Examining APA’s Proposed Redesign of the DSM
Can the future DSM overcome the epistemic arrogance of its predecessors?
My article ‘The Future DSM: Bold Redesign, Lingering Blind Spots’ appeared as the March 2026 cover story in Psychiatric Times. In it, I examine the reports from the Future DSM Strategic Committee and its subcommittees, recently published as a series of papers in the American Journal of Psychiatry. I’m republishing the piece here for readers of Psychiatry at the Margins.
The American Psychiatric Association (APA) has established a road map for updating its official diagnostic manual. In a series of articles published in the American Journal of Psychiatry, the Future DSM Strategic Committee has presented details of its progress and strategic priorities since beginning work in May 2024 [1-5]. Their publication represents the most comprehensive rethinking of psychiatric classification since DSM-III in 1980. In this article, I offer an overview of the proposed changes, and following that, I will offer some personal commentary and make a brief case for some suggestions of my own that I hope the readers will find illuminating and that the Future DSM Strategic Committee will take into consideration.
The overview paper by the committee, chaired by Maria A. Oquendo, MD, PhD, starts off by identifying multiple critiques of existing DSM editions that the future DSM seeks to address [1]. First is the manual’s atheoretical stance regarding causal factors. DSM-III emerged when “there were several conflicting hypotheses about causal mechanisms of mental illness, with limited empirical data and divergent views.” The atheoretical approach allowed psychiatry to sidestep these debates and focus on describing disorders with reliable criteria.
The new committee grapples with whether this agnosticism remains justified. They acknowledge that although there is broad agreement that disorders arise from the interplay of neurophysiological, developmental, experiential, sociocultural, and environmental factors, scientific understanding of how they interact is still premature. Nonetheless, they believe it is time to move from an atheoretical position toward one that explicitly acknowledges this multifactorial causal interplay. The DSM committee also recognizes that DSM constructs are not natural kinds, but the clinical and scientific work of classification can still proceed meaningfully by adopting a pragmatic stance. We cannot wait for perfect knowledge of valid boundaries before providing diagnostic tools for clinical practice.
Another fundamental issue concerns DSM’s categorical structure, where disorders are either present or absent. The committee notes that “many clinical presentations manifest symptoms that occur along a spectrum of severity and that some symptoms, such as anxiety and anhedonia, are transdiagnostic.” However, they point to a practical reality in support of categorical diagnoses: “Clinical decision-making involves categorical choices among a set of finite options.” DSM-5 attempted to address the presence of transdiagnostic symptoms through crosscutting dimensional symptom measures, yet these tools were relegated to section III, “often deemed by readers to be optional,” limiting their impact. The future DSM wants to address this issue in a better way.
The committee has organized its work through 4 subcommittees, as follows:
Structure and Dimensions
Functioning and Quality of Life
Biomarkers and Biological Factors
Socioeconomic, Cultural, and Environmental Determinants of Mental Health
The committee plans to integrate “people with lived experience as experts” alongside traditional clinical and research expertise. They are considering moving “away from theoretical agnosticism and embracing biology and environment and their interactions as key determinants of mental disorders.”
The committee envisions DSM evolving continuously rather than undergoing periodic major revisions. They propose regular cycles of solicited improvements with transparent review criteria, arguing for incorporating progress as it happens rather than waiting for comprehensive overhauls. The goal is to transform DSM into an online dynamic manual that keeps pace with advancing knowledge. Even the manual’s name may change from Diagnostic and Statistical Manual of Mental Disorders to Diagnostic and Scientific Manual of Mental Disorders, recognizing that the goal of the manual is “no longer simply to provide for the collection of psychiatric hospital and census statistics.”
The 4-Domain Model
The Structure and Dimensions Subcommittee, led by Dost Öngür, MD, PhD, proposes a major redesign, introducing a 4-domain model [2]. This is intended to address the problem that although DSM is near universally used for communication, its categorical structure poorly reflects clinical reality. “Only a minority of patients present with a classic form of one disorder as described in DSM; most instead present with a mixture of problems along dimensions such as mood, anxiety, psychosis, addiction, and so on.”
