Medication for depression – Is it all in the mind?

by Martin

in Depression Medications

An interview with Jay Fournier, leading researcher into the effectiveness of anti-depressant medication

In a previous piece on anti-depression medication, I looked at the findings of some recent studies that seem to call into question the effectiveness of modern medication for depression.

The Kirsch study

Some of the leading work in this field has been carried out by Professor Irving Kirsch, principally reported in his 2008 paper, Initial severity and antidepressant benefits: A meta-analysis of data submitted to the Food and Drug Administration and in his book “The Emperor’s New Drugs”.

Professor Kirsch’s research looked at the results of clinical trials for a number of anti-depressant drugs and compared the extent of improvement in depressed patients who were given the medication with that of patients who were given placebos.

What he found was that:

“Drug-placebo differences increased as a function of initial severity, rising from virtually no difference at moderate levels of initial depression to a relatively small difference for patients with very severe depression, reaching the criteria for clinical significance only for patients at the upper end of the very severely depressed category.”

Even for those patients in the ‘very severe category’, the drug-placebo differences were found to be relatively small.

In the Emperor’s New Drugs, Professor Kirsch describes the controversy caused by his 2008 paper and the resistance to its findings from some quarters of academia and medical practice. Many commentators took the view that Kirsch’s findings simply flew in the face of what they could readily observe in clinical practice – that antidepressants do work.

I was therefore very interested in some further research on the subject carried out at the University of Pennsylvania by Jay Fournier Ph.D. and others.

The Fournier study

The 2010 Fournier study, “Antidepressant Drug Effects and Depression Severity – A Patient-Level Meta-analysis”, was published in the Journal of the American Medical Association. It concluded as follows:

“The magnitude of benefit of antidepressant medication compared with placebo increases with severity of depression symptoms and may be minimal or nonexistent, on average, in patients with mild or moderate symptoms. For patients with very severe depression, the benefit of medications over placebo is substantial.”

In the light of the controversy surrounding Kirsch’s work, and to the extent that this research seemed to back up his Kirsch’s conclusions, this seemed to me to be a pretty significant finding. I wanted to know more about this.

I therefore contacted Dr Fournier and he very kindly agreed to answer some questions that I put to him.

anti-depressant medication for depression
How effective is medication for depression? Dr Jay Fournier gives us some insights

The Interview


First, I was interested in the differences between your study and that of Kirsch et al. You state in your paper :

“Despite differences in methods, our findings are consistent with those of both Kirsch et al and Khan et al that ADM/placebo differences increase as initial severity increases. We used individual patient data and included patients with less severe depression, whereas both Kirsch et al and Khan et al analyzed group means that largely excluded patients with HDRS [Hamilton Depression Rating Scores] scores below 20.”

Could you just briefly explain the differences in your methodology and that of Kirsch et al and Khan et al? Perhaps more particularly, can you explain how the methodology used in your study takes this line of research further?

Jay Fournier

Both Kirsch and Khan conducted what are referred to as meta-analyses, in which they combined data (at the group level) from a number of studies in order to examine the question. Essentially, each study that Kirsch and Khan included in their analyses provides a small amount of data: the means for each treatment group, the variation around those means, and the number of participants involved.

Our study approached the question in a different way. Instead of obtaining the summary values from each study we included, we went back and asked for the data from each individual participant in each study.

This allows for a much more fine-grained analysis of the relationship between initial severity and treatment outcome. An additional advance in our study is that we were able to obtain data from clinical trials that had included individuals with lower baseline severity scores than were available in either the Kirsch study or the Khan study. Again, this allows for a more careful examination of the relationship between baseline severity and treatment outcome – over a wider range of baseline severity scores.


What I don’t understand is that if Kirsch’s study excluded patients with Hamilton Rating Scale for Depression (“HDRS”) scores below 20, then doesn’t that mean that his conclusions related to the lack of efficacy of anti-depressants on severe cases of depression, which sounds to me like a different outcome to your study?

Jay Fournier

Actually, that is not how I would interpret his findings. It is standard practice in pharmaceutical trials to exclude patients with scores below 20. That was the data to which Dr. Kirsch had access for his analyses. What he found was that the medications outperformed placebo for those depressed patient at the higher end of the severity spectrum. We found the same thing, despite our different methods.

I believe the cut-off for a clinically significant difference in his study was an initial severity score of 28. Using the same threshold he used, our estimate was an initial score of 27. So, both studies found that true drug-placebo differences only emerge at the higher end of the depression severity spectrum.

Because our study was able to include data from participants with lower intake severity scores, we were able to test the claims made by Kirsch that the drug-placebo difference was negligible for patients with lower scores (scores that fell below the range to which Kirsch had access in the data that he used). Indeed, that is what we found.


In light of your research, are you able to provide any other comments on the effectiveness and use of anti-depressants and the current debate on the subject that I highlighted in the previous post

Jay Fournier

Nicely done on the blog. I agree with much of what you say.

Again, I don’t necessarily see our results as contradicting those of Kirsch and the others who have come before. I think the overall pattern is remarkably similar. The medications do not seem to offer an added benefit over placebo for patients in the mild and moderate range.

It is only patients with more severe depressions who seem to receive the added benefit from the active ingredients of the medications.

What we attempted to point out in the paper, is that given what we know from other research, the majority of patients who are prescribed medications in the community likely do not have severity levels high enough to register in the range of severity in which the medications were shown to have specific benefit.

