Evaluation of epigenetic age acceleration scores and their associations with CVD related phenotypes in a population cohort
Authors:
- Olga ChervovaOCOlga Chervova
- Elizabeth ChernyshevaECElizabeth Chernysheva
- Kseniia PanteleevaKPKseniia Panteleeva
- Tyas Arum WidayatiTWTyas Arum Widayati
- Natalie HrbkovaNHNatalie Hrbkova
- Jadesada SchneiderJSJadesada Schneider
- Andrew RyabikovARAndrew Ryabikov
- Taavi TillmannTTTaavi Tillmann
- Hynek PikhartHPHynek Pikhart
- Martin BobakMBMartin Bobak
- Vitaly VoloshinVVVitaly Voloshin
Preprint
Abstract
Background Epigenetic age acceleration (EAA) is a measure that can be used to investigate the relationship between molecular, clinical and phenotypic data. However, which EAA score is best suited for which phenotype(s) is an open question. To address this issue, we have conducted a comprehensive comparison of multiple EAA scores using currently understudied Eastern European ageing population cohort (HAPIEE) which is richly annotated for diverse phenotypes. Our analysis was based on a subset (n = 306; aged between 45 and 69 years) of samples with available DNA methylation (DNAm) and phenotypic data. Results We calculated nine established EAA measures and performed statistical hypothesis testing to investigate associations between these scores and 18 cardiometabolic phenotypes. This was implemented by splitting the data into groups with positive and negative EAAs. We observed strong association between all EAA scores and sex, suggesting that in any analysis EAAs should be adjusted by sex. We found that some sex-adjusted EAA scores were significantly associated with several phenotypes (not necessarily the same phenotypes for different EAAs), indicating that some EAA scores are more phenotype-specific than others. We were able to replicate some of the associations in sex-specific subsets. Furthermore, the most of the EAA associations with cardio-vascular (except presence of Carotid Plaque (CP)) and lipids phenotypes were established in females but not males (and vice versa for CP and alcohol intake). This demonstrates that even after adjusting EAAs for sex, EAA-phenotype associations remain sex-specific, which should be taken into account in any downstream analysis involving EAAs. We observed that in some EAA-phenotype associations, negative EAA scores (i.e. epigenetic age below chronological age) indicated more harmful phenotype values, which is counter-intuitive. Overall, GrimAge was associated with more phenotypes than any of the other EAA scores. Conclusions EAAs are sex-specific and should be adjusted for sex in EAA-phenotypes association studies. Associations between EAAs and cardiometabolic phenotypes are sex-specific, even after adjusting for sex. For some EAAs, the direction of the association with phenotype is counter-intuitive. Our results can be used as a guidance on which EAA score to use for which phenotype(s).Keywords
All Related Versions
- Published version: Biology, 12 (1), 68.
This publication has 52 references indexed in Scilit:
- Prevalence and Predictors of Carotid Wall Triple Line Pattern in a General Population SampleArteriosclerosis, Thrombosis, and Vascular Biology, 2011
- Gamma-glutamyltransferase as a cardiovascular risk factorEuropean Heart Journal, 2006
- Sex differences in the relationships between BMI, WHR and incidence of cardiovascular disease: a population-based cohort studyInternational Journal of Obesity, 2006
- Increased Cardiac Expression of Tissue Inhibitor of Metalloproteinase-1 and Tissue Inhibitor of Metalloproteinase-2 Is Related to Cardiac Fibrosis and Dysfunction in the Chronic Pressure-Overloaded Human HeartCirculation, 2005
- Gamma Glutamyl TransferaseCritical Reviews in Clinical Laboratory Sciences, 2001
- Genomic Structure of Human Adrenomedullin GeneBiochemical and Biophysical Research Communications, 1994
- A Comparison of a Prospective Diary and Two Summary Recall Techniques for Recording Alcohol ConsumptionBritish Journal of Addiction, 1989
- Multiple Comparisons Among MeansJournal of the American Statistical Association, 1961
- THE GENERALIZATION OF ‘STUDENT'S’ PROBLEM WHEN SEVERAL DIFFERENT POPULATION VARLANCES ARE INVOLVEDBiometrika, 1947
- On the Interpretation of χ 2 from Contingency Tables, and the Calculation of PJournal of the Royal Statistical Society, 1922