Brain scans reveal two distinct types of autism

Researchers examining neuroimaging data from approximately 1,000 individuals with autism spectrum disorder have identified evidence suggesting the condition encompasses at least two distinct biological subtypes, each characterized by fundamentally different patterns of neural communication. Through systematic analysis of brain scans combined with comparative studies using 20 genetically modified mouse models, scientists have delineated what they term a "hyperconnectivity" variant, in which brain regions demonstrate elevated levels of intercommunication, and a contrasting "hypoconnectivity" variant, marked by reduced neural signaling between brain regions. This discovery challenges the conventional understanding of autism as a single diagnostic entity and suggests that current clinical approaches may be conflating neurobiologically distinct conditions under a single diagnostic umbrella. The research represents a significant methodological advance in neuroscience, integrating human neuroimaging with animal model validation to establish biological foundations for previously recognized clinical heterogeneity. These findings emerge from collaborative research efforts that have systematized the analysis of large-scale neuroimaging datasets, providing quantifiable neurobiological markers that could reshape both scientific understanding and clinical practice.

The conceptual framework underlying autism diagnosis has long acknowledged substantial variation in presentation and severity among individuals meeting diagnostic criteria. However, the field has traditionally approached this heterogeneity as phenotypic variation arising from a common underlying etiology rather than as evidence of fundamentally distinct biological mechanisms. The push toward identifying discrete neurobiological subtypes reflects broader momentum in neuroscience toward precision medicine models, wherein treatments and interventions are increasingly tailored to specific biological substrates rather than applied uniformly across diverse patient populations. Psychiatric and neurological research has witnessed growing recognition that diagnostic categories established through behavioral observation may obscure important biological distinctions; previous research has suggested that conditions ranging from depression to schizophrenia likely encompass multiple biologically distinct presentations. In the context of autism, such distinctions become clinically urgent given the condition's high prevalence, its profound impact on individuals and families, and the current absence of interventions specifically designed for distinct biological subtypes. The timing of this research is particularly significant as the field grapples with improving intervention efficacy and understanding why certain therapeutic approaches demonstrate variable effectiveness across the autism population.

The research methodology employed represents a notable advancement in translational neuroscience. The analysis incorporated brain imaging data from nearly 1,000 individuals with autism, a sample size substantial enough to permit identification of patterns that might otherwise remain obscured in smaller studies. The integration of data from 20 genetically engineered mouse models provided crucial validation, allowing researchers to establish causal relationships between specific genetic variations and the observed neural communication patterns. This dual approach, combining human neuroimaging with mechanistic investigation in animal models, addressed a persistent challenge in neuroscience: distinguishing correlation from causation when examining human brain function. The identification of distinct hyperconnectivity and hypoconnectivity subtypes was not based on clinical presentation or symptom profiles but rather on objective measurement of neural connectivity patterns, suggesting these subtypes reflect genuine biological differences rather than subjective categorization.

For neuroscientists and clinicians engaged with autism spectrum disorder, these findings carry immediate practical implications. The existence of neurobiologically distinct subtypes suggests that current one-size-fits-all intervention approaches may be inherently limited in their potential efficacy. Individuals with hyperconnectivity patterns, characterized by excessive inter-regional communication, might theoretically benefit from interventions designed to modulate excessive neural signaling, while those with hypoconnectivity patterns might require fundamentally different therapeutic strategies aimed at enhancing neural communication. This distinction directly addresses a persistent clinical puzzle: why certain interventions demonstrate robust efficacy in some autistic individuals while proving ineffective or even counterproductive in others. Diagnostic stratification based on connectivity patterns could enable more rational allocation of healthcare resources and more informed prognostication. Furthermore, the findings suggest that future pharmaceutical development might benefit from targeting the specific neurobiological substrates underlying each subtype rather than pursuing broad-spectrum interventions. For individuals and families affected by autism, these discoveries potentially herald a transition from diagnostic approaches based on behavioral symptomatology to approaches grounded in underlying biological mechanisms.

These research findings illuminate a broader pattern evident across contemporary neuroscience: the progressive decomposition of behaviorally defined psychiatric and neurological syndromes into collections of biologically distinct conditions. The identification of autism subtypes parallels similar movements within other diagnostic categories, where advances in neuroimaging and genetics have revealed hidden biological heterogeneity beneath uniform diagnostic labels. This pattern carries profound implications for nosology, the scientific classification of diseases, suggesting that diagnostic manuals may require fundamental reorganization to reflect biological rather than purely behavioral distinctions. The hyperconnectivity and hypoconnectivity subtypes identified in autism research exemplify this shift, as they transcend observable behavioral patterns to identify mechanistic differences in how neural systems organize and communicate. The convergence of neuroimaging, genetics, and animal neuroscience in establishing these subtypes demonstrates the power of multi-method approaches in modern neuroscience. Additionally, the research contributes to ongoing philosophical and scientific debates regarding the nature of neurodiversity, raising questions about whether autism subtypes represent distinct conditions or variations along continuous biological dimensions.

Moving forward, several developments warrant close monitoring by researchers and clinicians. The National Institute of Mental Health and comparable international research bodies will likely prioritize replication and extension of these findings in independent cohorts, examining whether the identified subtypes remain stable across diverse populations and demographic groups. Researchers should anticipate that the next 24 to 36 months will see intensified investigation into the genetic foundations of hyperconnectivity and hypoconnectivity subtypes, potentially identifying specific molecular pathways amenable to targeted intervention. Pharmaceutical companies and academic institutions will probably commence development of connectivity-specific therapeutic candidates, representing a significant departure from current drug development approaches in autism. The clinical translation of these findings remains uncertain; implementation of connectivity-based diagnostic stratification would require development of standardized neuroimaging protocols, establishment of clear biological thresholds distinguishing subtypes, and integration of these approaches into clinical practice settings. Additionally, longitudinal studies tracking individuals across developmental periods will be essential for determining whether connectivity patterns remain stable across the lifespan or whether they shift in response to maturation or intervention. Stakeholders including the Autism Society and research-focused institutions should prepare for the likelihood that current understanding of autism's biological basis will undergo substantial revision in the coming years, with significant implications for research priorities, clinical practice standards, and resource allocation across the autism research and care ecosystem.