Tuesday, May 5, 2015

The Heuristic is Dead - Long Live The Heuristic

One of the latest non-scandals to rock monolithic psychiatry and psychiatrists everywhere is David Healy's commentary on serotonin and depression (1).  Certainly nothing new there from the Healian perspective - SSRIs don't work, SSRIs don't work for melancholic depression, no evidence that SSRIs raise or lower serotonin levels, SSRIs pushed tricyclic antidepressants out of the market, and SSRIs forced the abandonment of research on biological disturbances in depression like hypercortisolemia.  The only thing I did not see was the idea that SSRIs are somehow addicting because they cause discontinuation symptoms in a subgroup of people who take them.  Feel free to call that a withdrawal syndrome, but withdrawal does not constitute an addiction.  All of these premises allow him to reject serotonin and it's role in depression on the vaguest possible grounds while stating that it is not irrelevant:

"Serotonin is not irrelevant.  Just as with noradrenaline, dopamine, and other neurotransmitters, we can expect some correlation with personality and temperament."

Indeed - but I expect those to be a very weak correlations probably not on par with the mood disorder work.  I was glad to see his reference to the late Marrku Linnoila (2) or more specifically Linnoila and Virkkunen.  Linnoila had done some of the outstanding work in this area and work on serotonin metabolites (CSF HIAA) in alcohol use disorders remains a classic.  He talks about the serotonin based research running into the sand.  I did a Medline search of serotonin and major depression and plotted the papers per year and there does not seem to be any precipitous fall off (the 2015 references are through March).  If anything I would expect the research to increase with the availability of ligands for additional receptors and the further characterization of serotonin transporter (SERT) variants.

But the main problem with Healy's argument is that he is talking about serotonin as if there is a chemical floating around in the brain and that is it.  He is basically describing his own brand of chemical imbalance theory.  So if he is saying there is no serotonin chemical imbalance theory I would of course have to wholeheartedly agree with him.  And in fact, I would tell him the same thing I told the Prozac rep back in the 1980's:  "The brain is far more than a bag of chemicals.  Unless you can say something about brain structure and physiology naming a neurotransmitter is meaningless."  There is no such thing as a chemical imbalance theory and reading through the index of any psychopharmacology text printed in the past 30 years will confirm that.

Healy can't stop himself at that level.  He goes on to describe the impact of his chemical imbalance theory on clinicians:

"This history raises a questions about the weight doctors and others put on biological and epidemiological plausibility.  Does a plausible (but mythical) account of biology and treatment let everyone put aside clinical trial data that show no evidence of lives saved or restored function?"

Things get very tenuous at that point.  As a clinician who has made a career out of treating patients with the most severe problems  I don't care at all about the "biological and epidemiological plausibility" of a theory of a medication.  I got past that in the 1990s when the elegant Nobel laureate mechanism of action of aminophylline was debunked and the former first line drug for asthma and COPD exacerbations was suddenly relegated to tertiary status.  And even then, with the people I was running up on gurneys from the emergency department while calculating the parameters of their aminophylline drip - not a patient of mine was lost.  I suppose you could say it was all an elaborate placebo response, treating all of those people with acute shortness of breath bad enough to be brought in by paramedics.  I was after all using a tertiary treatment with a disproven molecular mechanism.  But really?  And my living, breathing pulmonary patients did not seem to mind.

The only thing that matters to me is if the medication works in my hands, has few side effects, and I am using it exclusively to save lives and restore functioning.  And without keeping anyone in suspense, I do prescribe SSRIs and they do work, and I do fully acknowledge to anyone willing to listen that the mechanism of action is unknown. Not surprisingly many patients want more than that and I can discuss speculative mechanisms as well as the next psychiatrist.  I don't think that makes me a drug company stooge or an idiot, because I am the only person with any accountability in the entire scheme of things.  It really doesn't matter to me what David Healy or anybody else thinks.  I am seeing very ill people and it is my job to get them better and not cause them side effects with the medication and not have the drug interact with one of their underlying medical conditions.  When you get right down to it, I don't need a mechanism of action to use a medication - only the approval of a regulatory body like the FDA.  At some point somebody may say that  SSRIs are no better for depression than aminophylline was for asthma, but so far (apart from Healy and various sympathizers) that has not happened.  I just attended a CME conference last weekend and they were still recommended.  As far as I know the FDA has not sent out any letter to doctors telling them not to prescribe them any more.

Healy also asks the irrelevant question:  "Do clinical trial data marketed as evidence of effectiveness make it easier to adopt a mythical account of biology ?"  Or I guess the real question is do these data make it easier to accept this strawman?  I strongly encourage any psychiatrist to read and reread the FDA approved package insert on any drug they prescribe and any drug their patient happens to be taking.  I encourage reading and rereading those package inserts especially if they are updated and sent out in a mass mailing by the FDA.  These package inserts need to be studied because they are not the final word in safe prescribing.  If you are an experienced psychiatrist you will routinely be called on to decide if a medication (based on all of the available evidence) is safe to prescribe to a patient with cardiac, liver and renal disease.  You will have minimal data to go by in most cases, because people with significant medical illness are typically eliminated from clinical trials.   Any psychiatrist reading those package inserts will recognize that for some time now, the package inserts have contained clinical trials data and that data is very useful.  It's utility has nothing to do with putative biological mechanisms, it has to do with safely monitoring and prescribing the medication.  The other consideration implicit in this question is that clinical psychiatrists are unaware that clinical trials are really a primitive technology.  Given that awareness exists, why would anyone accept clinical trials data as proof of much of anything.  How in the world does a double-blind placebo controlled study with an intent to treat analysis have anything at all to do with my practice of psychiatry?  It is totally irrelevant clinically and even less relevant theoretically.  Its only use is to get a new drug in my hands that my patients and I will accept as useful or reject as too toxic or worthless.

