As readers of Psychiatry at the Margins are well aware, I have been interested in understanding the phenomenon of overdiagnosis in psychiatry, and I have been frustrated by the tendency of critics to assert or imply that vast proportions of human suffering and impairment that come to clinical attention are simply undeserving of clinical recognition. It is common for critics to invoke “normality,” protecting “normal,” searching for “normal,” with “normal” as being under threat from out-of-control, ever-expanding DSMs. I was once sympathetic to this line of thinking, but now I am skeptical of invocations of “normality,” and I am of the view that the DSM thresholds are rather conservative. The DSM falls short of a comprehensive description of psychopathological presentations, and the categories are not broad enough to capture all clinically relevant forms of psychological suffering and disabilities for which diagnosis is needed. The criteria for specified diagnoses exclude a significant number of patients who experience considerable distress and impairment in their daily lives.
People fundamentally misunderstand the dynamic behind increasing rates of psychiatric diagnosis when they focus on the DSM as being the driver of this expansion. By and large, clinicians are not imposing labels or disorder judgments on reluctant patients who would rather not have a name for what they are experiencing (although it does happen). There is a tremendous demand and appetite for diagnostic labels. Psychiatry legitimized psychological distress and disability as healthcare problems deserving of recognition and treatment, but the profession was unprepared for the magnitude of clinically significant distress out there, and the diagnostic manuals have struggled to keep up. Much clinical practice involves clinicians responding to real suffering inadequately captured by official criteria. This leads to stretching existing diagnoses, not from diagnostic carelessness but due to genuine gaps in the diagnostic framework.
I have organized these ideas into a hypothesis of psychiatric overdiagnosis as diagnostic stretching in response to clinical need. The article describing this hypothesis is currently available as a preprint on PsyArXiv.
I encourage you to read the preprint for details.
Below is some additional overview of my argument.
Psychiatric overdiagnosis, for the purposes of my hypothesis, refers to the observation that there has been a rapid rise in the rate of mental health diagnoses in recent decades (including internalizing disorders such as depression, anxiety, and trauma-related diagnoses, and neurodevelopmental disorders such as autism and ADHD), combined with the suspicion or skepticism that a substantial proportion of these diagnoses are illegitimate, unwarranted, or unnecessary.
Note that this definition differs from how other authors have defined overdiagnosis. For example, Suzanne O’Sullivan (2025) defines overdiagnosis as a diagnosis that, while technically correct, does not benefit the patient and may do harm.
Since the 1990s, there has been a stretching of specified diagnostic categories (e.g. major depressive disorder, ADHD, bipolar disorder) in clinical practice to cover patients who have the relevant symptoms and have clinically significant distress and/or impairment but do not otherwise meet the official diagnostic thresholds for the specified categories. The presentations are either subthreshold (the fall below the required number of symptoms or temporal thresholds) or they are phenomenologically adjacent to the diagnosis (e.g. mood dysregulation being adjacent to bipolar disorder; focus problems being adjacent to ADHD).
Mental health awareness campaigns legitimized mood difficulties, anxiety, focus problems, and social difficulties as legitimate reasons for consultation, on par with bodily complaints such as dyspepsia, back pain, headache, and dizziness. Increased awareness of mental health problems has lowered the help-seeking threshold such that rates of diagnosis are now converging on the population burden of distress/impairment. Specified diagnoses are being stretched to allow that convergence.
This stretching is a clinical response to legitimate patient needs interacting with official classifications that are too narrow (not too broad, contrary to popular perception) and whose thresholds are semi-arbitrary and do not correspond to real-world distress or impairment. The suggested solution is that our clinical classifications should more accurately characterize clinically significant forms of distress and impairment instead of entrenching current gaps in the diagnostic schema.
The hypothesis can be empirically evaluated by measuring the population burden of psychiatric distress and/or impairment directly, benchmarking against common conditions from general medicine, and quantifying the risk of serious adverse outcomes. Because diagnostic stretching is hypothesized to be a response to institutional dependence on specified diagnoses, it also makes predictions that differ from competing hypotheses. My hypothesis doesn’t take a position, one way or the other, on whether diagnosis and treatments do more good than harm.
