This study critically reevaluates the utility of brain-age models within the context of detecting neurological and psychiatric disorders, challenging the conventional emphasis on maximizing chronological age prediction accuracy. Our analysis of T1 MRI data from 46,381 UK Biobank participants reveals a paradox: simpler machine learning models, and notably those with excessive regularization, demonstrate superior sensitivity to disease-relevant changes compared to their more complex counterparts. This counterintuitive discovery suggests that models traditionally deemed less accurate in predicting chronological age might, in fact, offer a more meaningful biomarker for brain health by capturing variations pertinent to disease states. Our findings challenge the traditional understanding of brain-age prediction as normative modeling, emphasizing the inadvertent identification of non-normative pathological markers over precise age prediction.

Beyond Accuracy: Refining Brain-Age Models for Enhanced Disease Detection