This 2016 Finnish human study was a followup to A study of DNA methylation and age:

“At the 2.55-year follow-up, we identified 19 mortality-associated CpG sites that mapped to genes functionally clustering around the nuclear factor kappa B (NF-κB) complex. None of the mortality-associated CpG sites overlapped with the established aging-associated DNAm sites.

Our results are in line with previous findings on the role of NF-κB in controlling animal life spans and demonstrate the role of this complex in human longevity.”

I was again impressed with the researchers’ frankness in the Discussion section:

“Our data do not provide a mechanistic link between the hypomethylation of these CpG sites and the risk of mortality.

Our data do not allow us to determine whether disrupted regulation of chromatin permissiveness underlies the increased mortality risk.

None of our top 250 mortality-associated methylomic sites were among the 525 common age-associated CpG sites that have been observed in more than one study.”

Regarding the lack of confirmation at the 4-year followup:

“The number of mortality-associated CpG sites was markedly reduced from the 2.55-years follow-up to the 4-years follow-up.

A substantial part of the genomic CpG sites might be constantly remodeled, and during 4 years, their methylation levels are likely to change to an extent that their predictive ability in our population is reduced. The longer follow-up time also allows more time for stochastic mortality determinants, such as trauma, to operate, which may thus weaken the role of the genomic predictors.”

The epigenetic clock method was investigated:

“The DNAm age has also been recently demonstrated to predict all-cause mortality in four different cohorts of elderly individuals and in Danish twins. However, the DNAm age was not predictive of mortality in our study.

One reason for the negative finding might be that individuals in our cohort were all very old at baseline (90 years), and death at this age likely has different underpinnings than at younger old ages and when assessed in cohorts with wider age spectra.”

http://www.impactjournals.com/oncotarget/index.php?journal=oncotarget&page=article&op=view&path[]=8278&path[]=24504 “Methylomic predictors demonstrate the role of NF-κB in old-age mortality and are unrelated to the aging-associated epigenetic drift”