The model consists of 4 interconnected domains:
Domain I: Contextual factors. This includes socioeconomic, cultural, and environmental determinants; developmental factors; medical comorbidities; functioning levels; and patient-reported quality of life (QOL). These factors, currently relegated to background information, would become central to diagnosis.
Domain II: Biomarkers and biological factors. This represents the first systematic inclusion of biological measures in DSM classification, encompassing “all factors related to the biology of brain and body measured using any modality—including neuroimaging, genetics, metabolomics, cognition, digital phenotypes.” Although validated biomarkers remain rare, the structure would be ready “to accept new, rigorously studied biomarkers as they become available.”
Domain III: Diagnoses. This introduces variable specificity. Clinicians could diagnose at a broad “major category” level (such as psychosis or trauma-related disorders) when information is limited or provide a “specific diagnosis with specifier(s)” when sufficient information exists.
Domain IV: Transdiagnostic features. This allows clinicians to document common problems that transcend diagnostic boundaries, such as anxiety or cognitive difficulties, even when these do not warrant separate diagnoses.
Öngür et al. provide a concrete example: a patient with prolonged grief disorder, posttraumatic stress disorder with dissociative symptoms, and alcohol use disorder. The evaluation would document contextual factors (low income, threat of job loss, history of childhood physical abuse), medical comorbidity (irritable bowel syndrome), biological factors (amygdala hyperreactivity on brain imaging), all 3 diagnoses with severity ratings, and transdiagnostic anxiety.
Integrating Biological Measures
The Biomarkers and Biological Factors Subcommittee, led by Bruce Cuthbert, PhD, and Anissa Abi-Dargham, MD, faces a daunting challenge: how to incorporate biological measures into a manual that has remained largely based on symptoms since 1980. With the recent exception of Alzheimer disease, “no biomarkers have had the needed specificity and sensitivity for use in routine psychiatric diagnosis.”
The report highlights 4 areas showing potential:
Psychosis biotypes: The Bipolar-Schizophrenia Network for Intermediate Phenotypes consortium identified 3 distinct biotypes among individuals with psychosis using a battery of tests including electroencephalogram (EEG), brain imaging, eye tracking, and cognitive testing. These biotypes cut across traditional diagnoses. One finding suggests patients with low “intrinsic EEG activity” may respond better to clozapine, pending validation.
Genetic risk scores: Polygenic risk scores aggregate thousands of genetic variants to calculate disease risk. For schizophrenia, individuals in the top 10% have 2.3 times higher odds of diagnosis. However, these scores “still have limited predictive power, may suffer from population bias, and may not add much more than already established risk factors, such as family history.”
Inflammatory markers: C-reactive protein (CRP), a readily available blood test, may identify an “inflammatory subtype” of depression. Approximately 27% of patients with depression show elevated CRP and respond better to certain antidepressants than selective serotonin reuptake inhibitors (SSRIs).
Brain circuitry: Findings from functional MRI studies identified a “cognitive biotype” representing 25% of patients with depression who show reduced activation in cognitive control regions. These patients “do not respond well to standard SSRI treatment” but benefit from cognitive behavior therapy or transcranial magnetic stimulation.
The next phase for the subcommittee involves defining criteria for listing candidate biomarkers “so that the process is structured, rigorous, and guided by robust science,” with attention to cultural and ethnic considerations.
Functioning and Quality of Life
The Functioning and Quality of Life Subcommittee argues that psychiatric diagnosis remains incomplete without systematic assessment of how patients function in daily life and perceive their well-being. Mental illnesses represent “8 of the top 25 causes of years lived with disability worldwide,” yet DSM has treated functional assessment inconsistently.
The Global Assessment of Functioning scale conflated symptoms with functioning: A patient could score in the same range due to either “serious symptoms” or “serious impairment,” creating ambiguity. DSM-5 recommended the World Health Organization Disability Assessment Schedule (WHODAS 2.0), but placement in section III “sent an unintended message to the field that the WHODAS 2.0 was not ready for routine use.”
