Predicting Alcohol-Related Memory Problems in Older Adults: A Machine Learning Study with Multi-Domain Features.

Area of Science:

• Neuroscience
• Psychiatry
• Machine Learning

Background:

• Memory problems are prevalent in older adults with a history of alcohol use disorder (AUD).
• Identifying predictors of alcohol-induced memory impairment is crucial for early intervention.
• Previous research has not comprehensively integrated multi-domain data for this classification.

Purpose of the Study:

• To investigate the utility of multi-domain features in classifying individuals with and without alcohol-induced memory problems.
• To identify specific features contributing to memory decline in older adults with AUD.
• To explore the neural underpinnings of alcohol-related memory impairment.

Main Methods:

• A machine learning framework, specifically a random forests model, was employed.
• 94 older adults (ages 50-81) with alcohol-induced memory problems were compared to a matched control group.
• Multi-domain features included resting-state brain connectivity, polygenic risk scores for AUD, alcohol consumption history, health consequences, and personality traits.

Main Results:

• The random forests model achieved high classification accuracy (AUC = 88.29%).
• Key predictors included hyperconnectivity in default mode network regions (except anterior cingulate cortex), polygenic risk scores for AUD, past alcohol consumption and consequences, and personality traits (neuroticism, harm avoidance).
• Hyperconnectivity in default mode network and hippocampal regions suggests neural dysregulation in individuals with memory problems.

Conclusions:

• Multi-domain features, including historical brain connectivity, genetics, lifestyle, and personality, are important for predicting alcohol-related memory problems in later life.
• The findings highlight the complex interplay of factors contributing to cognitive decline in older adults with AUD.
• This approach offers a promising avenue for early identification and potential prevention strategies for alcohol-induced memory impairment.