A single strand of hair can act like a 'biological time machine' to predict autism in babies only a month old.
March 19, 2026
Original Paper
Machine-learning analysis of temporal molecular dynamics stratifies autism likelihood - a multinational study
medRxiv · 2025.11.19.25340581
The Takeaway
Using lasers to map elemental levels at 800 different points along a hair strand, scientists reconstructed a baby's chemical exposure history over time. This temporal 'tree ring' approach allowed them to identify children at high risk for autism with 96% sensitivity, long before behavioral symptoms usually emerge.
From the abstract
Absence of autism risk-stratification tools under 18 months hampers early intervention. In a multinational sample of 1697 participants, aged one month and older, we provide proof-of-concept that temporal molecular dynamics can stratify autism likelihood. Using laser-ablation-inductively-coupled-plasma-mass-spectrometry, we measured elemental intensities along growth increments of single hair strands at ~800 timepoints. We developed a first-stage model to stratify individuals into a lower autism