The subcommittee concludes that functioning and QOL are related yet distinct domains requiring separate assessment. Functioning refers to objective capacity to perform activities and fulfill roles, such as going to work, maintaining relationships, managing daily tasks. QOL adds the patient’s subjective perspective on their well-being and life satisfaction, defined by WHO as “one’s perceptions of their position in life, contextualized by the culture and value systems in which they live.” Two patients might have similar functional impairments but dramatically different QOL ratings, depending on their values, expectations, and life circumstances.
The subcommittee reviewed numerous assessment tools but found that “no single instrument met all desirable characteristics.” The challenge is balancing thoroughness with feasibility. Even brief versions of recommended scales, the 12-item WHODAS-2.0 for functioning and the 16-item Quality of Life Enjoyment and Satisfaction Questionnaire for QOL, “may still prove infeasible in many clinical and research settings.”
The committee’s central recommendation is unambiguous: Functioning and QOL “must be included in the essential elements of DSM and not be relegated to the ‘Emerging Measures and Models’ section.” This marks a shift from treating these domains as optional supplements to recognizing them as core components of psychiatric diagnosis.
Integrating Social Context
The Socioeconomic, Cultural, and Environmental Determinants of Mental Health Subcommittee, led by Milton Wainberg, MD, proposes systematic integration of contextual factors historically treated as supplementary. These conditions span 5 domains: demographic factors, economic stability, neighborhood and built environment, environmental events, and social/cultural context.
The subcommittee outlines 3 potential approaches: incorporating screening into routine intake, embedding factors into clinical decision trees where “high burden triggers enhanced case management,” and developing “risk-adjusted diagnostic models” where patients meeting the same criteria but facing intersecting stressors warrant more intensive intervention.
The authors acknowledge barriers: Many systems “do not routinely collect” such data, resource shortages are common, and there is “risk of tokenism, where simple checklists without further evaluation make diversity efforts symbolic rather than structural.”
Summary of Proposed Changes to the Future DSM
I. Structural and conceptual framework changes
Transition to a living document with updates at briefer intervals
Possibly rename to Diagnostic and Scientific Manual to emphasize evolution away from statistical recordkeeping origins
Soften theoretical agnosticism to explicitly embrace biology, environment, and their interactions
With the new diagnostic construction model, organize assessments and diagnostic formulations into 4 domains: contextual factors, biomarkers, diagnoses, and transdiagnostic features
II. Integration of biomarkers and biological factors
Include candidate biomarkers derived from a wide range of methods, including neuroimaging, genetics, inflammatory markers, electrophysiology, and digital phenotypes
Use biomarkers to identify biotypes/subtypes for treatment selection (eg, inflammatory subtype of depression)
III. Enhancement of functioning and QOL
Make functioning and QOL essential elements of psychiatric diagnosis
Use brief, feasible instruments for clinical use
IV. Socioeconomic, cultural, and environmental determinants (SCE-DOH)
Move beyond V/Z codes to systematically integrate SCE-DOH into psychiatric formulations
Implement risk-adjusted diagnostic models that account for social stressors when estimating severity and treatment needs
V. Diagnostic specification and dimensionality
Allow variable specificity levels from major category (eg, psychosis) to specific diagnosis (eg, schizophrenia)
Integrate transdiagnostic dimensions to better reflect comorbidity and symptom heterogeneity
VI. Procedural and stakeholder changes
Include people with lived experience, global perspectives, and Indigenous epistemologies as experts in the revision process
Personal Reflections and Suggestions
Diagnostic Formulation Needs to Be Linked to Treatment
The future DSM is moving toward diagnostic formulation. This is a welcome change and a nod back to DSM-III and DSM-IV’s multiaxial diagnosis. In my opinion, for this to be meaningful, along with the atheoretical stance, DSM also needs to break its silence on treatment implications. Historically, DSM has restricted itself to diagnosis, but for diagnostic formulations to be useful and implemented by clinicians, DSM must explain how different aspects of the model will guide specific treatment. The manual does not need to become a collection of treatment guidelines, but where diagnostic distinctions have treatment implications, this needs acknowledgment and elaboration.
