Future DSM Repeats Past Mistakes: Validity Is Ignored

CONCEPTS IN PSYCHIATRY

Recent articles in the American Journal of Psychiatry¹ highlighted the activities of the Future DSM Strategic Committee and outlined planned changes to the diagnostic manual. They described 4 organizational adjustments—revisions to architecture, attention to biomarkers, contextual factors, and measurement of functioning and quality of life. Yet despite these promising areas of focus, the Committee, like the articles, largely overlooks DSM’s core scientific problem: its lack of validity.

Before reliability, before the debate over categories vs dimensions, and before we add biomarkers or social determinants, the central question remains: Are the diagnoses real? DSM‑III acknowledged that most diagnoses lacked validation; DSM‑IV repeated the structure; DSM‑5 avoided the question; and the future DSM appears poised to do the same.

If the diagnostic system is valid, psychiatry can advance scientifically. If it is invalid, everything that depends on it—biomarkers, epidemiology, therapeutics, reimbursement, and research—rests on an uncertain foundation. One of us (S.N.G.) raised these concerns directly with leaders of the Future DSM Committee; unfortunately, they received little weight. This article explains why DSM remains scientifically weak, why the new proposals do not address the underlying issue, and why psychiatry must adopt a scientific model based on empirical data rather than consensus.

Past and Present

We have been repeating essentially the same DSM system since 1980. We now live in an era in which, for the first time, the leaders of the DSM Task Force are no longer members of the Baby Boomer generation, but of the next cohort, Generation X. DSM-III and its successors were created by a generation that came of age in the 1960s and 1970s, and that group retained substantial influence over each revision through DSM-5 in 2013. Many of those figures have since retired or passed away. Leadership of the next DSM Task Force has now shifted to a new generation, those who came of age in the 1980s and 1990s. This generational transition might have been expected to bring meaningful change, consistent with Max Planck’s² observation that scientific progress often occurs not by persuading individuals but through generational replacement, with younger cohorts more open to new ideas.

One of us (S.N.G.) has met with members of the Future DSM leadership, whom he personally knows as peers through his professional career, unlike prior DSM-5 leaders, with whom his acquaintance was more remote. Yet despite this shift, the central problem that has limited progress in psychiatric diagnosis remains unchanged: the diagnostic system is not grounded primarily in validity, that is, in the reality or truthfulness of its definitions. In this article, we review this problem, explain why it matters, and argue that the only viable solution is to adopt the standard approach used in other areas of medicine: diagnostic definitions should be determined by scientific experts on the basis of empirical research, rather than by consensus or preference.

Without Validity, Any Future DSM Is an Obstacle to Progress

Back in the 1970s, the original precursor to DSM-III was the Research Diagnostic Criteria (RDC), which argued that psychiatric diagnosis should be based on validity and reliability (in that order); 14 valid diagnoses were identified.³ Diagnostic validators were proposed, based on the fact that an external validator is always needed for symptoms. Symptoms are nonspecific: different symptoms can happen in the same disease, and the same symptom can happen in different diseases. General medicine solved this problem in the 19th century with the advent of pathology and rise of autopsy. Psychiatrists like Theodor Meynert and Emil Kraepelin and others tried to apply to pathology to psychiatric diagnoses, mostly unsuccessfully. They then turned to other independent lines of evidence to be able to correctly evaluate symptoms and identify valid diagnoses: they emphasized the long-term course of illness and heredity (genetics). The RDC group formalized this approach, which in later years has evolved to 4 basic diagnostic validators: symptoms, genetics, course of illness, and biological markers.⁴

Diagnostic Validators

• Symptoms
• Genetics
• Course of Illness
• Biological Markers

No single validator is sufficient to make a diagnosis valid. The more the validators point in the same direction, identifying a clinical condition that is distinguishable from other clinical conditions and from normality, the more valid a diagnosis.

The RDC group was subsumed into the DSM-III Task Force, and in 2 years, 14 valid diagnoses were expanded to 292 diagnoses, which means that 278 (95.2% of the total) were invalid. They were invented, created out of thin air, based on the wishes and beliefs of clinical leaders in psychiatry. The DSM-III leadership admitted this fact,⁵ and argued that if the 95% invalid diagnoses could be defined well (reliability), then future studies could change them to make them more valid. That did not happen of course, because the leaders of DSM-IV in 1994 refused to make major changes to DSM-III, and the leaders of DSM-5 in 2013 refused to make major changes to DSM-IV. Reliability has not led to validity, and yet the future DSM leadership continues to claim that reliability matters more than validity. A dictionary is useful if the reliable definitions in them are true (valid).⁶ A false dictionary is worse than useless, it is misleading.

