Li Rihui, assistant professor at the Centre for Cognitive and Brain Sciences of the University of Macau (UM), and his research team have achieved significant breakthroughs in addressing the heterogeneity of autism spectrum disorder (ASD). By using functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) technologies, the team has revealed the atypical brain connectome of both symptom-based and genetic-based ASD subtypes. This provides a new perspective to address the heterogeneity in ASD research and can be applied to precise healthcare management and therapeutic interventions for individuals with ASD.

As one of the most prevalent neurodevelopmental disorders, ASD affects approximately 1 to 2 per cent of the population. Individuals diagnosed with ASD typically exhibit challenges in social communication and restricted repetitive behaviours. Understanding the intricate neural mechanisms underlying ASD has been a longstanding goal for researchers. However, the substantial heterogeneity in both causes and manifestations within the ASD population has posed a considerable challenge in finding a consistent neural signature for the condition. A consensus has emerged that ASD comprises various subgroups with diverse etiologies and developmental trajectories. Investigating the distinctive characteristics of these subgroups therefore offers promise in predicting symptom development and advancing precision intervention strategies for ASD.

One of the studies by the research team delved into the atypical functional connectivity patterns of patients with Fragile X syndrome (FXS). FXS is a genetic disorder frequently associated with ASD, serving as one of the most common heritable risk factors of ASD (about 40 to 50 per cent of patients with FXS will be diagnosed with ASD). The team collected resting-state fMRI data from 38 girls diagnosed with FXS, and compared them to 32 girls (control group) matched in IQ, social function, and age. Their results showed that compared with the control group, girls with FXS showed significantly greater resting-state functional connectivity of the default mode network, lower nodal strength at the right middle temporal gyrus, stronger nodal strength at the left caudate, and higher global efficiency of the default mode network. These aberrant brain network characteristics map directly onto the cognitive behavioural symptoms commonly observed in girls with FXS. An exploratory analysis suggested that brain network patterns at a prior time point were predictive of the longitudinal development of participants’ multidomain cognitive behavioural symptoms. These findings contribute valuable insights into the impact of FXS on the brain’s connectome and offer a potential tool for assessing the developmental trajectory of behavioural phenotypes in girls with FXS. Moreover, the study highlights the specific brain connectome patterns linked to the different subtypes and risk factors of ASD. The study has been published in Biological Psychiatry, a leading journal in psychiatry.

Prof Li is the first and corresponding author of the research, with Allan Reiss, professor at Stanford University, as the senior author. The research is supported by the National Institute of Mental Health in the US (File no: R01MH050047 and T32MH019908), the Maternal & Child Health Research Institute at Stanford University (File no: 1220552-152-DHPEU), the Start-up Research Grant of UM (File no: SRG2023-00015-ICI), and the Canel Family Fund. The full version of the article can be viewed at

In another study, the team employed a symptom-based subgroup identification method to explore the resting-state functional connectivity of ASD subgroups. They collected resting state EEG data and evaluated autism behavioural symptom scores from 72 children diagnosed with ASD, alongside 63 typically developing children matched in gender and age. Using this approach, which was based on clustering and participants’ multidomain autism symptom clinical scale scores (SRS-2—Social Responsiveness Scales-Second Edition, and Vineland-3—Vineland Adaptive Behavior Scales-Third Edition), two ASD subgroups were identified: one with more severe symptoms (sASD) and another with milder symptoms (mASD). Subsequent analysis of their EEG data uncovered divergent abnormalities in the resting-state functional connectivity of these two subgroups. Specifically, mASD group displayed increased functional connectivity in the beta band, while sASD group exhibited decreased connectivity in the alpha band. Significant between-group differences in global and regional topological abnormalities were identified in both alpha and beta bands. These results imply that variations in symptom severity might account for previous inconsistencies in research regarding the functional connectome of ASD children. The study has introduced a fresh perspective for addressing heterogeneity in ASD research and has been published in Cerebral Cortex, a leading journal in neuroscience.

This research is co-led by Prof Li and the team of Prof Yang Guang from the Chinese People’s Liberation Army (PLA) General Hospital. Li Yuhang, a doctoral student at UM, and Zhu Gang, a postgraduate student at the Chinese PLA General Hospital, are the co-first authors of the research. The research is supported by the Start-up Research Grant of UM (File no: SRG2023-00015-ICI), the General Project of Beijing Natural Science Foundation (File no: 7222187), and the Special Scientific Research Project of Military Family Planning (File no: 22JSZ20). The full version of the article can be viewed at

Group differences in the mean functional connectivity (FC) of selected intrinsic networks

Correlational analyses between brain intrinsic network FC and multidomain behaviours in the FXS (blue) and control (orange) groups

The clinical scale scores of two ASD subgroups

Comparison of global topological properties between mASD, sASD, TD, and ASD groups