The Missing Domain: Psychological Factors and Personality Traits
It is notable that the proposed structure lacks a dedicated domain for psychological functioning. This seems like an inexcusable oversight. Classification systems serve to inform probabilistic reasoning about presentation, trajectory, and therapeutic response. Without designated space to document psychological capacities and patterns, clinical formulations will inadequately capture person-level psychological characteristics that influence prognosis and treatment planning.
By psychological factors here I am referring to measurable, reliable, relatively enduring dispositions and clinically meaningful psychological capacities that provide information beyond what diagnosis and symptom dimensions convey. These include personality trait profiles (Big 5 personality traits such as neuroticism, agreeableness), patterns of attachment, internalized relational models, reflective functioning abilities, metacognitive beliefs, recurrent cognitive distortions, defense mechanisms, and psychological resources supporting resilience.
Such constructs routinely inform practicing clinicians’ decisions about psychotherapy customization, patient-modality matching, treatment pacing and emphasis, potential for therapeutic relationship difficulties, interpretation of treatment nonresponse, and selection among skills-focused, insight-oriented, family-involved, or integrated approaches. These aspects are particularly emphasized in the Psychodynamic Diagnostic Manual (PDM). If DSM neglects them, it will lose ground among psychotherapy clinicians.
The solution is fortunately straightforward: Incorporate “Psychological Factors” as an explicitly named domain, supported by concise implementation guidance and validated brief assessment instruments.
DSM’s Notion of Dimensionality Is Underdeveloped
Hopwood et al. (2023) have described 3 different meanings of dimension in psychopathology literature [6]. The first simply refers to continuous variables where indicators can be summed or averaged rather than used categorically, such as counting depression symptoms on a continuous scale (eg, Patient Health Questionnaire-9) rather than making a binary diagnosis. The second, more restrictive meaning requires that these indicators cohere statistically, fitting a unidimensional factor model where a single superordinate construct explains their covariance. The third and most restrictive meaning adds the requirement that the dimension be empirically distinguishable from related constructs within a multidimensional framework. Hierarchical Taxonomy of Psychopathology (HiTOP) exemplifies such a statistically coherent, multidimensional framework. The Distress subfactor in the Internalizing spectrum, for instance, is not only continuous and homogeneous but also demonstrably distinct from the Fear subfactor.
These conceptual distinctions regarding dimensionality matter because debates about dimensional vs categorical diagnosis become muddled when authors use the same term to mean different things. Some may consider simple continuous measures to be sufficient for dimensionality (what DSM is currently doing), but psychiatrists and psychologists interested in grounding classification in structural evidence envision psychometrically validated constructs embedded in comprehensive structural models as appropriately dimensional (the HiTOP approach).
Major Categories Should Not Become the New “Unspecified”
Öngür et al. note that for the major category level, such as depression or psychosis, they “anticipate that currently available diagnostic codes, such as unspecified depressive disorder, unspecified schizophrenia spectrum, and other psychotic disorder, will be used in combination with the severity measure.” [2]
Prior DSM editions have delegated unspecified categories to almost second-class status. Large swaths of psychopathology simply are not named in the DSM, either because presentations are subthreshold or because no corresponding category exists. This is why clinicians use unspecified categories so frequently. The more I practice, the more I believe specified DSM criteria are rather conservative. The diagnostic manual struggles to keep up with the magnitude of clinically significant distress that exists.
The proposal to use major categories is a step forward in officially recognizing the large domain of clinically relevant presentations. Although reliance on unspecified International Classification of Diseases (ICD) codes makes practical sense, I strongly recommend that the future DSM not use the word unspecified in the names of major categories to ensure that these major categories are not deprioritized in the same way as unspecified categories have been in DSM-5 (and the way “not otherwise specified” categories were in DSM-III and DSM-IV before that).
