The conversation around neurodivergent support needs lacks a consistent framework. Autism has the DSM-5’s three-level system for autism severity — requiring support, requiring substantial support, requiring very substantial support — but this framework doesn’t extend to ADHD, dyslexia, dyspraxia, dyscalculia, and other non neurotypicality conditions and profiles. The result is a flattened discourse where all neurodivergence gets treated as equivalent, and where rising diagnosis rates trigger panic about “overdiagnosis” without any structural distinction between different levels of support need.
In this piece Ronnie Cane proposes a three-level framework for understanding neurodivergent severity that applies across conditions, is operationally defined rather than clinically assessed, and clarifies what the diagnostic expansion of recent decades has actually been capturing.
Why neurodivergence needs a severity framework beyond autism
The DSM-5 introduced autism’s current three-level system in 2013 when it collapsed the previous separate diagnoses — autistic disorder, Asperger’s syndrome, PDD-NOS — into a single Autism Spectrum Disorder diagnosis with support need specifiers. The levels describe functional impact: how much support does this person require to navigate social communication and manage restricted or repetitive behaviours?
ADHD has no equivalent framework. It has presentation types — inattentive, hyperactive-impulsive, combined — and technically has severity specifiers (mild, moderate, severe), but these are rarely used clinically and don’t describe support needs in any operationally meaningful way. Dyslexia, dyspraxia, dyscalculia, and other neurodivergent conditions have no severity framework at all. They are diagnosed as present or absent, with severity left to clinical impression rather than structured assessment.
This creates a problem for public discourse. When someone says they are autistic, ADHD, or dyslexic, we have no shared framework for understanding what level of support that implies. The person who manages their ADHD through calendar systems and a quiet workspace occupies the same diagnostic category as the person who cannot maintain employment without stimulant medication occupies the same category as the person who cannot perform basic self-care without full-time support.
The absence of a cross-condition severity framework fuels the “overdiagnosis” panic. Critics point to rising diagnosis rates and ask whether thresholds have been lowered, whether diagnosis has become fashionable, whether we are “medicalising” ordinary human variation. But this critique treats all diagnoses as equivalent — as though a Level 1 autism diagnosis and a Level 3 autism diagnosis represent the same claim, as though ADHD managed through self-knowledge and ADHD requiring daily medication are the same phenomenon.
A framework that distinguishes support needs across neurodivergent conditions clarifies what is actually being counted and what the diagnostic expansion has been capturing.
Neurodivergent Level 1 (Diagnosed): recognition without intervention
Level 1 describes neurodivergent individuals for whom diagnosis provides framework and self-understanding, but who do not require medical or pharmacological intervention to reach functional baseline. They may benefit from accommodation — flexible working arrangements, sensory-considerate environments, explicit communication norms, etc — but they can and do function without it when necessary.
What diagnosis provides at this level is explanation, not treatment. The person understands why they have always struggled with small talk, why open-plan offices drain them, why they lose track of conversations that others follow easily, why they need to see the whole picture before engaging with details. The framework organises a lifetime of experiences that previously lacked coherence. It allows them to stop blaming themselves for difficulties that are structural rather than personal.
Level 1 is not “mild” neurodivergence in the sense of being barely present. The neurological difference is real. The cognitive architecture is genuinely distinct from neurotypical processing. What makes it Level 1 is that the mismatch between that architecture and environmental demands can be navigated through understanding, strategy, and selective accommodation rather than medical intervention.
This is the level where the diagnostic expansion has been most dramatic. The person who would have been called “quirky” or “intense” or “difficult” in previous generations now has a framework. The woman whose ADHD presented as anxiety and disorganisation rather than hyperactivity now has an explanation for why willpower never worked. The man whose autism was masked by learned social scripts now understands why social interaction has always cost him energy it doesn’t seem to cost others.
Level 1 is not overdiagnosis. It is recognition catching up with reality. These people were always neurodivergent. They simply had no framework for understanding their experience and no language for communicating their needs.
Neurodivergent Level 2 (Medicated): pharmacological support for functional baseline
Level 2 describes neurodivergent individuals for whom diagnosis alone is insufficient — who require pharmacological intervention to reach functional baseline. Without medication, the mismatch between their cognitive architecture and environmental demands becomes disabling. With medication, functioning is possible, sometimes exceptional.
The distinction between Level 1 and Level 2 is not about the “severity” of the underlying neurodivergence. It is about what is required to function. A Level 2 individual is not “more ADHD” than a Level 1 individual in any meaningful sense. They may have similar cognitive architecture, similar processing patterns, similar strengths and challenges. The difference is that their particular configuration, in their particular environment, requires pharmacological support to navigate.
For ADHD, this typically means stimulant medication — methylphenidate, lisdexamfetamine — that modulates dopamine and norepinephrine to bring executive function within workable range. For autism with co-occurring anxiety or depression, it may mean SSRIs or other medications that address the secondary conditions created by living in a mismatched environment. For some, it means sleep medication, because the circadian differences associated with neurodivergence create sleep disruption that cascades into everything else.
