This work challenges deficit-oriented narratives around perception in autism and highlights shared strengths of implicit perceptual cohesion across neurodiversity.

In this large-scale (n = 470) pre-registered online behavioural study involving two different tasks - no significant group differences emerged.

@plos.org

This one was featured in the @natcomms.nature.com editors highlights page - a lovely recognition of the quality and importance of this work. Well done team!

This opens exciting avenues: for example, exploring whether similar neurochemical-computational links exist in value-based or affective learning settings (where anxiety might play a stronger role).

Even though uncertainty processing is tightly linked to anxiety, we found these computational signatures were not simply explained by trait anxiety. However, individuals higher in trait anxiety were faster after the reversal of the task-structure suggesting they adapted more quickly to change.

In the context of a motor learning task, we show that the way people update beliefs under uncertainty maps onto excitatory neurochemistry in motor cortex. Higher M1 Glx was tied to stronger prediction-error updates and lower inferred volatility.

Suggests a mechanism by which postnatal maternal anxiety could shape early sensitivity to environmental unpredictability.

For anyone interested in developmental neurocognition, early risk for anxiety, and prediction-error processing, this opens up a compelling pathway.

Key findings:

Prediction-error responses localised to infant mPFC and only evident when controlling for infant attention via gaze.

The strength of these responses scaled with maternal trait anxiety – infants whose mothers were more anxious showed greater activation to unexpected stimuli.

Just found Brónagh on BlueSky - go give her some love
@bronaghm.bsky.social

Suggests that when studying anxiety and learning, we should consider both sources of uncertainty (noise + volatility), not just volatility alone.

Great advance for computational psychiatry and understanding how anxiety influences adaptation in changing environments.

Key take-aways:

Noise and volatility interact in shaping learning behaviour and modelled learning rates.

Anxious traits modulated how people responded especially when noise was high and volatility low: more lose-shift, less adjustment.