A child in a wheelchair receives ramps, accessible toilets, modified desks, extra time between classes, and assistance with materials. No one suggests these accommodations are indulgent. No one proposes that the child should try harder to walk.
A child with PTSD or sensory processing difficulties who has an emotional outburst in class receives exclusion. The outburst is a disability manifestation — involuntary, neurological, documented. But because the disability is invisible, the system classifies it as behaviour. And behaviour gets punished.
This distinction — between disabilities that receive support and disabilities that receive sanction — is not grounded in science. It is grounded in which disabilities make institutions uncomfortable. Understanding why it persists, and what it costs, is the foundation of everything else in this series.
The neuroscience is not ambiguous
The premise underlying the double standard is that emotional dysregulation, sensory processing differences, and trauma responses are fundamentally different from recognised learning disabilities — less biological, more controllable, closer to character than to condition.
This premise does not survive contact with the evidence.
Research published in Nature Neuroscience in 2024 identified for the first time the precise brain regions dedicated to emotion regulation — specific neural circuits in the prefrontal cortex, anterior cingulate, and orbitofrontal areas (Bo et al., 2024). These are measurable neurological structures with chemical substrates involving serotonin, cannabinoids, opioids, and GABA systems. They are not willpower. They are not character. And developmental studies show they mature on the same trajectory as other executive functions — following the exact same pattern as the cognitive processes we already accommodate.
Sensory processing disorder affects 5–16% of school-aged children — as common as ADHD. Groundbreaking UCSF research in 2013 used brain imaging to demonstrate measurable structural differences and altered white matter connectivity in children with SPD (Owen et al., 2013). Yet SPD has never been formally recognised in the DSM or ICD, leaving millions without diagnostic legitimacy.
Trauma responses involve documented structural brain alterations — smaller hippocampal volume, overactive amygdala, underactive prefrontal cortex (Bremner, 2006). These responses are involuntary. The body reacts to perceived danger before the thinking brain can intervene. It is not a matter of choice.
A 2018 NeuroImage study compared brain activity in children with dyslexia and dyscalculia. Despite different behavioural presentations, brain activity profiles were more similar than expected, showing substantial neural overlap in how they deviated from controls (Peters & De Smedt, 2018). The same structural parallels exist between learning disabilities and emotional dysregulation: both involve brain structure differences, disrupted functional connectivity, neurotransmitter system variations, and involuntary manifestation.
The science settled this. What hasn’t caught up is institutional practice.
What visibility buys you
The disparity is driven by three interlocking biases, none of which have a scientific basis.
Visibility creates credibility. Physical disabilities involve wheelchairs or hearing aids. Learning disabilities demonstrate through testing. Emotional dysregulation lacks visible markers. When the condition doesn’t announce itself, institutions default to scepticism.
Perceived controllability shapes response. Weiner’s Attribution Model explains the mechanism: conditions perceived as uncontrollable, like dyslexia, elicit support. Conditions perceived as controllable, like behavioural differences, elicit blame. The perception is wrong — the neuroscience confirms these responses are not under voluntary control — but it determines institutional behaviour nonetheless.
The moral model persists. Children with behavioural differences continue to be labelled “bad,” “lazy,” “defiant,” or “attention-seeking.” These are moral categories applied to neurological conditions. As the law firm Kingsley Napley concluded in 2025, there is “something inherent within our education and youth justice systems that punishes difference.”
Thomas Nagel’s concept of moral luck applies directly: people cannot be morally assessed for what is not their fault. Being born with a visible disability versus an invisible one represents what Nagel calls “constitutive moral luck” — unchosen characteristics that profoundly affect how you are judged. If we should not hold people responsible for factors beyond their control, treating emotional dysregulation as more blameworthy than a learning disability is philosophically incoherent.
The scale of the failure
The consequences of this distinction are documented in the sectors’ own data.
In Scottish schools, pupils with Additional Support Needs are excluded at a rate of 28.9 per 1,000, compared to 5.8 for pupils without ASN — a gap of nearly five to one (Scottish Government, 2024/25). Across England and Wales, 80% of children cautioned or sentenced within the youth justice system have special educational needs or neurodivergence (UCL Institute of Education / Michael Sieff Foundation, 2025). Of young people in custody, an estimated 12% have ADHD, 15% are autistic, and between 60% and 90% have speech, language, or communication difficulties. These conditions are routinely not identified until after a child enters the justice system — routine screening before sentencing does not exist.
The trajectory is consistent. A child whose emotional dysregulation is treated as a behaviour problem rather than a disability manifestation is excluded from school. The excluded child enters the justice system at higher rates, at earlier ages, receives longer sentences, and reoffends more frequently.
In employment, the pattern continues. The disability employment gap for autistic adults stands at 31.4% — the lowest employment rate of any disability group. The disability pay gap is 12.7% overall, rising to 27.9% for autistic employees. Adults with ADHD are 60% more likely to be dismissed. These are not the outcomes of individual failure. They are the outcomes of systems that do not accommodate invisible disabilities with the same seriousness as visible ones.
When conditions multiply
The evidence above describes individual conditions. But neurodivergent adults rarely manage one condition in isolation.
Approximately 80% of adults with ADHD have at least one co-occurring psychiatric disorder (Choi et al., 2022). For autistic adults, a UK-led meta-analysis reported current anxiety prevalence of 27% and lifetime anxiety of 42% (Hollocks et al., 2019). The co-occurrence of ADHD and autism itself is remarkably common — a pooled prevalence of 38.5% among autistic individuals (Rong et al., 2021).
