One of the most contentious topics in psychiatry is the concept of disease. It might seem odd that a medical specialty should debate the concept of disease, which is so central to medicine itself. Even in the ancient Hippocratic perspective, it was held that the art of medicine had three parts: the doctor, the patient, and the disease. To deny the disease concept would be to deny scientific medicine [1]. But for two millennia, most physicians, following Galen, did just that. Galen said there was only one disease: variations of abnormalities of the four humours. One could not be more specific. Further, the one physical disease of the humours differed from person to person since the specific mixes of the four humours that were abnormal could vary infinitely in different persons. Galen “individualized” diagnosis for each person.

Psychiatry today is Galenic, not Hippocratic. The four humours have become a half dozen neurotransmitters, whose rise and fall we speculatively manipulate with drugs. Careful clinical observation and nosology of disease, the hallmark of Hippocratic thinking, have been replaced by penny-in-the-slot drug-for-symptom practice. This pseudoscience is justified on humanistic grounds as being individualized to the patient. We forget that such extremist individualization, which is the opposite of science, produced 2000 years of dehumanizing, harmful bleeding and purging.

It would take the Enlightenment for physicians to begin to rethink this rejection of disease, and return to the central role the idea held in the Hippocratic vision. Morgagni in the 17th century made the classic case for disease as reflecting a pathology of an organ in the body which would be the same in all individuals. Humours were not involved; individual differences need not matter. If you have cirrhosis of the liver, it looks the same in the king as in the pauper, in the male as in the female.

Virchow later codified Morgagni's view in the notion that medical disease involves organ pathology expressed in clinical syndromes. Kraepelin took up this mantle. After a century of detailed French nosology based on clinical symptoms, without much progress in corresponding pathology, Kraepelin made the guess that organ pathology would match better with the clinical course, not symptoms per se.

In the intervening century, under the distorting influence of time, psychiatrists have often reproached Kraepelin unfairly, saying that a century of research has proven him wrong. We have not found the pathology of dementia praecox or manic-depressive insanity, as he defined them, or as redefined later in schizophrenia and bipolar/unipolar illnesses. The reproach is unfair because Kraepelin proved to be correct in two major diseases: Alzheimer's dementia (named after Kraepelin's colleague who worked on those who had a chronic course beginning in old age, unlike dementia praecox, which began in young age), and general paralysis of the insane, which proved to be neurosyphilis in Kraepelin's later years, and was completely cured by penicillin within two decades after his death [2].

It is true that the two other major “disease processes”, schizophrenia and manic-depression, have not been definitively proven to be diseases based on clear pathology (as in Alzheimer's dementia) or clear etiology (as in neurosyphilis). We can say, however, that after one hundred more years of research, a huge biological database has been created that confirms a major biological pathogenesis and probable biological etiology to both schizophrenia and manic-depression: ventricular enlargement, white matter abnormalities, amygdalar enlargement, hippocampal atrophy, second-trimester infections, and 80% heritability replicated in dozens of twin studies [3, 4]. This literature is not small, and it is consistent. We do not have the etiologies, but we do not have them for Alzheimer's dementia or migraine or epilepsies or hypertension or lupus either.

In this sense, I think Kraepelin has been proven correct: there are diseases of the mind, and schizophrenia and manic-depression are among them.

It is important to appreciate that Kraepelin did not classify diseases only; his approach, not unique to him, was to view diagnoses as two basic types: disease processes (Krankheitsprozessen) and clinical pictures (Zustandsbilden) [5]. If we can scientifically validate a diagnosis – meaning we can delineate it from other diagnoses based on the classic validators of symptoms, course, genetics, biological markers, and/or treatment effects [6] – then we can say there is a clinical picture. To take the next step to claim a disease process, we would have to do research on that clinical picture and find a large amount of biological pathophysiology or biological etiology or both. This has been done more or less successfully with a few conditions: schizophrenia, manic-depression, obsessive-compulsive disease, autism. They are psychiatric diseases. But many clinical pictures may be scientifically valid, and yet not represent disease processes: they don't have major biological pathophysiology and etiology. These include: substance abuse and alcoholism, hysteria and its variants (post-traumatic stress illness, borderline personality), antisocial personality, neurotic depression and its variants (the many anxiety “disorders” of DSM), simple phobias, attention-deficit/hyperactivity disorder, anorexia/bulimia, grief, and extremes of personality traits (neuroticism, extraversion/introversion, risk-taking) [6]. These latter clinical pictures may have a biological component, but they also have environmental components that are equal if not larger (unlike the psychiatric diseases above) [7, 8]. Environmental trauma is a prerequisite to post-traumatic stress. The vast majority of persons with borderline personality have sexual trauma as a major etiology, a social cause [9]. These clinical pictures are legitimate as clinical pictures, but illegitimate as diseases.

A major problem with contemporary psychiatry is that, after DSM-III in 1980, American nosology refused to distinguish disease processes from clinical pictures. By claiming to be “atheoretical”, the term “disorder” was applied to all 300–400 diagnoses, so that clinicians and researchers are unclear as to what is what [10].

Red skies are not red apples; they are different things, despite sharing redness. But bipolar “disorder” is seen as similar to borderline personality “disorder”, partly because the word “disorder” puts them at the same ontological plane, ignoring the fact that one condition is almost completely genetic, while the other is less than half genetic, and that one condition has a huge biological pathophysiology and appreciable animal modeling, while the other has a large social etiology, much more limited biological pathophysiology, and zero animal modeling [6].

