New research has provided compelling evidence that autism may encompass at least two distinct biological subtypes, differentiated by the strength of brain connections. Scientists at the Italian Institute of Technology in Rovereto found that some autistic individuals exhibit unusually strong links between different brain regions, a phenomenon termed 'hyperconnectivity', while others present with weakened connections, or 'hypoconnectivity'. These findings underscore the complex and varied nature of autism, challenging previous attempts to identify a single brain 'signature' for the condition.
The study, led by Alessandro Gozzi, investigated these connectivity patterns by first examining 20 strains of mice, each carrying a mutation in a gene associated with autism in humans. Functional magnetic resonance imaging (fMRI) scans of these mice revealed distinct patterns: 11 strains predominantly showed hypoconnectivity, while the remaining nine exhibited hyperconnectivity. Further analysis revealed that these opposing connectivity signatures were linked to different underlying biological mechanisms, with affected genes in hypoconnective mice interacting with proteins involved in synapses, and those in hyperconnective mice interacting with proteins related to gene regulation and the immune system.
Building on these animal model insights, Gozzi's team then analysed fMRI data from 940 autistic people and 1,036 age-matched control participants. Among the autistic cohort, 24 per cent displayed hypoconnectivity, and 17 per cent showed hyperconnectivity. This suggests that, at a minimum, there are two dominant and biologically distinct subtypes of autism. However, a significant proportion – 59 per cent – of the autistic individuals in the study did not neatly fit into either category, indicating that further diversity within the condition likely exists and warrants more exploration.
The research, which has been peer-reviewed, contributes significantly to the ongoing scientific understanding of autism. For years, brain imaging studies have sought a singular neurological characteristic for autism, but results have often been inconsistent, reporting either hyperconnectivity, hypoconnectivity, or a mix of both. Gozzi suggests that previous studies may have overlooked the inherent diversity within the autistic population. Natalie Sauerwald at the Flatiron Institute in New York, while noting that the exact number of autism subtypes remains unclear, commented that this study is crucial for explaining the heterogeneity of the condition and its biological underpinnings.
While the mouse models provide valuable insights, researchers acknowledge limitations. Human autism is influenced by hundreds of genes, each with small effects, meaning animal models may not capture the full spectrum of the condition. Additionally, some genes selected for the mouse study are also linked to developmental delay, potentially limiting the generalisability of these specific findings to all autistic people. The long-term goal for researchers is to create a comprehensive picture of autism's diversity, linking genetic factors to brain connectivity and, ultimately, to observed behaviours.