Why are reasonable people interested in serotonin despite Healy's commentary?

The complexity of serotonin or more appropriately serotonergic signaling for one.  Absolute levels of neurotransmitters are rarely the issue in complex disorders.  In the case of serotonin any bare bones discussion needs to consider the fact that there are 14 serotonin (5-HT) receptor subtypes coded by 18 genes.  Those receptors are organized into 7 families, 6 of which are G-protein receptor linked superfamily.  The remaining family is part of the ligand-gated ion channel superfamily.  The combinatorics of these receptor types, their specific locations, and their genetic variants can lead to a dizzying number of signaling variations that can easily be found in psychiatric research (4).  As an example, this is from a 2013 review of serotonin signaling in depression and suicide by Mann (3):

".......Part of the neurobiology observed in recurrent major depression, that is present between episodes as well as during episodes, is a series of abnormalities in the serotonergic system [19]. Cerebrospinal fluid (CSF) levels of the main serotonin metabolite, 5-hydroxyindoleacetic acid (5-HIAA), are low in more lethal suicide attempters, and predict the risk for future suicide with an odds ratio of 4.6 [20].  Convergent evidence comes from some reports of low CSF 5-HIAA in suicide attempters with other diagnoses such as schizophrenia [21,22], bipolar disorder [23] and personality disorders, although there are fewer studies and less consensus in these other disorders [24,25]. Nevertheless, if correct, such findings suggest that less serotonin transmission, as implied by less 5-HIAA, appears to be associated with suicidal behaviour across psychiatric diagnostic boundaries, as would be expected if it were associated with the diathesis for suicidal behaviour instead a specific psychiatric disorder. A review of post-mortem brain studies of suicides has found much the same thing: most studies find suicides have lower levels of serotonin and/or 5-HIAA in the brainstem serotonin neurons compared with psychiatric-matched groups [26–28]."
Apart from all of the disputed technical details, what is the importance of the Healy commentary?  It follows the political tradition of other commentaries mentioned on this blog by prominent psychiatrists.  People have sent me e-mails and asked me: "What do you mean by political?"  In the simplest analysis it is taking one end of a polarized position and running with it and bringing everything that you can to support that position.  That is the classic way that politicians work.  They are fueled by ideology.  On the surface, it seems that science operates like that as well, but ongoing scientific arguments don't include comments about marketing or words like "neurobabble".  They don't make assumptions about how clinicians think or whether they are confusing science and hypthothesis testing with clinical work.  The best science encompasses complexity and contradictory data and we are past the point where neuroscience can be watered down for the masses in a 2 minute sound bite.

Neurobabble happens when a neuroscience discussion is reduced to a sound bite and debated the way that Democrats and Republicans discuss just about anything.  That is what happens when we are all polled "Serotonin - yes or no?"

If you really want to get your opinion out in the press that way that is one thing.  Let's not pretend that psychiatrists really think this way or that psychiatrists are just flunkies for Big Pharma talking about some crazy chemical imbalance theory or that we based our clinical practice on heuristics or for that matter marketing.

Psychiatrists are really bright people, time to stop pretending that we are not.

 George Dawson, MD, DFAPA


1: Healy D. Serotonin and depression. BMJ. 2015 Apr 21;350:h1771. doi: 10.1136/bmj.h1771. PubMed PMID: 25900074.

2:  View my collection, "Marku Linnoila" from NCBI

3: Mann JJ. The serotonergic system in mood disorders and suicidal behaviour. Philos Trans R Soc Lond B Biol Sci. 2013 Feb 25;368(1615):20120537. doi: 10.1098/rstb.2012.0537. Print 2013. Review. PubMed PMID: 23440471

4:  KEGG.  The Serotonergic synapse.

5: Albert PR, Benkelfat C. The neurobiology of depression--revisiting the serotonin hypothesis. II. Genetic, epigenetic and clinical studies. Philos Trans R Soc Lond B Biol Sci. 2013 Feb 25;368(1615):20120535. doi: 10.1098/rstb.2012.0535. Print 2013. PubMed PMID: 23440469

6: Albert PR, Benkelfat C, Descarries L. The neurobiology of depression--revisiting the serotonin hypothesis. I. Cellular and molecular mechanisms. Philos Trans R Soc Lond B Biol Sci. 2012 Sep 5;367(1601):2378-81. doi: 10.1098/rstb.2012.0190. PubMed PMID: 22826338


  1. Great piece!

    Healy has unfortunately gone off the deep end about SSRI's - he's starting to sound like Ssazz or Breggin. Anyone who has prescribed them to patients who meet the criteria for MDD (and not just dysthymia) knows they work about as well as tricyclics for most patients, with fewer side effects.

    There's a huge difference between "widespread clinical experience in multiple settings" and "clinical anecdote," but you'd never know that from listening to everyone using the scientific-sounding but increasingly absurd term "evidence-supported medicine" nowadays. Psychotherapy researchers are doing it too.

    Clinical trials vs widespread clinical experience by doctors who know the difference between a true drug effect versus a side effect like sedation? I know which "evidence" I think is more powerful.

    1. Isn't Dr. Healy saying the same thing but for different outcomes in his reply here:


    2. In another thread I brought up the triptans for migraines. Turns out that psychiatry was not the only discipline promoting serotonin depletion as an easy explanation for disease:



      These are all over the place and not hard to find in a Google search.

    3. And in the process we also learned that 5-HT1x receptor agonists were potent causes of chronic daily headache and that for many people Excedrin or ibuprofen worked as good. Damn - isn't that a scandal that we can blame on monolithic neurology?