From infantile psychosis to autistic spectrum: An evolving understanding

Notice: Blog posts are not subject to review by MedLink Neurology’s Editorial Board. MedLink acknowledges using artificial intelligence to assist in the creation of blog posts.

The diagnostic boundaries of autism have shifted repeatedly over the past century, reflecting changing conceptions of brain, behavior, and development. Once viewed as a rare childhood variant of schizophrenia, autism is now recognized as a diverse neurodevelopmental spectrum. Understanding how its classification evolved provides insight into broader trends in psychiatry and neurology toward biologically grounded, dimensional approaches to mental and developmental disorders.

Early descriptions and the psychotic model (1910s–1950s)

The term autism first appeared in psychiatry in 1911, when Eugen Bleuler used it to describe withdrawal and self-absorption in adults with schizophrenia. For Bleuler, autismus represented a loss of contact with external reality—a symptom of psychosis rather than a developmental trait.

Three decades later, two independent descriptions reframed the term. In 1943, Leo Kanner at Johns Hopkins reported eleven children with “early infantile autism,” characterized by profound social aloofness, echolalia, insistence on sameness, and atypical object use (Kanner 1943). The following year, Hans Asperger in Vienna described “autistic psychopathy,” involving social detachment but preserved verbal ability and intelligence (Asperger 1944).

Despite these distinctions, both conditions were initially conceptualized within the language of childhood psychosis. In the mid-twentieth century, “infantile psychosis” remained a common label, influenced by psychoanalytic interpretations that emphasized cold, emotionally distant mothering (“Refrigerator Mother,” a term coined by child psychologist Bruno Bettelheim was emblematic) rather than neurodevelopmental pathology.

From psychosis to developmental disorder (1960s–1980s)

By the 1960s, advances in developmental psychology and the rejection of purely psychogenic explanations began to shift autism toward the domain of neurodevelopmental disorders.

• DSM-II (1968) classified autism under “schizophrenic reaction, childhood type,” continuing the psychotic framing.
• DSM-III (1980) introduced a separate category, “infantile autism,” within “pervasive developmental disorders,” marking a conceptual break from schizophrenia.

This reclassification paralleled emerging research demonstrating early onset, strong heritability, and specific patterns of language and social impairment distinct from psychotic illness. The concept of pervasive developmental disorders was further elaborated in DSM-III-R (1987), introducing diagnostic criteria that emphasized behavioral observation over inferred psychodynamics.

The expansion of subtypes (1990s–2010s)

DSM-IV (1994) expanded the pervasive developmental disorder category to include autistic disorder, Asperger disorder, childhood disintegrative disorder, Rett disorder, and pervasive developmental disorder not otherwise specified (PDD-NOS ). This model aimed to capture phenotypic diversity but often produced inconsistent diagnoses across clinicians and research centers.

During this period, neuroimaging and genetic studies reinforced the view of autism as a continuum of social-communication and behavioral traits. Findings of widespread cortical connectivity differences and polygenic influences blurred categorical boundaries, suggesting shared neurodevelopmental mechanisms across subtypes.

The spectrum model (2013–present)

DSM-5 (2013) unified all previous subtypes under a single category: autism spectrum disorder. This revision adopted two core domains—(1) deficits in social communication and interaction, and (2) restricted, repetitive patterns of behavior or interests—with dimensional severity ratings based on required support levels. The change aimed to improve diagnostic reliability and align with research showing that language ability, cognition, and adaptive functioning vary continuously rather than by discrete subtype.

Although the spectrum model improved consistency, it also generated controversy. Some clinicians and patients identified with Asperger disorder expressed concern about the loss of a distinct label and potential effects on service eligibility. Nevertheless, the move reflected a broader nosologic trend toward dimensional classification and away from rigid categorical constructs (not “yes” or “no,” but “how much”).

Future directions: toward biologically based stratification

Contemporary frameworks, such as the National Institute of Mental Health’s Research Domain Criteria (RDoC) and the World Health Organization’s International Classification of Diseases, 11th Revision (ICD-11), emphasize continuous traits (eg, social cognition, sensory reactivity, and behavioral rigidity) across diagnostic boundaries.

Emerging subtyping efforts integrate:

• Genomic clustering, identifying synaptic and chromatin-modifying gene networks.
• Electrophysiologic markers, such as altered mismatch negativity and gamma oscillation synchrony.
• Neuroimaging signatures, including atypical long-range connectivity and cerebellar-frontal circuit differences.

These biologically grounded models may eventually refine autism nosology beyond behavioral description, yielding “autism-related neurodevelopmental conditions” defined by molecular and circuit-level mechanisms.

Table. Evolution of Autism Classification in DSM and ICD

Key trends over time

• Shift from psychosis to development. Early psychotic models gave way to recognition of congenital, neurodevelopmental origins.
• Behavioral specificity. DSM-III introduced operationalized behavioral criteria, improving reliability.
• Spectrum unification. DSM-5 and ICD-11 consolidated subtypes under a single spectrum to capture dimensional variation.
• Toward biology. Future revisions are expected to integrate genetic and neural circuitry data to define subgroups mechanistically.

Conclusion

Over a century, autism has evolved from Bleuler’s symptom of psychosis to Kanner’s infantile syndrome to today’s multidimensional spectrum. This transformation parallels the trajectory of psychiatry and neurology toward neurodevelopmental, biologically informed frameworks. As genetic, imaging, and computational data converge, the next phase of classification may finally anchor autism in neural systems biology—transforming diagnosis from descriptive taxonomy to mechanistic understanding.

Selected references

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 3rd to 5th editions.

Asperger H. Die "autistischen Psychopathen" im Kindesalter. Arch Psychiatr Nervenkr 1944;117:76-136.

Kanner L. Autistic disturbances of affective contact. Nerv Child 1943;2:217-50.

Lombardo MV, Lai MC, Baron-Cohen S. Big data approaches to decomposing heterogeneity across the autism spectrum. Mol Psychiatry 2019;24(10):1435-50. PMID 30617272

Lord C, Elsabbagh M, Baird G, Veenstra-Vanderweele J. Autism spectrum disorder. Lancet 2018;392(10146):508-20. PMID 30078460

Rutter M. Incidence of autism spectrum disorders: changes over time and their meaning. Acta Paediatr 2005;94(1):2-15. PMID 15858952

Are you interested in contributing a post or becoming a guest blogger for MedLink? Contact us at editorial@medlink.com.