A Computational Cognitive Biomarker for Early-Stage Huntington's Disease

Area of Science:

• Neuroscience
• Computational Psychiatry
• Genetics

Background:

• Huntington's disease (HD) is a genetic disorder with variable symptom onset.
• Early intervention is challenging due to uncertainty in predicting disease progression.

Purpose of the Study:

• To develop quantitative predictors for Huntington's disease before symptom onset.
• To utilize computational modeling for early diagnosis and understanding of HD.
• To differentiate pre-manifest HD (pre-HD) from healthy controls.

Main Methods:

• Neurocognitive phenotyping using the antisaccade conflict task.
• Computational modeling (accumulator-based model) of reaction time and error rates.
• Classification analysis based on model parameters.

Main Results:

• Symptomatic HD subjects showed clear behavioral differences (reaction time, error rates) compared to controls.
• Pre-HD individuals did not exhibit significant behavioral differences from controls.
• Classification models identified executive control parameters as key differentiators between pre-HD and controls.
• Response inhibition parameters declined only after symptom onset.

Conclusions:

• Computational modeling offers a powerful tool for predicting and classifying neurodegenerative disorders like HD.
• Executive control deficits are an early indicator of Huntington's disease, preceding symptom manifestation.
• Findings provide insights into the neural mechanisms underlying pre-symptomatic HD.