The Definition of Disorder (and the Proposed Name Change)
DSM’s reliance on the term “disorder” and its sloppy formal definition in the manual have led to a lot of unnecessary conceptual confusion. What is the difference between “mental health problems” that merit clinical treatment due to distress and impairment and “disorders” as classified in the DSM? By the official DSM definition, the distinction involves psychological or biological “dysfunction.” Dysfunction is not defined further, but diagnostic criteria make clear the notion is commonsensical and folk-psychological: Something is “not doing what it is supposed to do” based on everyday norms of what is expected and typical. [7]
Two strategies are possible here:
1. Make the DSM meaning of dysfunction explicit and clear, differentiating it from other notions of dysfunction (such as failure of evolved mechanisms or biostatistical deviation from species-typical norms). This will help address the popular misconception that the DSM assumes the existence of pathological processes inside the individual. But it raises the additional question: Why should a scientific manual design its scope around such a commonsensical, folk-psychological notion?
2. Abandon the commitment to a folk-psychological notion of dysfunction, avoid references to dysfunction in the formal definition, and recognize the manual’s interest in a wide range of experiential and behavioral states of distress, impairment, and harm to others which come to clinical attention, warrant management, and have been characterized with some degree of reliability and rigor (meeting DSM’s evidential standards) by the scientific community.
If we go the second route, a more accurate title for the book would be Diagnostic and Statistical/Scientific Manual of Mental Disorders and Related Mental Health Problems, akin to ICD’s full name: International Statistical Classification of Diseases and Related Health Problems.
Indicate the Degree of Empirical Validation for Specific Diagnoses
Currently, browsing through DSM gives the false impression that all conditions have equal clinical and scientific legitimacy. Disinhibited social engagement disorder appears to have the same status as posttraumatic stress disorder. It is like a house of mirrors that flattens everything to the same size [8]. Validity and utility are not equally distributed among DSM diagnoses. Schizophrenia and disruptive mood dysregulation disorder are continents apart in terms of validation. The future DSM needs to communicate this meaningfully in some manner. One way to approach this could be to synthesize the evidence of interrater reliability, predictive validity, and diagnostic stability for each specified diagnosis.
Explain Why Diagnostic Thresholds Are What They Are
Unlike dimensional diagnoses in general medicine (such as hypertension thresholds that optimize cardiovascular risk management), DSM thresholds do not seem to clearly optimize anything. Some thresholds seem semi-arbitrary by design. Spitzer famously said 5 criteria for depression were chosen as a threshold because “4 seemed like not enough and 6 seemed like too much.” Others are based on nonempirical considerations: Prolonged grief’s 12-month threshold was reportedly set more conservatively than research supported to avoid public backlash.
There is nothing wrong with using best guesses or expert opinion as preliminary thresholds, but we cannot treat these thresholds as sacred or conduct research programs assuming they capture meaningful etiological differences. The future DSM needs transparency about what evidence, if any, supports particular thresholds, and if the threshold can be set differently to optimize different clinical goals, that information should be disclosed.
Include HiTOP in the DSM Appendix
DSM categories have advantages for clinical communication, but they rest on semi-arbitrary thresholds, produce heavy comorbidity, and lump heterogeneous presentations together. HiTOP offers a dimensional alternative that starts from psychometric data rather than historical categories, with research showing advantages at the spectrum level, with better stability, cleaner psychometric structure, and stronger validator links.
The DSM appendix, where the Alternative Model for Personality Disorders currently lives, would be a reasonable place for inclusion of HiTOP. This would legitimize psychometric approaches to classification and provide a bridge between categorical and dimensional models.
DSM, including the future DSM, should not be seen as the one true classification but as one historically contingent, fallible effort. The clinical and scientific reality is that we now exist in a landscape of nosological pluralism, where traditional DSM serves as a “good enough” shared language that coexists with alternatives such as HiTOP and PDM. A plurality of legitimate and useful classifications is already here. It is heartening to see Oquendo et al. write, “Applied with all due epistemic humility, DSM can continue to play an important role in clinical care and research.” I hope that the future DSM can overcome the epistemic arrogance of its predecessors and show us that the manual has philosophically matured.