One possible path forward, advanced by one of us (M.R.), is for the future DSM to establish a validity-based hierarchy in which diagnoses are listed hierarchically according to their empirical support.⁷ Diagnoses with robust validity would be listed at or near the top of the hierarchy, while diagnoses with limited or poor validity would be listed at the bottom of the hierarchy, encouraging clinicians to utilize more valid diagnoses. Another path, advocated by one of us (S.N.G.) is to replace the DSM system with a new clinical research diagnostic criteria (CRDC),⁸ using the 4 diagnostic validators to identify and define diagnoses, as with the RDC of the 1970s. A third alternative, perhaps the best, is to simply eliminate DSM altogether and for clinicians to rely on textbooks and the scientific research to guide diagnosis, as is done in every other branch of medicine. Insurance reimbursement, as with the rest of medicine, would depend on ICD-10 or ICD-11 codes; there is no need for a DSM coding at all for anything.

All 3 approaches differ in practical details, but they share a common premise: psychiatric diagnosis should be governed by validity rather than by consensus, preference, or institutional convenience. As one of the RDC leaders, Samuel B. Guze, MD, observed, “Validity is the gold standard for judging any diagnostic category and comparing different diagnostic approaches. Esthetic and political consideration and a priori assumptions must give way to it.”⁹

DSM-III was a social construction; it was not a scientific document.¹⁰ It was created for the social and political purposes of the American psychiatric profession. That remains the case. DSM revisions happen when the American Psychiatric Association needs to renew its main source of income—sales of DSM. The economic motivation for continuing the DSM enterprise is hard to ignore.

Instead, psychiatry should reconsider the continued centrality of DSM to clinical practice and research. The current system remains heavily shaped by administrative, economic, and legal considerations, particularly insurance reimbursement, rather than by the scientific question of diagnostic validity. Yet diagnostic concepts should be determined primarily by empirical evidence concerning the nature of psychiatric illness, not by professional consensus or institutional convenience. If psychiatric diagnoses are to advance scientifically, they must be grounded in reality rather than maintained for pragmatic or bureaucratic reasons. Biology and nature does not care what we want; it won’t adjust to our DSM wishes. We need to adjust to reality, not the other way around, and find out which psychiatric illnesses really exist. The DSM approach is abetted by a widespread antiscientific postmodernist attitude that we never will know how to reach validity, and thus we can make things up as we like. Our patients deserve better.

Dr Ghaemi is a lecturer on psychiatry at Harvard Medical School, Cambridge Health Alliance, and is employed by Bristol Myers Squibb. The views expressed in this article are solely those of the authors and do not necessarily reflect the official policy or position of their employers.

Dr Ruffalo is an assistant professor of psychiatry at the University of Central Florida College of Medicine in Orlando, where he is director of psychotherapy training in the adult psychiatry residency program. He is also an adjunct assistant professor of psychiatry at Tufts University School of Medicine in Boston, Massachusetts, and on the faculty of The New Jersey Institute for Training in Psychoanalysis in Teaneck, New Jersey.

References

1. Oquendo MA, Abi-Dargham A, Alpert JE, et al. Initial strategy for the future of DSM. Am J Psychiatry. 2026;183(5):292-300.

2. Planck MK. Scientific Autobiography and Other Papers. Philosophical Library; 1950.

3. Spitzer RL, Endicott J, Robins E. Research diagnostic criteria: rationale and reliability. Arch Gen Psychiatry. 1978;35(6):773-782.

4. Robins E, Guze SB. Establishment of diagnostic validity in psychiatric illness: its application to schizophrenia. Am J Psychiatry. 1970;126(7):983-987.

5. Klerman GL, Vaillant GE, Spitzer RL, Michels R. A debate on DSM-III. Am J Psychiatry. 1984;141(4):539-553.

6. Ghaemi SN. Utility without validity is useless. World Psychiatry. 2016;15(1):35-37.

7. Ruffalo ML, Pies RW. Why We Need a Diagnostic Hierarchy in DSM-6. Manuscript submitted for publication.

8. Ghaemi SN, Angst J, Vohringer PA, et al. Clinical research diagnostic criteria for bipolar illness (CRDC-BP): rationale and validity. Int J Bipolar Disord. 2022;10(1):23.

9. Guze S. Why Psychiatry Is a Branch of Medicine. Oxford University Press; 1992.

10. Decker HS. The Making of DSM-III: A Diagnostic Manual’s Conquest of American Psychiatry. Oxford University Press; 2013.