Brain age as a biomarker for pathological versus healthy ageing - a REMEMBER study

Area of Science:

• Neuroimaging and computational neuroscience.
• Biomarkers for neurodegenerative diseases.

Background:

• Brain age prediction models offer a novel approach to assess brain health.
• Existing methods require validation for clinical utility and intervention assessment.

Purpose of the Study:

• To evaluate the clinical value of a brain age prediction model as an interpretable variable for brain condition.
• To explore brain age as a potential outcome measure for lifestyle intervention studies.

Main Methods:

• Utilized T1-weighted MRI data from the REMEMBER study (N=742) including cognitively healthy, SCD, MCI, and AD dementia subjects.
• Employed automated brain volumetry software (icobrain dm) to extract structural brain data.
• Developed a linear model combining volumetric features and sex to predict brain age and brain predicted age difference (BPAD).

Main Results:

• Mild cognitive impairment (MCI) and Alzheimer's disease dementia (ADD) patients exhibited increased brain age relative to chronological age.
• Brain age demonstrated superior classification accuracy across the Alzheimer's disease spectrum compared to BPAD and chronological age.
• Increased BPAD was observed in heavy alcohol drinkers, suggesting a link between lifestyle and brain aging.

Conclusions:

• Brain age and BPAD show promise as indicators of brain health.
• These metrics can potentially be used to evaluate the effectiveness of lifestyle modifications and interventions on brain health.