Connectome models of attention and autism show phenotype-specific features vary by age and diagnosis

An analysis examined feature consistency in transdiagnostic connectome-based models predicting sustained attention and autism symptoms across different datasets, fMRI scan types, and participant populations. The study included youth and adult participants with and without an autism diagnosis, though the specific sample size, study design, and setting were not reported. No specific intervention, comparator, or primary outcome was detailed.

The main findings were associative. Model similarity was associated with participant characteristics and predicted behavioral measures. Models predicting attention phenotypes shared more similar features with each other than with models predicting autism symptoms. Furthermore, the predictive features overlapped more between datasets that included participants of similar ages (youth vs. adult) and similar diagnostic status (autism vs. no diagnosis). No effect sizes, absolute numbers, or statistical measures (p-values or confidence intervals) were reported for these associations.

Safety and tolerability data were not reported. Key limitations include the unreported study type and sample size, the lack of statistical metrics to assess the strength of findings, and the associative nature of the results which preclude causal inference. The practice relevance is not reported. The analysis suggests functional connectivity patterns predicting individual behavioral differences may be phenotype-specific and vary with age and clinical diagnosis, but these are preliminary observations from an unreported methodological framework.