Overinterpretation of SSRI study results: Halaris et al. 2015

Halaris and colleagues published a paper in the Journal of Psychiatric Research in which they studied the impact of the SSRI antidepressant escitalopram (ESC) in a group of 30 depressed patients. Only 20 participants completed the trial, and there was no placebo group.

Halaris et al. 2015

The authors tracked the level of a number of inflammatory markers and neurotoxic markers over the course of 12 weeks, along with depressive symptoms. They find that neurotoxic markers and symptoms decrease, but not inflammatory markers.

This leads the authors to conclude:

  1. “The results indicate that ESC may exert its antidepressant effect in part through inhibition of synthesis of certain neurotoxic kynurenine metabolites and possibly also through reduction of the inflammatory response.”
  2. “Escitalopram induced a high remission rate but no change in inflammation biomarkers.”

kirk what

As Captain Kirk has already figured out, these conclusions are quite problematic.

  1. It is absolutely unclear if ESC actually had an antidepressant effect on the 20 patients at all – many studies have found that antidepressants do not outperform placebo. Meta-analysis usually estimate the response rate of depressed patients to antidepressants at around 35%, while the placebo response lies around 30% (see also here and here). Only if ESC is statistically related with the symptom reduction do the subsequent interpretations make sense – which the authors could not show.
  2. There is no evidence that reduction in neurotoxic markers is associated with ESC – such a conclusion would require a placebo group. Yet, the authors interpret it causally, stating ESC “induces” changes in biological markers.
  3. The authors conclude that changed markers may be moderators between ESC and antidepressant response (“ESC may exert its antidepressant effect in part through …”). There is no evidence for this, since the analyses are purely correlational.
  4. I am confused that the authors believe their results may indicate reduced inflammatory markers, since they do not find reduced inflammatory markers.

Additionally, the newsletter of the American Psychiatric Associations quotes the authors with the following statement:

“[…] one clear observation is that, at week 8, escitalopram monotherapy induced reduction of both clinical symptom scores and inflammatory and neurotoxic biomarkers.”

If this is not a mistake on the side of the journalist, the statement is both factually false (no reduction of inflammatory markers was discovered) and highly speculative (there is no actual evidence for ESC “inducing” anything).

Let me recycle the following section from a previous post:

Psychiatry is desperate, and understandably so, especially when it comes to depression research. After half a century of drug development, the best antidepressants only marginally outperform placebos. After over 3 decades of biomarker research, there is no single biomarker robustly associated with depression diagnosis.

But this despair does not justify lowering scientific standards regarding antidepressant research.


Halaris, A., Myint, A.-M., Savant, V., Meresh, E., Lim, E., Guillemin, G., … Sinacore, J. (2015). Does escitalopram reduce neurotoxicity in major depression? Journal of Psychiatric Research, 66, 118–126. doi: 10.1016/j.jpsychires.2015.04.026. (URL)

3 thoughts on “Overinterpretation of SSRI study results: Halaris et al. 2015

  1. Pingback: Ketamin as promising treatment for suicidal thoughts? | Eiko Fried

  2. Pingback: Adjunctive Nutraceuticals for depression treatment | Eiko Fried

  3. Pingback: Clinical trials continued: Adjunctive Brexpiprazole without control group | Eiko Fried

Leave a Reply

Your email address will not be published. Required fields are marked *