A lipidome Aging Clock shows Age Acceleration in individuals with Autism

Abstract

Recent advancements in lipidomics and machine learning have been leveraged to investigate the prediction of biological age in individuals. This study delves into age acceleration patterns, entropy, and the potential role of dolichol as an aging biomarker. We introduce a novel aging clock combined with explainable AI that utilizes the lipid composition of the prefrontal cortex to predict the biological age of individuals, both those without known neurological conditions and those with autism, schizophrenia, or Down syndrome. Notably, significant age acceleration was observed in individuals with autism. Furthermore, entropy exhibits a significant increase around the age of 40, indicating potential dysregulation in the mevalonate pathway. Lastly, dolichol emerges as a potential biomarker. These findings underscore the feasibility of predicting biological age using lipidomics data, paving the way for further investigation into the intricate relationship between lipid alterations and prefrontal cortex aging, while offering valuable insights into the associated molecular mechanisms.