Autism Diagnosis Today (2): Dediagnosing

In the past two decades, our societies have become more fluent in the language of mental health. Public campaigns encourage people to “speak out,” schools teach children about anxiety and depression, and workplaces stress the importance of wellbeing.  Yet a paradox has emerged: despite this rising awareness, mental health level around the world continues to decline (Schrijvers et al. 2024). What was intended to alleviate suffering may instead make it worse.

This paradox is not confined to mental health alone. Medicine in a broad sense has entered what neurologist Suzanne O’Sullivan (2025) has called an “age of diagnosis.” Conditions are defined ever more broadly, and everyday struggles are reclassified as disorders. The result is a double bind: while diagnoses can explain and legitimise distress, they also risk medicalising ordinary human variation and loss.

The expanding reach of diagnosis

A diagnosis does important work. It can explain troubling experiences, legitimise suffering, provide access to care, and guide treatment (Page & Etherton-Beer 2019). But diagnosis also has less favourable consequences: it can stigmatise, narrow identity, generate unnecessary treatments, and burden health systems. Lea & Hofmann (2022) have proposed “dediagnosing” as “a framework for removing diagnoses that do not contribute to reducing the persons’ suffering and should be introduced to make people less ill.”

The problem is not merely the number of diagnoses, but their elasticity. Disease definitions expand over time, lowering thresholds for what counts as pathological (Moynihan et al. 2019). ADHD is a telling example. As Thomas, Mitchell, and Batstra (2013) observed, successive editions of the DSM have broadened the criteria, doubling prevalence rates in some countries and sharply increasing demand in medication. While diagnosis can be validating, overdiagnosis risks harm: drugs with side effects, reduced expectations from teachers and parents, and the internalisation of a label that can shape identity in limiting ways. The story of ADHD, they argued, raises a critical question: should some psychiatric diagnoses need to become less fixed, more temporary, and open to dediagnosis when circumstances change?

Autism and the question of “clinically significant impairment”

These issues resonate strongly in the field of autism as well. Since DSM-5 (2013), a diagnosis of ASD requires evidence of “clinically significant impairment.” The criterion was meant to separate clinically relevant cases from broader expressions of neurodiversity – subclinical autistic traits. Yet, as Jellett & Muggleton (2022) argue, this interpretation clashes with the neurodiversity paradigm, which maintains that autistic traits are not inherently impairing but may become disabling when social environments are unaccommodating.

This produces a conceptual tension. On one hand, clinicians must diagnose autism only when impairment is present. On the other hand, many autistic people resist the idea that their condition is defined by deficit. The diagnosis itself becomes unstable: is autism a disorder, a disability, a difference, or an identity?

The problem of autistic identity

For many people, an autism diagnosis is not a medical label but an identity marker. It can bring clarity, belonging, and community. Autistic self-advocacy movements have strengthened this framing, advancing the language of neurodiversity, pride, and acceptance.

But this very success sharpens contradictions. If autism is fundamentally an identity—a way of being—why should it depend on a psychiatric diagnosis? If diagnosis unlocks social rights, accommodations, and recognition, what about those who identify as autistic without a formal label? And what about those with profound impairments who require 24/7 support?

This paradox reveals one of the deep risks of our diagnostic culture: the merging of medicine and identity. A diagnosis can be empowering, but it can also be constraining, casting a person’s sense of self in clinical terms that may not align with their lived experience. As Lea & Hofmann (2022) observe, diagnoses do not simply describe conditions; they shape who people believe themselves to be. In the case of autism, this can be liberating for some, but it can also lock others into a framework that emphasises impairment rather than diversity.

Towards a post-diagnostic future?

If we live in an age of diagnosis, what might a post-diagnostic age look like? O’Sullivan is surely right: such a shift would require profound reforms in healthcare systems and a cultural recalibration of how we think about health, difference, and suffering. It would mean taking seriously the possibility that diagnoses, while helpful for some, can also harm – by overstating pathology, entrenching stigma, and straining systems with unnecessary treatment.

Dediagnosis offers one promising step. It reframes diagnoses as provisional tools – open to revision, reversal, or retirement when the label may cease to be useful – when they no longer serve wellbeing. In autism, this might mean acknowledging the tension between medical models of impairment and identity-based understandings of neurodiversity. In psychiatry more generally, it means resisting the drift toward ever-expanding definitions and reclaiming diagnosis as a careful, bounded practice.

If diagnosis is primarily a tool for identifying impairment, then there are clearly circumstances where the label may cease to be useful – when a person has adapted, found supportive environments, or no longer requires services.

Concrete examples:

• Adulthood reassessments: Some adults who were diagnosed as children later undergo re-evaluation and are told they no longer meet criteria—either because their early difficulties have diminished, or because diagnostic thresholds have tightened. This can bring relief for some but a profound sense of loss or invalidation for others, especially when autistic identity has become central to self-understanding.

• Access to services: In certain health systems, diagnosis is pursued primarily to secure support in education or employment. When those supports are no longer needed – or when resources are cut – clinicians may withdraw the diagnosis, effectively “dediagnosing” by administrative decision rather than clinical reality.

For some, dediagnosis can be liberating, signalling that they no longer require medical oversight. For others, it can feel like an erasure of identity and loss of recognition.

Rethinking the role of diagnosis

The question, then, is not whether autism itself can be “undiagnosed,” but whether culture should stop tying identity, rights, and recognition so tightly to medical labels. It might mean that support and accommodations depend on demonstrated needs rather than DSM categories.

It might mean recognising one’s identity without requiring any clinical impairment as the price of legitimacy.

And it might mean reserving diagnosis for those with significant clinical impairments, while allowing people to embrace their identity without needing to justify it in medical terms.

Such a shift would not be easy. But it may be the only way to resolve the tensions between diagnosis, impairment, and identity in the contemporary autism debate.

The challenge is enormous. But the first step is intellectual honesty: to admit that diagnosis, once seen as a path to clarity and care, can also deepen confusion. To build a healthier culture, we need not only the courage to diagnose, but also the wisdom to dediagnose.