Autism was first described as a clinical condition in the mid-twentieth century and was long considered of low prevalence. Since the 1990s, autism prevalence has increased substantially, particularly in high-income countries. This rise has prompted public and professional debate, including concerns of over-diagnosis and suggestions of an “autism epidemic.” Current research and conceptual analysis suggest that these interpretations are not supported by the available evidence (Mandy, 2025). Instead, changes in diagnostic practices, alongside shifts in social, educational, and economic environments, have reshaped how autism is recognised and understood (Fombonne, 2020; Lord et al., 2020).
Rising rates of autism diagnosis reflect changing diagnostic practices and social contexts rather than an epidemic.
The origins of autism as a clinical diagnosis
Autism was first identified as a distinct clinical syndrome in the 1940s by child psychiatrist Leo Kanner, who described a group of children with early-emerging differences in social engagement, communication, and behaviour, alongside a strong preference for sameness. These features continue to form the core of contemporary diagnostic definitions (Kanner, 1943; Mandy, 2025). Despite advances in genetics, neuroscience, and cognitive science, autism is still defined at the behavioural level. There is currently no biological marker that can be used for diagnosis confirmation. Identification relies on observed behaviour and developmental history, consistent with other psychiatric conditions (Lord et al., 2020).
Autism in contemporary diagnostic frameworks
Current diagnostic systems, including DSM-5-TR and ICD-11, define autism by persistent differences in social communication and social interaction, and restricted or repetitive patterns of behaviour, interests, or activities (American Psychiatric Association, 2022; World Health Organization, 2019). There has been a major conceptual shift in recent decades: autism is now widely understood as a dimensional condition rather than a discrete category (Mandy, 2025). Autistic traits are continuously distributed across the population, with diagnostic thresholds applied when traits are associated with clinically significant impact (Constantino & Charman, 2016; Lord et al., 2020). This dimensional perspective helps explain why diagnostic boundaries are necessarily flexible and why prevalence estimates are sensitive to changes in criteria and practice.
Autism is now widely understood as a dimensional rather than categorical condition.
The broadening of diagnostic criteria since the 1990s
Rising prevalence of autism diagnosis may be primarily explained by changes in diagnostic conventions (Mandy, 2025). Since the 1990s, criteria have broadened to include a wider range of presentations, including individuals without intellectual disability and those whose difficulties become apparent later in development (Fombonne, 2020). Earlier frameworks prioritised more severe and externally visible presentations, often associated with language delay. Current criteria allow for recognition of individuals with fluent language and more subtle social differences, particularly in adolescence and adulthood (Lai & Szatmari, 2020). Individuals diagnosed under these broader criteria continue to experience high levels of distress, functional impact, and autistic traits (Mandy, 2025). Empirical studies consistently signal elevated mental health needs and reduced quality of life among autistic individuals diagnosed today (Lord et al., 2020; Zeidan et al., 2022).
Broader diagnostic criteria have increased recognition without diluting clinical significance.
Disability, distress, and life outcomes
Autistic individuals show elevated rates of anxiety, depression, self-harm, and suicidality, alongside increased risks of educational under-achievement, unemployment, and social isolation (Hirvikoski et al., 2020; Cassidy et al., 2020). These difficulties often intensify during adolescence and adulthood, when social and occupational demands increase (Mandy, 2025). Longitudinal research suggests that many autistic adults experience cumulative disadvantage across health, economic, and social domains (Lord et al., 2020). Therefore, findings suggest that autism remains a condition that can have a significant functional impact for many people, challenging narratives that rising diagnosis reflects trivialisation or diagnostic excess.
High levels of disability and distress are consistently observed among autistic people.
The functions of an autism diagnosis
For many individuals, particularly those diagnosed later in life, autism provides a framework for understanding lifelong experiences of difference, difficulty, or marginalisation (Huang et al., 2020; Mandy, 2025). It may also support identity development and belonging, and, as research suggests, autistic identity can be associated with improved self-understanding and reduced self-blame, particularly when embedded in supportive social contexts (Cooper et al., 2022). Furthermore, a diagnosis can facilitate reasonable adjustments and accommodations in education, employment, and healthcare, improving person–environment fit rather than addressing individual deficits alone.
Diagnosis may function as both an explanatory and enabling process.