Red skies are not red apples. The term “disorder” has confused our profession to the point that often we do not call diseases those conditions which are, and we often call diseases those which are not. Or, more commonly, we just reject the concept of disease, and practice as we like, justifying it, if asked, by biopsychosocial eclecticism [11].

DSM-III onward has produced a system that is proudly called “pragmatic” [12] by its leaders, but which reflects in fact an abdication of scientific responsibility. We reject the disease concept, or we apply it indiscriminately. Either way we do not take it seriously. Two generations of mostly failed biological research in etiology, pathogenesis, and pharmacology cannot be laid at the feet of Nature, for creating mental illness to be so complex that we fail to understand it. We should be willing to blame ourselves, for artificially making up “pragmatic” diagnoses without a serious attempt to try to understand Nature, to identify when diseases are present and when they are not. Van Praag warned the profession two decades ago, just before DSM-IV was produced [13]. Our prior leaders in DSM-III and DSM-IV did not appear to be aware of this problem [14], and now they simply close their eyes to it.

Pragmatism has led to our current eclecticism, where psychiatrists practice as they wish, based on their personal opinions and dogmas, rather than practicing as scientific knowledge guides them. We cannot obtain that level of scientific knowledge until we take the disease concept seriously in psychiatric diagnosis.

This disease-oriented approach does not mean that we will presume that all psychiatric diagnoses are discrete diseases, as many have criticized Kraepelin for presuming. Van Praag [13] is correct that dimensional definitions may be more appropriate for some conditions, like extremes of personality. But this is an empirical, not a conceptual, matter. Let us do the scientific work and go where the data lead us, sometimes to categorical diseases, sometimes to dimensional extremes of the norm.

It is important not to get nihilistic, as have some biological researchers [13], and some postmodernist-oriented critics of biological psychiatry [15]. Critics will point to the tortuous history of psychiatric diagnosis, and conclude that all diagnostic classifications are doomed to fail because mental illness is too complex: it is biopsychosocial [11], or “hybrid” [15], or multidimensional [13]. Some psychiatric clinical pictures do not represent simple diseases, certainly, but the claim that none ever does is disproven by history [2, 16]: neurosyphilis was indistinguishable in many of its phases from what we see today in bipolar illness or schizophrenia. It was complex, polysymptomatic, and variable. And yet it was caused by a single pathogen.

Thus, I would part with Van Praag and with the NIMH Research Domain Criteria (RDoC) approach in the assumption that we should focus on biological or psychopathological dimensions alone [13, 17]. Progress will be made in that approach, as Wernicke long ago argued [18]. But some progress still will require the categorical clinical nosology approach that Kraepelin promoted, and which has been proven valid in so many medical illnesses, dating from Morgagni to Alzheimer and the spirochete.

The key issue is not categories versus dimensions; it is science versus pragmatism [2]. Will we continue to deny the primacy of science in favor of the pragmatic utilities of the profession, as the leader of DSM-IV explicitly advocates [12]? Or will we return to the fold of scientific medicine, and base our diagnoses on our best science to date, even if it has limitations or errors?

Scientific truth, after all, is nothing but corrected error. One cannot reach the truth if one is afraid to err.


  1. Top of page
  2. References
  • 1
    Jouanna J. Hippocrates. Baltimore: Johns Hopkins University Press, 1999.
  • 2
    Ghaemi SN. Taking disease seriously: beyond “pragmatic” nosology. In: Kendler KS, Parnas J (eds). Philosophical issues in psychiatry II: Nosology. Oxford: Oxford University Press, 2012:42-53.
  • 3
    Goodwin F, Jamison K. Manic depressive illness, 2nd ed. New York: Oxford University Press, 2007.
  • 4
    Bienvenu OJ, Davydow DS, Kendler KS. Psychiatric ‘diseases’ versus behavioral disorders and degree of genetic influence. Psychol Med 2011;41:33-40.
  • 5
    Boestrom A. Zustandsbild und Krankheit in der Psychiatrie. Klinische Wochenschrift 1923;2:1728-31.
  • 6
    North C, Yutzy S. Goodwin and Guze's psychiatric diagnosis, 6th ed. New York: Oxford University Press, 2010.
  • 7
    Kendler KS, Prescott C. Genes, environment, and psychopathology. New York: Guilford, 2006.
  • 8
    Eaves L, Eysenck H, Martin N. Genes, culture, and personality. London: Academic Press, 1989.
  • 9
    Zanarini M. Role of sexual abuse in etiology of borderline personality disorder. Washington: American Psychiatric Press, 1997.
  • 10
    Decker H. The making of DSM-III. New York: Oxford University Press, 2013.
  • 11
    Ghaemi SN. The rise and fall of the biopsychosocial model: reconciling art and science in psychiatry. Baltimore: Johns Hopkins University Press, 2009.
  • 12
    Frances A. DSM in Philosophyland: curioser and curioser. AAPP Bulletin 2010;17:3-7.
  • 13
    Van Praag HM. Make-believes in psychiatry: or the perils of progress. New York: Brunner Mazel, 1992.
  • 14
    Frances A. The past, present and future of psychiatric diagnosis. World Psychiatry 2013;12:111-2.
  • 15
    Berrios GE. The history of mental symptoms. Cambridge: Cambridge University Press, 1996.
  • 16
    Ghaemi SN. On depression: diagnosis, drugs and despair in the modern world. Baltimore: Johns Hopkins Press, 2013.
  • 17
    Insel T, Cuthbert B, Garvey M et al. Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiatry 2010;167:748-51.
  • 18
    Wernicke C. Grundriss der Psychiatrie in klinischen Vorlesungen. Leipzig: Thieme, 1906.