Gibbons and Mann failed to be specific in their age groups. By young adults, they meant 18 to 24 years olds, but the 2006 warning also covered 15 to 17 years olds. That's an error that I am surprised got by peer reviewers and journal editors.

Posted by Philip Dawdy at February 4, 2009 12:01 AM

del.icio.us Digg it reddit

Comments

Antidepressants can cause bipolar. Just ask Jane Pauley - no, wait, her speaking fee is more than we could afford.
http://www.playingfieldpromotions.com/Jane-Pauley.php

Posted by: Lilly NC at February 4, 2009 03:06 AM

Follow the money trail ...

If you google the lead authors of this new "study" you will find "no competing interests declared" , but it seems the when you google the co-authors (whose names are not mentioned in relation to their part) you find that the co-authors are usually the ones on the pharmaceutical industry payroll...

So therefore , while these studies look like they are independent , they rarely are... here's another study by the same people on the effectiveness on paroxetine (paxil-seroxat) ..
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2211353

Competing interests: None declared for Corrado Barbui and Andrea Cipriani. Toshiaki Furukawa has received research funds and speaker fees from Asahi Kasei, Astellas, Dai-Nippon, Eisai, Eli Lilly, GlaxoSmithKline, Janssen, Kyowa Hakko, Meiji, Organon, Pfizer, Tsumura, Yoshitomi and Zelia. His research has also been funded by the Japanese Ministry of Education, Science and Technology and by the Japanese Ministry of Health Labor and Welfare.

Effectiveness of paroxetine in the treatment of acute major depression in adults: a systematic re-examination of published and unpublished data from randomized trials
Corrado Barbui, MD, Toshiaki A. Furukawa, MD, and Andrea Cipriani, MD

Posted by: truthman30 at February 4, 2009 06:47 AM

Peace be with you

I read these studies, but mostly believe what I actually see around me. As I have said before I live in a rural, California, experimental mental health, county, Humboldt. Our suicide rate went from 9 per 100,000 to 21 since we became "experimental." One person even committed suicide in the parking lot of our mental health department.

Another thing I see is drug overdoses. An adult child of a friend of mine overdosed on heroin two days after being court ordered to take anti-psychotics. It might be that people have reactions with their street drugs and anti-psychotics, or it might be that they are taking their "golden trip" as a result of the suicidal tendencies of these drugs. How obvious is every suicide attempt? How obvious is every suicide for that matter? And what about all the people who say these drugs make them think about suicide?

love eternal
tad

Posted by: tad at February 4, 2009 10:16 AM

Very informative post - thanks.

Posted by: Ward at February 4, 2009 10:27 AM

truthman - yes. technically, if a SPECIFIC study was not funded by a drug company, then the authors can declare we had no conflict in THIS study. Your method is recommended. Just determine the manufacturer of the drug, then google the last name plus the company. Bingo. Done.

Alrming: elderly suicides: on versus off SSRIs: case-control studies have their weaknesses, and I generally don't like them. but this result is intriguing, and certainly would prevent me from believing that antidepressants have no effect in elderly. This design has one benefit: pharmacy records-based assessment of SSRI prescription, versus blood testing on autopsy (if ssri withdrawal leads to suicidality, the drug not no longer be detectable on autopsy):

Juurlink DN, Mamdani MM, Kopp A, Redelmeier DA. The risk of suicide with selective serotonin reuptake inhibitors in the elderly. Am J Psychiatry. 2006 May;163(5):813-21.

OBJECTIVE: The authors explored the relationship between the initiation of
therapy with selective serotonin reuptake inhibitor (SSRI) antidepressants and completed suicide in older patients. METHOD: The authors linked population-based coroner's records with patient-level prescription data, physician billing claims, and hospitalization data for more than 1.2 million Ontario residents 66 years of age and older from 1992 to 2000. For each suicide case, four closely matched comparison subjects were selected using propensity score methods. The authors determined the odds ratio for suicide with SSRIs versus other antidepressant treatment, calculated at discrete monthly intervals from the start of treatment.
RESULTS: Of 1,329 suicide cases, 1,138 (86%) were each fully matched to four
comparison subjects using propensity scores. During the first month of therapy, SSRI antidepressants were associated with a nearly fivefold higher risk of completed suicide than other antidepressants (adjusted odds ratio: 4.8, 95% confidence interval=1.9-12.2). The risk was independent of a recent diagnosis of depression or the receipt of psychiatric care, and suicides of a violent nature were distinctly more common during SSRI therapy. Numerous sensitivity analyses revealed consistent results. No disproportionate suicide risk was seen during the second and subsequent months of treatment with SSRI antidepressants, and the absolute risk of suicide with all antidepressants was low. CONCLUSIONS:
Initiation of SSRI therapy is associated with an increased risk of suicide during the first month of therapy compared with other antidepressants. The absolute risk is low, suggesting that an idiosyncratic response to these agents may provoke suicide in a vulnerable subgroup of patients.

PMID: 16648321

Posted by: MedsVsTherapy at February 4, 2009 10:33 AM

Personally, I think these studies have conflicts of interest written all over them. I must say it is a clever tactic they are using now...
Just because something has the guise of one thing doesn't mean it isn't harboring another agenda...

Posted by: truthman30 at February 4, 2009 10:51 AM

I can't help but suspect that the increase in children/adolescents vs. decrease in adults/elderly is probably due to the fact that they're looking at attempts and actual suicides, NOT ideation. Let's assume that ideation is increased equally in all age groups - children and adolescents are (obviously) less mature emotionally and physiologically, particularly in terms of frontal lobe development. So even if ideation increased across the board, children and adolescents are, for lack of better phrasing, more likely to succumb to these new, powerful, painful thoughts and feelings. Particularly since we know from other studies you've posted that doctors routinely do not discuss side effects or the increased vulnerability of the first two months when they write a new antidepressant script. So they don't have that warning in advance, they don't have the cognitive processing ability of adults, they are given to being taken by their emotions anyhow at that age, have a poor conception of life/death/forever like most kids and teens do, and therefore actual suicides and attempts are higher for their age group than for adults, who probably have more effective support structures than your average teenager.

Just a theory. And of course that would be assuming that the numbers from the study are accurate, which is far from certain.

Posted by: Jordan at February 4, 2009 12:19 PM