Changing environments and rising diagnostic rates
Social, educational, and economic environments have become more demanding in terms of social communication, flexibility, emotional regulation, and self-management (Mandy, 2025). From this perspective, rising diagnostic rates reflect a growing mismatch between autistic characteristics and contemporary environments rather than changes in underlying neurobiology (Mandy, 2025). Similar arguments have been made in broader disability research, emphasising the relational nature of disability (Shakespeare et al., 2022). This account reframes autism diagnosis as a response to changing social conditions as much as changing clinical awareness.
Environmental demands play a key role in shaping disability and diagnosis.
Autism and the neurodiversity perspective
A neurodiversity-informed framework conceptualises autism as a natural form of human variation rather than a pathology (Singer, 2017; Botha et al., 2024). From this perspective, social difficulties arise through interaction and relational mismatch. The “double empathy problem” further challenges deficit-based accounts by emphasising mutual misunderstanding between autistic and non-autistic people (Milton, 2012; Crompton et al., 2020). This framework supports a shift toward environmental adaptation and mutual understanding (Mandy, 2025). Neurodiversity perspectives do not deny disability or support needs but reframe them as context-dependent.
Neurodiversity perspectives emphasise relational and contextual factors.
Reframing debates about over-diagnosis
Debates about over-diagnosis often overlook the persistent and substantial needs of autistic people (Mandy, 2025). Higher prevalence rates may reflect improved recognition of previously overlooked groups, including individuals without intellectual disability and those diagnosed in adulthood (Fombonne, 2020). Current understanding of autism and associated research suggests attention should shift toward how societies respond to neurodevelopmental difference and whether environments adequately support autistic people (Mandy, 2025).
Debates about over-diagnosis risk obscuring ongoing unmet needs.
Conclusions
Contemporary autism diagnosis reflects broader criteria and changing environments rather than an epidemic or diagnostic excess (Mandy, 2025). Autistic individuals diagnosed today continue to experience significant disability and distress, underscoring the importance of inclusive environments and responsive support systems. Understanding autism “now” requires attention to both diagnostic practice and the social contexts that shape disability and participation (Mandy, 2025).
Where next?
The 2026 Emanuel Miller Conference will explore current debates around autism and ADHD, focusing on how scientific evidence, diagnostic practices, and rising demand for assessment intersect within contemporary neurodevelopmental care. Bringing together leading experts, the programme will examine how changes in diagnostic frameworks, service pressures, and social contexts shape who receives a diagnosis, when, and with what implications for young people, families, and services.
References
- American Psychiatric Association. (2022). DSM-5-TR: Diagnostic and statistical manual of mental disorders (5th ed., text rev.).
- Botha, M., Dibb, B., & Frost, D. M. (2024). Autistic identity and mental health: A systematic review. Autism, 28(1), 3–18.
- Cassidy, S., et al. (2020). Risk markers for suicidality in autistic adults. Molecular Autism, 11, 1–14.
- Constantino, J. N., & Charman, T. (2016). Diagnosis of autism spectrum disorder: Reconciling the syndrome, its diverse origins, and variation in expression. The Lancet Neurology, 15, 279–291.
- Crompton, C. J., et al. (2020). Neurotype-matching and communication. Autism, 24(8), 2106–2117.
- Fombonne, E. (2020). Epidemiology of autism spectrum disorders. Pediatric Research, 65, 591–598.
- Hirvikoski, T., et al. (2020). Premature mortality in autism spectrum disorder. The British Journal of Psychiatry, 217(1), 1–8.
- Huang, Y., et al. (2020). Late diagnosis of autism and its consequences. Autism, 24(8), 2058–2070.
- Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217–250.
- Lord, C., et al. (2020). Autism spectrum disorder. The Lancet, 392, 508–520.
- Mandy, W. (2025). What is autism, now? Proceedings of the Paris Institute for Advanced Study, 21. https://doi.org/10.5281/zenodo.17989621
- Milton, D. (2012). On the ontological status of autism. Disability & Society, 27(6), 883–887.
- Shakespeare, T., et al. (2022). The social model of disability: Thirty years on. Disability & Society, 37(3), 367–386.
- Singer, J. (2017). Neurodiversity: The birth of an idea.
- Zeidan, J., et al. (2022). Global prevalence of autism: A systematic review update. Autism Research, 15(5), 778–790.