Dr. Aftab is a psychiatrist in Cleveland, Ohio, and clinical associate professor of psychiatry at Case Western Reserve University School of Medicine. He is the editor of “Conversations in Critical Psychiatry” (Oxford University Press, 2024) and writes online at “Psychiatry at the Margins.”
See also:
References
1. Oquendo MA, Abi-Dargham A, Alpert JE, et al. Initial strategy for the future of DSM. Am J Psychiatry. 2026;appiajp20250878. Online ahead of print.
2. Öngür D, Abi-Dargham A, Clarke DE, et al. The future of DSM: a report from the Structure and Dimensions Subcommittee. Am J Psychiatry. 2026;appiajp20250876. Online ahead of print.
3. Cuthbert B, Ajilore O, Alpert JE, et al. The future of DSM: role of candidate biomarkers and biological factors. Am J Psychiatry. 2026;appiajp20250877. Online ahead of print.
4. Drexler K, Alpert JE, Benton TD, et al. The future of DSM: are functioning and quality of life essential elements of a complete psychiatric diagnosis? Am J Psychiatry. 2026;appiajp20250874. Online ahead of print.
5. Wainberg ML, Alpert JE, Benton TD, et al. The future of DSM: a strategic vision for incorporating socioeconomic, cultural, and environmental determinants and intersectionality. Am J Psychiatry. 2026:appiajp20250875. Online ahead of print.
6. Hopwood CJ, Morey LC, Markon KE. What is a psychopathology dimension? Clin Psychol Rev. 2023;106:102356.
7. Aftab A. Weaving conceptual and empirical work in psychiatry: Kenneth S. Kendler, MD. Psychiatric Times. May 26, 2020. https://www.psychiatrictimes.com/view/weaving-conceptual-and-empirical-work-psychiatry-kenneth-s-kendler-md
8. Aftab A. 6 suggestions for DSM-6. Psychiatry at the Margins. November 20, 2025. Accessed February 10, 2026. https://www.psychiatrymargins.com/p/6-suggestions-for-dsm-6
I read your summary of the working groups' proposals thinking "the next one is going to be about psychological factors... right?" Absolutely floored and somewhat aghast that you even need to make the suggestion for it to be its own section. It's a fundamental part of psychiatry!
I send my best wishes and condolences to the DSM redesigns, not necessarily in that order.
It may be an oversimplification to say that calling something a psychiatric disorder/illness vs NOT a disorder/illness has more social and political ramifications than is true for non-psychiatric medical conditions. It underlies the challenge of incorporating non-"experts" in the process. There are a good many people who consider themselves "survivors" of psychiatric care, and many of those flatly reject the existence of psychiatric disorders, much less the value of psychiatric care. If the APA intends to incorporate diverse views, it'll be tough - or impossible - to thread the needle of "all views are welcome, but our position is that there are conditions that we consider pathological and worthy of our treatment." Whether such conditions exist at all seems like a major fork in the road. I can imagine some of those ( eg madinamerica.com) yet again complaining that they've been disenfranchised once the new DSM says "yes, there are psychiatric illnesses." I think the APA needs to do more than offer lip service to engaging with that kind of thinking; it's not as if our treatment has been so spot-on that we can ignore this.
I'm intrigued by the concept of epistemic arrogance. I came across a commentary about a subtle but useful alternative category of 'epistemic confidence." https://blog.apaonline.org/2019/06/03/against-epistemic-arrogance/
Psychiatrists are experts, if anyone is, in mental illness diagnosis and treatment, so society gives us some rights to have our views have more sway. How we do that with humility and not convey a sense that " we really don't know all that much" is another tough needle to thread.
Parenthetically, it has long griped me that the press likes to call the DSM 'the psychiatric bible," and if the DSM revisers get out in front of that from the outset, it'll help with balancing our expertise with our humility. I know it's a cute catchphrase, but we let it go too often.
This is a real opportunity to make the DSM more useful and respectable. There is a real risk of groupthink limiting the redesigners' effectiveness in doing so. I'm sure the work groups will have their own internal politics, and I hope that it'll be safe for the group members to say so when appropriate.
If I never see "NOS" again, they can declare victory!