Level 2 is where the “medicalisation” critique has some purchase — but not in the way critics intend. The concern is usually that we are turning ordinary human variation into medical conditions requiring treatment. But Level 2 is not ordinary variation being medicalised. It is genuine neurological difference that creates genuine functional impairment that is genuinely ameliorated by medication. The person is not being medicated into normality. They are being medicated into functionality (due to the square peg-round hole mismatch) — their own functionality, on their own terms, with their own cognitive architecture intact.
The growth in ADHD medication prescriptions over recent decades reflects the growth in Level 2 recognition. These are not people who have been convinced they have a disorder they don’t have. They are people who have spent years or decades struggling with executive function, trying every non-pharmacological strategy, and finally accessing medication that makes the difference willpower never could.
Neurodivergent Level 3 (Disabled): intrinsic impairment regardless of support
Level 3 describes neurodivergent individuals for whom impairment persists regardless of diagnosis, medication, or accommodation. Support needs are intrinsic, high, and constant. This is the population that genuinely cannot function independently in any environment — for whom no amount of understanding, strategy, or medication produces independent functioning.
Level 3 is what the word “disabled” properly describes. Not disability as social construction, not disability as mismatch between individual and environment, but disability as intrinsic limitation that no environmental modification fully resolves. The person may be nonverbal, may be unable to perform basic self-care, may require round-the-clock supervision. Their needs exceed what accommodation or medication can address.
For autism, this population is approximately 10–15% of diagnosed individuals. About a third of autistic people have co-occurring intellectual disability, which provides a rough proxy for the Level 3 population — those who would struggle to survive independently in any historical or cultural context, regardless of social attitudes or environmental design.
The critical point is that this population is small and stable. The diagnostic expansion of recent decades has not occurred at Level 3. The prevalence of severe autism — nonverbal, intellectually disabled, requiring very substantial support — has remained relatively constant even as overall autism prevalence has increased dramatically. What has grown is Levels 1 and 2. The “epidemic” is not an epidemic of disability.
This distinction matters because the “overdiagnosis” critique implicitly conflates all levels. When critics worry that rising diagnosis rates represent medicalisation of ordinary variation, they are treating Level 1 and Level 2 diagnoses as though they were claiming Level 3 status. They are not. They are claiming the available recognition of (“square peg”) genuine neurological difference that creates genuine difficulty in navigating (“round hole”) neurotypical environments — difficulty that understanding, medication, or both can substantially address.
What this new three-level framework for neurodivergent severity reveals about the "overdiagnosis" debate
The Fonagy Review interim report, published March 2026, frames rising ADHD and autism diagnoses as potentially concerning — raising questions about “medicalisation,” about whether diagnosis has become “incentivised,” about whether thresholds have shifted. The report notes that epidemiological estimates of underlying prevalence appear stable while administrative diagnoses have increased dramatically.
This framework clarifies what that divergence means. Epidemiological prevalence captures the full neurodivergent population — Levels 1, 2, and 3. Administrative diagnosis captures those who have sought and received formal recognition. The growth in administrative diagnosis represents Level 1 and Level 2 individuals finally being counted — not an expansion of Level 3, not threshold creep capturing people who don’t really have the condition, but recognition catching up with a population that was always there.
The “incentive” framing in the Fonagy Review gets the causation backwards. People are not seeking diagnosis because diagnosis has become fashionable or because they have been convinced they have conditions they don’t have. They are seeking diagnosis because diagnosis provides access to framework, to self-understanding, to medication where needed, and to accommodation in educational and employment contexts that would otherwise exclude them. The incentive is not artificial. It is the natural consequence of a system that gates support behind diagnostic thresholds.
The framework also addresses the strongest critique of neurodiversity perspectives. Critics often ask: “What about severely disabled autistic people? Isn’t it dismissive to frame autism as just ‘difference’ rather than disability?” The three-level framework answers this directly. Level 3 is real. It is disability-primary. It requires language of impairment and support needs that the “just difference” framing cannot capture. But Level 3 is 10–15% of the autistic population. The remaining 85–90% — Levels 1 and 2 — are where the mismatch framing applies. Their difficulties are real, but they are substantially context-dependent. With understanding, accommodation, or medication, functioning changes.
This is not a hierarchy of legitimacy. Level 1 is not “autism-lite” or “not really ADHD.” The neurological difference is genuine across all levels. What differs is what is required to navigate that difference — recognition alone, recognition plus medication, or support needs that exceed what either can provide.
This new neurodivergent severity framework makes visible what the diagnostic expansion has actually been capturing: a population that was always neurodivergent, always struggling, always present — but never counted, never understood, and never supported. That population is now being recognised. This is not overdiagnosis. It is the opposite.
Citations
American Psychiatric Association (2013) — Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). Arlington, VA: American Psychiatric Publishing.
Fonagy, P. et al. (2026) — Independent review into mental health conditions, ADHD and autism: interim report. Department of Health and Social Care.
