Trustworthy ML/AI for Aging Clocks: Preventing Systematic Prediction Bias in Biological Age Estimation

Machine learning (ML)- and artificial intelligence (AI)-based aging clocks are increasingly used to quantify physiological and molecular aging from omics and medical imaging data as distinct from chronological age. Here, we characterize a fundamental but underappreciated statistical limitation of commonly used ML/AI regression models for continuous outcomes: systematic prediction bias and its propagation to downstream association estimates. This issue becomes more challenging when the true outcome, biological age, is latent and therefore unobserved during ML/AI model training. We demonstrate that systematic prediction bias can distort and, in some cases, even reverse downstream association analyses that use aging clocks as ML/AI-predicted outcomes to assess their associations with exposures or clinical factors. For example, it can produce spurious associations suggesting that older predicted brain age is linked to better cognitive performance, or that older epigenetic age is associated with better kidney function. To address this problem, we introduce a principled and broadly applicable ML/AI regression framework based on constrained optimization, yielding better calibrated aging-clock estimates and valid downstream inference.