Autism spectrum disorder is a prevalent neurodevelopmental condition characterized by substantial social difficulties. Animal studies have shown that oxytocin can reduce autism-like social deficits in rodent models by acting on a distributed set of brain regions often referred to as the social salience network (SSN). Whether a homologous SSN exists and has clinical relevance in humans with autism had not been established. The present study aimed to test SSN functional organization in adult men with autism, examine relationships between SSN connectivity and behavioral traits, and assess modulation of SSN connectivity by intranasal oxytocin.
The investigators assembled a human SSN atlas by collating open-access cortical and subcortical parcellations to approximate the rodent-derived SSN in humans. This atlas was applied retrospectively to a resting-state fMRI dataset collected from adult men with autism that originated from a previously published randomized, placebo-controlled oxytocin trial. The dataset included measures of social and repetitive behaviors and salivary oxytocin concentrations collected at baseline and after administration.
Using the SSN atlas, the authors computed region-of-interest (ROI) to ROI functional connectivity (FC) across the network from resting-state fMRI data. Analyses tested SSN-wide and sub-network FC correlations with behavioral trait measures (social and repetitive behavioral scores) and with salivary oxytocin concentrations at baseline and post-administration. Treatment-specific effects on FC following intranasal oxytocin versus placebo were evaluated using ANCOVA. The authors emphasize that the present work is a retrospective analysis of data from a randomized, placebo-controlled trial reported previously.
At baseline, several SSN sub-networks showed FC correlations with measures of social traits and repetitive behaviors. These associations indicate that variability in resting-state connectivity within the SSN relates to clinical expression of core autism features in the studied sample of adult men. The specific sub-networks implicated included circuits connecting somatosensory cortex with subcortical regions.
Baseline salivary oxytocin concentrations were associated with SSN connectivity in particular circuits. Notably, connectivity between the somatosensory cortex and the nucleus accumbens, and between the somatosensory cortex and the paraventricular nucleus, demonstrated strong sensitivity to oxytocin at baseline. These baseline associations suggest physiological relevance of peripheral oxytocin measures to central SSN functional organization in this cohort.
Following nasal spray administration, the principal oxytocin-induced change within the SSN was a strengthening of an amygdala–somatosensory circuit. The authors report this amygdala–somatosensory enhancement as the largest treatment-related FC shift detected in their analyses. Treatment effects were quantified using ANCOVA comparing post-administration connectivity while accounting for relevant covariates.
A key observation was that baseline FC within the amygdala–somatosensory circuit strongly predicted the magnitude of post-treatment FC change: individuals with lower baseline connectivity in this circuit tended to show greater increases after oxytocin. This pattern supports the idea that baseline network state may serve as a biomarker to forecast which individuals are most likely to exhibit neural responsiveness to intranasal oxytocin.
These results provide the first evidence described by the authors that a human SSN is clinically relevant in men with autism and that specific SSN circuits may offer candidate biomarkers for predicting oxytocin responsiveness. The analysis was retrospective and based on a previously published randomized, placebo-controlled trial; the preprint has not undergone peer review and should not be used to guide clinical practice. The authors report that all data produced in the study are available on reasonable request.
The authors stated that all procedures and data sharing were approved by the Universitair Ziekenhuis/Katholieke Universiteit Leuven Ethics Committee for Biomedical Research and conducted in accordance with the Declaration of Helsinki. The clinical trial is registered (Eudract 2014-000586-45). The paper lists multiple funding sources supporting the work and declares no competing interests.
In adult men with autism, retrospective SSN-focused analyses of resting-state fMRI revealed correlations between SSN connectivity and social/repetitive behavioral measures, baseline sensitivity of nucleus accumbens–somatosensory and paraventricular nucleus–somatosensory circuits to oxytocin, and an oxytocin-induced strengthening of an amygdala–somatosensory circuit. Baseline connectivity in this amygdala–somatosensory pathway predicted the degree of post-treatment connectivity change, suggesting potential utility as a predictive biomarker for neural response to intranasal oxytocin. The findings are presented in a preprint that has not yet been peer reviewed; data are available on request from the authors.