Keen to hear if anyone has compared how well Horvaths' DNA methylation-based ‘epigenetic clock’ relates to Blackburn's telomere length, as an assessment of the ageing process in different cultures/peoples.
Edinburgh certainly have data on both of those in the ageing cohort Lothian Birth Cohort 1936, if it was something you were in.
1: http://www.genomebiology.com/2015/16/1/25
2. http://www.ncbi.nlm.nih.gov/pubmed/21194798
Hi Paul,
That particular topic is something I am very interested in. I am currently working with a unique aging population that has telomere length and 450k data. Manuscript to come soon hopefully!
Lisa
Thank you, Donald and Lisa ,
I know your work Donald, appreciate you sending and wonder if you have directly compared the EC and TL?
Lisa, I look forward to reading about the results of your work in Vancouver - can you provide any info ahead of publication about what you are doing?
Cheers, Paul
Paul - Still a work in progress. We have a cohort from a unique population of centenarians, along with their 450k methylation profiles as well as telomere length amongs other variables. Hoping to have a manuscript out in the new year.
DNA methylation age and telomere length have only a weak negative correlation after you correct for chronological age. As an aside I mention that it is important to correct for chronological age before one relates DNAm age to telomere length because chronological age confounds the relationship between the two variables.
We correlated DNAm Age with telomere length in *adipose* tissue (Horvath et al 2014, PNAS, Title: Obesity accelerates epigenetic aging of human liver") and only found a weak and insignificant negative correlation between the two variables after correcting for chronological age (e.g., r = −0.28, P = 0.22). We and others have seen similar correlation coefficients in blood tissue but these appear to be unpublished.
Bottom line: the epigenetic clock relates to a biological process that is largely independent of telomere attrition and cellular senescence.
In the language of multivariate linear models:
if you fit the following regression model:
lm(DNAmAge~AgeAtBloodDraw+TelomereLength) then TelomereLength will only have a marginally significant p value.
Conversely, if you fit
lm(TelomereLength~AgeAtBloodDraw+DNAmAge) then DNAmAge will only have a marginally significant p-value. Caveat: bmulti-collinearities between the 2 covariates will make the coefficient estimates unstable.
Best,
Steve
Thank you, Steve, very much appreciated. And what of the input samples - is this US-based or from a wider sample? Kindly, Paul
Hi Paul,
the weak/insignificant relationship between leukocyte telomere length and (age adjusted) DNA methylation age can also be observed in European populations. The following article may be of interest:
Breitling LP, Saum K-U, Perna L, Schöttker B, Holleczek B, Brenner H. Frailty is associated with the epigenetic clock but not with telomere length in a German cohort. Clinical Epigenetics. 2016;8:21. doi:10.1186/s13148-016-0186-5.
Best,
Steve
- Badges
- Science topic
- Similar topics
- Biological Science
- Botany
- Pharmaceutical Botany
- Plant Extracts