That said, I think it’s important to be careful with what we take away from these findings. One of the unfortunate messages that was reported in some media outlets as a result of the paper was that antidepressants don’t work.

I think that’s inaccurate. They clearly work (in that they reduce symptoms of depression), it’s just that for a large number of depressed individuals, placebos work just as well.

This is an important point. For a large number of depressed patients, elements of the placebo treatment (e.g., taking one’s illness seriously, speaking with a mental health professional about it, following through with treatment recommendations, etc.) provide sufficient benefit that the active ingredients of the medications cannot provide anything over and above those placebo effects.


In his book Professor Kirsch calls into question the idea that anti depressants operate to correct some kind of chemical imbalance in the brain. I’ve also read (for example in Elliot Valenstein’s book Blaming the Brain) that there is actually little solid evidence to support the theory that mental disorders are even caused by such chemical imbalances. Do you have any views upon this based upon your research or otherwise?

Jay Fournier

I hesitate to comment on the chemical-imbalance theory directly, as it is not the focal part of my expertise. What I will say is that there is almost certainly something that is different about the functioning of the brain of an individual who is depressed compared to that of a non-depressed person.

There is growing evidence that both psychiatric medications AND particular psychotherapies, such as cognitive behavioural therapy, can have an impact on brain function – and can reverse some of the functional abnormalities that are observed during a depressed state.

I think the truth is that we have much more to learn about the functioning of the brain during a depressive episode and how treatments work to correct or augment those processes.


What would you say to anybody who has recently been prescribed anti-depressants or who is thinking about ceasing to take their anti-depressant medication?

Jay Fournier

I think people need to exercise caution when thinking about changing and or stopping their medications. Although the medications may not outperform placebo at treating depression at lower levels of depression severity, they do have a profound impact on the brain, and they can affect other things (e.g., anxiety, neuroticism, see the Tang paper, I refer to below).

I think seeking and following the medical advice of a doctor is crucial. If I were such a patient, I would want to know why my doctor thought an antidepressant was indicated, and I would be curious as to how she reached that decision and whether she considered other alternatives.

What our results show, I think, is that it is important for persons who are suffering from depression to take an active role in their care, regardless of the severity of their symptoms. That is, even placebo treatment helped most patients. When considering which treatment option to pursue, it would be important to consider other relevant research findings, for example the ability of certain psychotherapies to help the patient buttress him/herself against relapse.

Finally, I would urge patients to reconsider (with their doctors) any of their treatment decisions if they continue to struggle with symptoms after attempting a particular treatment and allowing it a reasonable amount of time to work.

For a list of empirically supported psychotherapies for depression and other conditions, see this list from the Society of Clinical Psychology.

I also highly recommend two additional papers. I think they are quite relevant to the questions you pose:

Selective Publication of Antidepressant Trials and Its Influence on Apparent Efficacy, Erick H. Turner, M.D., Annette M. Matthews, M.D., Eftihia Linardatos, B.S., Robert A. Tell, L.C.S.W., and Robert Rosenthal, N Engl J Med 2008; 358:252-260 January 17, 2008

Personality Change During Depression Treatment: A Placebo-Controlled Trial Tony Z. Tang, PhD; Robert J. DeRubeis, PhD; Steven D. Hollon, PhD; Jay Amsterdam, MD; Richard Shelton, MD; Benjamin Schalet, MA Arch Gen Psychiatry. 2009;66(12):1322-1330


I’d like to thank you for taking the time to answer these questions. I think it’s really helpful for the layman patient to understand as much about their condition and its treatment as possible. It’s therefore really valuable to be get some direct responses like this from somebody who has undertaken the core research. And thanks also for researching in this area. It seems to me that there is still a long way to go as far as depression and medication is concerned.

Jay Fournier

I agree! My hope is that all of these recent papers and books have revitalized the quest to find effective and lasting treatments for depression.


Kirsch, I., Deacon, B. J., Huedo-Medina, T. B., Scoboria, A., Moore, T. J., & Johnson, B. T. (2008), Initial severity and antidepressant benefits: A meta-analysis of data submitted to the Food and Drug Administration (2008).. PLoS Medicine, 5(2))

Kirsh, I, The Emperor’s New Drugs: Exploding the Antidepressant Myth

Khan A, Leventhal R, Khan S, et al. (2002), Severity of depression and response to antidepressants and placebo: an analysis of the Food and Drug Administration database. J Clin Psychopharmacol 22: 40–45

Jay C. Fournier, MA; Robert J. DeRubeis, PhD; Steven D. Hollon, PhD; Sona Dimidjian, PhD; Jay D. Amsterdam, MD; Richard C. Shelton, MD; Jan Fawcett, MD (2010), Antidepressant drug effects and depression severity – A patient-level meta-analysis JAMA.2010;303(1):47-53. doi: 10.1001/jama.2009.1943

Elliot Valenstein, Blaming the Brain: The Truth About Drugs and Mental Health

Main image: Jerry Butler in Brockwell Park singing “Are You Happy?” by Martin Grover

Pills Image: posterize /

Go from this page to the main ‘Depression Medications’ category page
Go from this page to the ‘Depression Help’ Home page

About the Author

Father, husband, writer and website publisher, discontented in his day-job, he writes here about depression - his own and in general. You can follow Too Depressed on Twitter. Please share the content on this site with all your friends, followers and contacts using the buttons above.

Contact the author

Last revised on February 22, 2012

{ 2 comments… read them below or add one }

Leave a Comment

Previous post:

Next post: