Peer-reviewed article

Methylation profile scores in child and adolescent health: a practical guide.

Isabel K Schuurmans, Serena Defina, A P C Hermans, Alexander Neumann, Matthew Suderman et al. · 25 Aug 2026

Abstract

Methylation profile scores (MPSs) aggregate the effects of many DNA methylation (DNAm) sites across the genome into a single continuous value per individual. These scores are rapidly gaining traction in health research, as they translate DNAm patterns into summary measures that can be used to address many clinical and epidemiological questions. Most current research, however, is focused on adults, with little consideration given to transferability to children and adolescents. In this Review, we aim to provide a practical guide introducing MPSs and to discuss the current status of MPSs in child and adolescent populations. First, we introduce the diverse applications of MPSs, including the following: (1) use as exposure proxies to estimate exposures that are missing, under-reported, or hard to measure (eg, an MPS for exposure to maternal prenatal smoking); (2) use as biological proxies to summarise physiological processes such as inflammation (eg, an MPS for C-reactive protein); (3) use in risk stratification, where MPS variation is associated with risk of a future health outcome (eg, cardiometabolic disease); (4) use in diagnostics, where MPSs are used to detect disease states (currently most established for rare Mendelian syndromes and paediatric brain tumours); and (5) use in pharmacoepigenetics and treatment monitoring, an emerging area where MPSs are used to predict treatment response or track symptom change over time. Second, we outline specific considerations that apply to the developmental context, which explain why child and adolescent MPSs often differ from their later-life counterparts. Third, we provide recommendations on how to critically judge paediatric MPS studies. Finally, we conclude that, at present, with a few notable exceptions, most MPSs are better characterised as research tools at an early stage of development rather than clinical tools, and we highlight future directions for the field.