This matters because when conditions co-occur, they do not simply add risk. They multiply it. Sun et al. (2019) followed 2.7 million individuals in Sweden and found mortality risk increased in a dose-response fashion with the number of psychiatric comorbidities. For ADHD with four or more co-occurring conditions, the hazard ratio was 25.22 — compared to 1.41 for ADHD alone. Someone managing multiple conditions is not in a slightly worse version of the same situation. They are in a qualitatively different risk category.
Two landmark UK studies quantify the outcome. O’Nions et al. (2023), using a primary care database of nearly 10 million people, found autistic people without intellectual disability had a life expectancy reduction of over six years. O’Nions et al. (2025), in the first worldwide study to calculate life expectancy for adults with diagnosed ADHD, found reductions of 6.78 years for males and 8.64 years for females.
Yet UK policy frameworks — the Equality Act, NICE guidelines, occupational health assessment protocols — all default to single-condition models. They assess and respond to conditions individually. A single reasonable adjustment — a written agenda, a flexible start time, asynchronous communication — may not be managing one condition. It may be managing load across multiple interacting conditions simultaneously. Current frameworks do not account for this.
What works
This is not a problem without solutions. Effective alternatives exist and are operating in UK schools now.
In Oxfordshire, where 70 schools have participated in restorative practice training since 2019, 90% of trained schools have seen a reduction in suspensions and 83% have seen a reduction in permanent exclusions. In Barnet, restorative justice trained schools achieved a 51% reduction in exclusions, while non-trained schools in the same authority saw a 65% increase over the same period.
Scotland’s trajectory provides the clearest evidence that policy commitment produces results. Permanent exclusions fell from 21 cases in 2012/13 to just 1 in 2022/23 — at a time when England’s permanent exclusions rose from 4,632 to 6,495. This reflects Scotland’s “needs not deeds” principle and the full incorporation of the UN Convention on the Rights of the Child into Scots law in 2024. The five-to-one ASN exclusion disparity shows the work is far from complete, but the direction of travel demonstrates what is achievable.
Sensory rooms, trauma-informed whole-school approaches, and nurture groups all show measurable impact. In Muskegon County, Michigan, pupils were 56% more engaged in classroom activities after using sensory rooms. The Oxfordshire Relational Schools Programme — combining restorative, trauma-informed, and attachment-aware practice — reports schools halving suspensions, with staff describing reduced panic about high-level behaviour.
These interventions work for all pupils, not just those with identified needs. They reduce socioeconomic disparities alongside disability disparities. And they are more cost-effective than the exclusion and custody pipeline they replace.
The structural argument
The neuroscience establishes that emotional dysregulation, sensory processing differences, and trauma responses have the same neurobiological basis as recognised learning disabilities. The multi-morbidity evidence establishes that these conditions interact physiologically and compound each other, with measurable consequences for life expectancy. The educational data establishes that the system excludes children for disability manifestations it has a legal duty to accommodate. The employment data establishes that the same pattern continues into adult life. And the intervention evidence establishes that effective alternatives exist.
What connects all of this is a single structural flaw: the systems designed to support disabled people were built around a model of disability that is visible, stable, and binary. Invisible disabilities break that model. When the condition doesn’t announce itself, the institution doesn’t see it. When the person announces it themselves, the institution uses their competence at self-announcement as evidence they don’t need help.
That mechanism — competence treated as counter-evidence — is the subject of the next piece in this series.
This piece is adapted from The Compassion Gap, a research paper examining the neurobiological, philosophical, and practical case for equal accommodation of invisible disabilities. The full paper, with 60+ sources, is available as a free download.
The Broken by Design series examines the gap between UK disability policy and practice across education, employment, and financial services. Previously published: Measured by Neglect. Next: The Competence Trap.
References
Bo, K., et al. (2024). A systems identification approach using Bayes factors to deconstruct the brain bases of emotion regulation. Nature Neuroscience. doi:10.1038/s41593-024-01605-7
Bremner, J. D. (2006). Traumatic stress: effects on the brain. Dialogues in Clinical Neuroscience, 8(4), 445–461.
Choi, W.-S. et al. (2022). The prevalence of psychiatric comorbidities in adult ADHD compared with non-ADHD populations. PLOS ONE, 17(11), e0277175.
Hollocks, M.J. et al. (2019). Anxiety and depression in adults with autism spectrum disorder. Psychological Medicine, 49(4), 559–572.
O’Nions, E. et al. (2023). Estimating life expectancy and years of life lost for autistic people in the UK. The Lancet Regional Health — Europe, 36, 100776.
O’Nions, E. et al. (2025). Life expectancy and years of life lost for adults with diagnosed ADHD in the UK. British Journal of Psychiatry, 226(5), 261–268.
Owen, J. P., et al. (2013). Abnormal white matter microstructure in children with sensory processing disorders. NeuroImage: Clinical, 2, 844–853.
Peters, L., & De Smedt, B. (2018). Dyscalculia and dyslexia: Different behavioral, yet similar brain activity profiles during arithmetic. NeuroImage: Clinical, 18, 663–674.
Rong, Y. et al. (2021). Prevalence of ADHD in individuals with autism spectrum disorder. Research in Autism Spectrum Disorders, 83, 101759.
Scottish Government. (2025). School attendance, absence and exclusions statistics 2024–25.
Sun, S. et al. (2019). Association of psychiatric comorbidity with the risk of premature death among children and adults with ADHD. JAMA Psychiatry, 76(11), 1141–1149.
UCL Institute of Education / Michael Sieff Foundation. (2025). Achieving justice for children with SEND and neurodivergence in the youth justice system.