Brain Age Gap as a Predictive Biomarker for Neurodegenerative Disease Progression: A Systematic Review and Meta-analysis
SAI VENKATA MANOJ KOTHARU1, Simranjeet S Nagoke3, Aditi Agarwal4, Binay Panjiyar5, Govind singh Mann6, DHEEKSHITA SAI KORLAKUNTA7, Korlakunta Bhavana2
1medicine, osmania medical college, 2osmania medical college, 3Government medical college jammu, 4Bharathi Vidyapeeth Medical college Pune, 5Northwell health, 6Sant Paramanand Hospital, 7kakatiya medical college
Objective:
To perform a meta-analysis assessing the diagnostic performance of MRI-derived brain age gap (BAG) in differentiating patients with Alzheimer’s disease (AD), mild cognitive impairment (MCI), and Parkinson’s disease (PD) from healthy controls.
Background:
One promising biomarker for neurodegenerative diseases is the brain age gap (BAG) , which is the difference between chronological age and brain age as determined by MRI. Although elevated BAG has been linked to Parkinson's disease (PD) , Alzheimer's disease (AD) , and mild cognitive impairment (MCI) , no meta-analysis has measured it's diagnostic value across disorders .
Design/Methods:
we conducted a comprehensive search of PubMed , PMC and relevent imaging journals up to October 2025 for studies that reported group-wise brain age gap (BAG) means and standard deviations, using structural T1-weighted MRI and machine learning to estimate brain age. A total of 3560 participants from six eligible studies ( AD=1240 ; MCI = 920; PD = 1080 ; controls = 320 ) were included. Pooled standardised mean differences ( SMD ) in BAG between patients and controls were calculated using random-effects meta-analyses.
Results:
The patients brain appeared to age faster than those of the control group, with a significantly higher brain age gap (BAG) (pooled SMD=1.12; 95% CI, 0.89–1.36; p<0.001). AD had the highest SMD (1.35; 95% CI, 1.08–1.61), with lower values seen in MCI (0.89; 95% CI, 0.65–1.13) and PD (0.98; 95% CI, 0.75–1.21). The model architecture (deep learning vs. atlas-based; p=0.03) contributed to the moderate heterogeneity ( I²=58%). Robustness was confirmed by sensitivity analyses.
Conclusions:
MRI-derived BAG effectively distinguishes between healthy ageing and neurodegenerative patients, with biggest effects in AD. Our results validate BAG as a transdiagnostic biomarker for tracking the course of a disease. Clinical translation and the creation of guidelines will be facilitated by standardised reporting and increased research in other disorders.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.