The individual ingredients of my hypothesis are not new; they have all been discussed in some form or other over the years (e.g. Haslam’s work on concept creep), but I have not come across an arrangement of these ingredients in the manner and with the valence that I see as most accurately characterizing the state of affairs.
The intention of my hypothesis is to offer an explanation for both why diagnoses have increased (this is in response to genuine clinical need and the population burden of distress/impairment) and why this invites skepticism (stretching beyond official thresholds is problematic for various reasons and is met with understandable suspicion).
Comments are open.
I haven't read the article yet. But I think you are on target. I was amused thinking about a lecture Nassir Ghaemi gave at the LAST APA convention in NYC, in which he "showed" statistically that we UNDERdiagnose bipolar disorder, not OVERdiagnose it. There is a similar argument we UNDERdiagnose ADHD not OVERdiagnose it, too. I realized both these arguments are no more than sophistry.
The fundemental problem with these discussions relates to the variable definition of "diagnosis" and even "normal", and unless these are defined at the start, assessing the overdiagnosis "problem" is hopeless. That is the heuristic problem. The clinical problem more relates to motivations of patients. When patients seek enlightenment and recommendations or treatment, "diagnosis" means one thing. When they actually come to the doctor with an agenda explicit or unspoken (like "stimulants"), "diagnosis" means something else.
In my clinical experience (and I am open-minded to other experiences), patients adopt or repeat or encourage certain diagnoses because it labels an anxiety they have and this makes them feel better (ie "autism"). In contrast, the patients who come to me seeking stimulants are less intent on being diagnosed with "ADHD" than obtaining their desired drug. I think that we document "ADHD" in various locations to address OUR anxiety about prescribing the stimulant. I personally don't care, I don't have this anxiety. But there are numerous places "ADHD" is populated or has to be written to finish the prescription process, and I write "ADHD" as a telegraphic way of describing why the person gets stimulants, to differentiate it from "narcolepsy" (for example) or "OSAHS", maybe my compulsion to write that is subconciously addressing my anxiety so that I don't appear to be giving uppers "because I feel like it".
Remember, precise diagnoses are rarely needed to provide psychiatric services or get paid. I might have to document a relatively precise diagnosis like "obstructive sleep apnea", but this is rare. I might have to endorse specific psychiatric diagnoses, say, twice per year in my practice, it is that rare, most commonly in pre-authorizations. And I don't take it seriously or concern myself with truthfulness in those instances either, I am just trying to finish the damn pre-auth and get the drug approved. But this is truly rare.
I personally don't have an "overdiagnosis" problem. Now patients bandy various labels sure, but that is just one more aspect of getting a history. Truthfully the problem isn't their throwing around jargon, it is their refusal to provide records, which has been a gradual trend in psychiatry such that most patients are now shocked when I request their records.
If we psychiatrists are "overdiagnosing" something, that comes from disagreement over what the diagnosis means, which I consider a problem of operationalization (a serious problem in psychiatric nomenclature). Examples are biopolar disorder or ADHD. I don't think there is a solid consensus in the profession or in society about what these mean. Compare the diagnosis of "Lewy Body Dementia". It is in the DSM, we all know what it means. No one would say it is over-used. I PERSONALLY think that "major depressive disorder" is very easy to operationalize, others might disagree, to give an example of something without a clear physiologic basis or pathology.
So I think that "overdiagnosis" is a problem only if you say it is.
With the breakdown of the nuclear family I have no doubt the average person is quite a bit more disturbed today than in the 1960s.
But in my opinion as any diagnosis can introduce iatrogenic harm, diagnostic threshold should not be objective across all generations… but should be in direct proportion to our ability to fix deep-seated problems. To diagnose is to involve yourself in treatment, otherwise it may cause harm, and although for example borderline traits are more common today, and psychiatrists as public health practitioners should comment on it, nevertheless, as people have gotten sicker a case can be made to pull back from diagnosis.
A diagnosis is like a Freudian interpretation: it is a stressor and may destabilize someone.
I spend a lot of time undiagnosing.