Saturday, November 12, 2016

Artificial kinds of mental disorder

Tsou (2016) makes a case for natural kinds of mental disorders. This seems to be dependent on his claim that there is "good evidence that the symptoms of schizophrenia are underwritten by stable neurobiological mechanisms". By this he means the dopamine theory of schizophrenia. Similarly he says "multiple lines of research indicate that the core signs of depression are underwritten by stable neurobiological mechanisms". Again, by this he means the monoamine hypothesis of depression. I guess if the dopamine and monoamine hypotheses are incorrect, his whole argument fails.

He does acknowledge the article by Kendler and Schaffner (2011) that I have mentioned before (eg. see previous post). This apparently doesn't undermine his faith in the dopamine hypothesis. The monoamine hypothesis is similarly not tenable (eg. see previous post). Even psychiatrists have described the 'chemical imbalance theory' as a kind of urban myth (see previous post and book review). Psychopharmacologists stopped believing in it a long time ago.

Tsou wishfully hopes that "mental disorders should be classified at a level of generality such that the characteristic signs of disorders are associated with stable biological mechanisms". His lack of clinical and scientific experience does not provide a sound basis for his philosophical theorising. Psychiatric diagnosis is inevitably purely descriptive. Considering the motivation for the revision of DSM-5 (see eg. previous post), information about the biological causes of mental disorders would have been incorporated if there is any, which there isn't.


Jonathan Tsou said...

Thank-you for discussing my article here Duncan.

Two points of response:

(1) I am optimistic about the dopamine hypothesis (at least more optimistic than Kendler and Schaffner are), although I don't think that dopamine will be the whole story to be told about the etiology of schizophrenia (e.g., glutamate systems are undoubtedly going to be important). All I want to commit myself to is the claim there are certain causal regularities in dopamine processes involved in schizophrenia (e.g., excessive dopamine activity in the mesolimbic pathway is responsible for psychotic symptoms, deficient dopamine activity in the mesocortical activity is responsible for negative symptoms). I think that it's important to acknowledge these causal regularities (or biological mechanisms) in schizophrenia because they will be relevant for treatment using pharmacological drugs. I've discussed these issues more comprehensively in my 2012 article, "Intervention, Causal Reasoning, and the Neurobiology of Mental Disorders: Pharmacological Drugs as Experimental Instruments" (Studies in History and Philosophy of Biological and Biodemical Sciences). The short story is that while I agree with much of Kendler and Schaffner's analysis in their 2011 paper (whose work I greatly admire), I think that they are working with an outdated philosophical ideal of evidence in reaching their skeptical conclusion that there is no evidence for the dopamine hypothesis.

(2) As I argue in my 2016 paper and elsewhere, my view that "mental disorders should be classified at a level of generality such that the characteristic signs of disorders are associated with stable biological mechanisms" is based on a a theory of the projectability of psychiatric classifications, viz., psychiatric classifications can only make reliable predictions if classifications capture a "natural kind" associated with certain causal regularities. At present, I think that DSM classifications are not reliably doing this and classifying some disorders (e.g., anoxeria nervosa) at too specific of a level to allow for reliable inferences about members of a kind (e.g., inferences about the prognosis of the disorder or about which treatment interventions to pursue).

Also, I'd like to note that my advocacy of a theoretical/ causal approach to psychiatric classification does not imply that we should get rid of the DSM's descriptive approach. Ultimately, I advocate an approach that integrates descriptive categories with a constraint that requires that DSM classifications ought to individuate classes of abnormal behavior that are underwritten by biological mechanisms. So it's a descriptive and theoretical (hybrid) approach that I defend; not a removal of descriptive categories. I of course accept responsibility for not expressing this point clearly enough in my paper.

Duncan Double said...

Thanks for your response, Jon.

Of course schizophrenia is a natural kind of mental disorder if the dopamine hypothesis is true. And you are right the hypothesis comes from the dopamine blocking action of the traditional neuroleptics. Still, the theory doesn't necessarily follow from this drug action. And, the atypical antipsychotics could be said to undermine the hypothesis. For example, aripiprazole is a partial dopamine agonist rather than blocker.

It doesn't seem right to me that a philosophical issue is reduced to an empirical question. And maybe both of us are too a priori in our belief about the neurobiological basis of schizophrenia. Still, I don't think you can say that the dopamine hypothesis, or the neurobiological basis of schizophrenia in general, is any more than a theory. The evidence is against it.

