Additional assays were performed between 2018 and 2019 for DNA methylation. Exclusion criteria included surviving a hematopoietic stem cell transplant, being non-English speaking, and having a genetic or neurodevelopmental syndrome linked to cognitive impairment.
Neurocognitive testing and clinical data collection
A comprehensive neurocognitive exam was provided to all participants alongside biospecimen collection. Researchers evaluated attention, processing speed, memory, and executive function. Additional data collected from medical records included cumulative chemotherapy, radiation doses and fields, and demographics.
Participants also provided body mass index and chronic health conditions during a clinical exam. The Infinium MethylationEPIC BeadChip (850 K CpG sites) was used to generate genome-wide DNA methylation profiling data.
There were 1413 adult survivors of childhood cancer and 292 non-cancer controls included in the final analysis, the former of whom were aged a mean of 26 years. The age at evaluation did not significantly differ between cases and controls.
Epigenetic age acceleration in survivors vs controls
The central nervous system (CNS) treatment group comprised 77% of acute lymphoblastic leukemia survivors. In comparison, Hodgkin lymphoma and Wilms tumor were more common in non-CNS treated patients, with rates of 32% and 11%, respectively.
Biological aging was assessed with the PCPhenoAge, PCGrimAge, DunedinPACE, Horvath, and Hannum epigenetic clocks. Significantly increased epigenetic age acceleration (EAA) was reported by all 5 clocks among survivors vs non-cancer controls, alongside reduced telomere length.
These associations remained when adjusting for covariates. Additionally, the PCPhenoAge, Horvath, and Hannum epigenetic clocks indicated significantly greater EAA among non-CNS-treated survivors vs CNS-treated survivors.
Associations between EAA and neurocognitive performance
In neurocognitive assessments, a performance of 0.19 standard deviations worse on a test of visual-motor processing speed was observed among non-CNS-treated survivors in the third tertile of PCPhenoAge EAA vs those in the first tertile. Worse non-verbal reasoning was also found in patients of the third tertile vs the first tertile.
Investigators also linked EAA to neurocognitive dysfunction across all domains when using PCGrimAge, with patients in the third tertile performing a third of a standard deviation worse on attention variability, visual-motor processing speed, and memory span vs the first tertile. These links remained significant when adjusting for education.
Worse neurocognitive results were also reported among CNS-treated survivors with EAA. Overall, these results indicated worse neurocognitive function among long-term childhood cancer survivors with EAA.
“EAA may identify survivors at risk for accelerated cognitive aging and serve as an efficacy biomarker for neurocognitive interventions in at-risk populations,” wrote investigators.
References
- Researchers raise concerns about faster aging, possible early-onset dementia, for children and young adult cancer survivors. University of Rochester Medical Center. January 7, 2026. Accessed January 13, 2026. https://www.eurekalert.org/news-releases/1111871
- Williams AM, Phillips NS, Dong Q, et al. Epigenetic age acceleration, telomere length, and neurocognitive function in long-term survivors of childhood cancer. Nat Commun. 2025. doi:10.1038/s41467-025-65664-5