Was Kraepelin right?

The classic diagnostic validators in psychiatry are symptoms, course of illness, genetics, and biological markers. Treatment response or pathophysiology often are used as alternatives to biological markers. The basic idea is that there is no gold standard, like pathology, to validate clinical diagnoses in psychiatry. So instead, we can use these four or five independent lines of evidence to validate a diagnosis, if they point in the same direction. DSM is based purely on symptoms, but symptoms of various diseases often overlap, hence the need for independent non-symptom validators.

One could apply classic diagnostic validators to Emil Kraepelin’s basic dichotomy of dementia praecox (schizophrenia) versus manic-depressive illness (MDI: bipolar illness plus unipolar depression).

Let’s take these one by one:

1. Genetics: Much of the current genome-wide association (GWAS) studies find a good deal of overlap between schizophrenia and bipolar illness, as defined by DSM-IV and 5. But they do not find them to be identical. Thus, they do not invalidate the Kraepelinian dichotomy, although at least in those with psychosis (which is a minority of bipolar illness), there is overlap. It should be noted that the same GWAS studies find notable overlap between bipolar illness and DSM-defined “major depressive disorder” (which is a conglomeration of many different depressive states). Thus, genetic research today can be seen as refuting the DSM-III/IV/5 anti-Kraepelinian refusal to combine bipolar illness and “MDD” together as one disease (manic-depressive illness).

2. Pathophysiology: A common error is to view similar pathophysiology as arguing for one diagnosis or disease. This is the false assumption of the anti-Kraepelinian tradition, as well aso o the current National Institute of Mental Health Research Domain criteria (NIMH-RDoC). The falsity of this idea has to do with the fact that the body functions as a final common pathway for many disease etiologies. Thus, streptococcal pneumonia and pneumococcal pneumonia both have a similar pathophysiology – lung inflammation – yet are different diseases. So to schizophrenia and MDI have common pathways in the limbic system and with monoamine or glutamate neurotransmitters, but that does not mean they are the same diseases or diagnoses. The same principles apply to cognitive sequelae.

3. Treatment response: This is by far the weakest validator of a diagnosis, because drugs are nonspecific. Yet many clinicians focus a lot on this factor. Drugs are nonspecific: aspirin works for 27 different diseases, as does propranolol. So we really should not emphasize this validator. But speaking to the examples given, lithium works in MDI, meaning both unipolar depression and bipolar illness; dopamine blockers (so-called “antipsychotics”) work for mixed states and thus for both depressive and manic symptoms – and thus both examples support Kraepelin’s concept of MDI, as opposed to the DSM anti-Kraepelinian bipolar/MDD dichotomy.

On the issue of the supposed efficacy of dopamine blockers in maintenance efficacy of bipolar illness, most clinicians and researchers take the word of the FDA and the pharmaceutical industry uncritically. It is an unfortunate fact that most of the profession still does not realize that the claim to maintenance efficacy with randomized discontinuation trials (RDTs) is invalid scientifically. In RDTs, patients are preselected to respond already to a drug, and then they are assessed again for response. If you like chocolate cake, but not vanilla cake, you are randomized to chocolate cake again or vanilla, and, not surprisingly, chocolate cake always wins. There never has been a RDT which disproves efficacy: whatever drug is “enriched”, always works. If you preselect for acute placebo response, placebo also is better than any drug in maintenance treatment of anything. (References are provided below.) In short, the claims about dopamine blocker efficacy in bipolar illness based on RDTs cannot be believed; those agents are not proven effective because the RDT design upon which such claims are based is invalid.

In summary, the newest research, properly analyzed, supports Kraepelin’s basic nosology, by and large, and it contradicts the DSM-III to 5 anti-Kraepelinian nosology.


SN Ghaemi, Clinical Psychopharmacology, 2019, Oxford University Press.

Ghaemi and Selker 2017:

Goodwin et al 2011:

334 views2 comments

Recent Posts

See All

Research and Practice: Twins, not Opposites

I never have understood clinicians who see clinical research as unrelated or even opposed to what they do. This attitude – that clinical research and clinical practice are opposites – even is enshrine

Suicide and lithium skeptics

Question: In sharing with a colleague your textbook on Psychopharmacology, she pointed out that for lithium to have any anti-suicidal effect, it must be used at least six months and that there is an