I'm sorry, you will just have to accept the uncertainty of psychiatric classification. I'm not one of the people that concludes psychiatric diagnosis should be abandoned because of this situation (see previous post). And anyway, what's more important in assessment is formulation of the reasons for people's problems.

Jonathan Tsou said...

Hello Duncan,

The action of atypical antipsychotics do not undermine the dopamine hypothesis (only a simplistic version of the hypothesis that asserts the schizophrenia is caused by excessive dopamine activity). From a historical point of view, second-generation atypical antipsychotics (which are not neuroleptics) demonstrated that the dopamine hypothesis needed to be expanded to account for the neurobiological basis of negative symptoms, not that we should reject the dopamine hypothesis. Contemporary proponents of the dopamine hypothesis hold that the positive symptoms of schizophrenia are the result of excessive activity in the mesolimbic dopamine pathway, while negative symptoms are caused by deficient activity in the mesocortical dopamine pathway; this latter mechanism was discovered through interventions with atypical antipsychotics. I discuss this in my 2012 paper on schizophrenia:

I agree that the dopamine hypothesis is a theory; however, I think that it’s a theory supported by the evidence. For me it’s not just an empirical question, but an empirical questioned coupled with assumptions regarding what counts as scientific evidence. I suspect we may ultimately disagree on this issue. In my 2012 paper, I defend a philosophical account of evidence (“robustness”) that I think is amenable to actual scientific practices (e.g., it can account for the fallible nature of scientific knowledge), rather than just an a priori philosophical account. Basically, this theory maintains that a theory is well-confirmed (“robust”) when multiple independent findings all support a single theory (or subset of theories) over rival theories. I argue that the dopamine theory is robust in this sense (e.g., it is support by various pharmacological interventions, brain-imaging studies, cognitive task studies, animal studies).

I strongly agree with what you say in your last paragraph. I’d add that the neurobiological underpinnings of schizophrenia are just one causal factor involved in schizophrenia, but it is a particularly important causal factor insofar as we can directly intervene with patients on that level using drugs. We may disagree about the benefits of using antipsychotics to treat individuals suffering from schizophrenia. While I do not think that these drugs are unproblematic, I think it is difficult to deny that they allow some people to live vastly improved lives.

Duncan Double said...

Jon, I'm glad you don't accept a simplistic hypothesis that schizophrenia is due to dopamine hyperactivity. You're right I don't think your interpretation of the empirical evidence is as robust as you think (eg. see Moncrieff, 2009). You're also correct that I'm more sceptical about the benefits of antipsychotic medication.

I was just surprised to see a philosophical paper so rigorously defending the biological basis of mental illness from "chemical imbalance" theories. As I said in my previous comment, I'm sure DSM-5 would have included this perspective if it thought it could 'stick its neck out' and agree with your conclusions. You seem to go beyond the speculations of biological psychiatrists. I think we can agree the evidence for the dopamine hypothesis is contradictory. I just worry that you go beyond this situation to conclude that mental disorders can be natural kinds. I think the philosophical arguments are against you.

Richard Hassall said...

I agree that the philosophical arguments for natural kind status of mental disorders are not strong. Any attribution of such status to a concept in advance of evidence supporting its explanatory role is just speculation, and it's sometimes difficult to see what might motivate such speculation. Richard Boyd, in his papers, emphasises repeatedly that natural kinds are only identified a posteriori - he also emphasises that they have to "accommodate our concepts to the causal structure of the world". If 'schizophrenia' can be shown to do that, and in particular if the dopamine hypothesis is strongly supported by evidence, then we have good grounds for considering it a natural kind. But that's obviously the crux of the matter, as Duncan's comments indicate. In other words, the concept has to do some non-trivial explanatory work. Many argue that the concept 'schizophrenia' does not do this. That's a separate issue, of course, from it's pragmatic use in everyday clinical contexts.

Interestingly, it's usually only philosophers who talk much about natural kinds. Most scientists just get on with the job of sorting out which concepts have explanatory value in their science and which don't.

Dr. med. Hans Bangen said...

Hello Duncan,

just wondering how do you know about Mr. Tsous "lack of clinical and scientific experience"?

Dr. med. Hans Bangen

Duncan Double said...

I agree it was mere assumption and may be nothing to do with why he holds the views he does. What I was meaning is that coming at this issue as a mere philosopher may not be the best experience but I'm sure there are plenty of clinicians and scientists who agree with him. I should try to avoid such potentially pejorative remarks.