Epigenetic Clocks Do Not Strongly Reflect Inflammatory Status?

I recall being surprised by the study from a few years ago showing that early epigenetic clocks are insensitive to physical fitness, as demonstrated in twin studies using fit versus sedentary twin pairs. Given that a higher epigenetic age than chronological age, epigenetic age acceleration, correlates with increased mortality, and fitness status is similarly well correlated with mortality, it seems interesting that the machine learning approaches used to generate the clocks from raw epigenetic data by age managed to produce this outcome. Today's study is similarly surprising, and perhaps more so. It suggests that epigenetic age is not strongly correlated with inflammatory status, and yet it is well demonstrated that increased chronic inflammation in aging drives all of the common age-related conditions, raises mortality risk, and is in general an important component in degenerative aging.

The true promise of epigenetic clocks (and similarly, transcriptomic and other clocks) is to be able to test potential rejuvenation therapies, determining quickly and efficiently whether or not they work, and how good they are relative to other options. As things stand today the research and development communities spend far too much time and effort on marginal therapies. Some process by which poor approaches are cost-effectively winnowed out early on in the development process is very much needed. Ideally, an epigenetic clock measurement would be taken before and after an intervention is attempted, either in mice or in human trials, and provide an unambiguous result. Unfortunately, epigenetic clocks cannot be used in this fashion for so long as they have these gaps, unknown until discovered, in which important aspects of aging are not well reflected in epigenetic age.

Inflammation and epigenetic ageing are largely independent markers of biological ageing and mortality

Limited evidence exists on the link between inflammation and epigenetic ageing. We aimed to 1) assess the cross-sectional and prospective associations of 22 inflammation-related plasma markers and a signature of inflammaging with epigenetic ageing; 2) determine whether epigenetic ageing and inflammaging are independently associated with mortality. Blood samples from 940 participants in the Melbourne Collaborative Cohort Study, collected at baseline (1990-1994) and follow-up (2003-2007) were assayed for DNA methylation and 22 inflammation-related markers, including well-established markers (e.g., interleukins and C-reactive protein) and metabolites of the tryptophan-kynurenine pathway. Four measures of epigenetic ageing (PhenoAge, GrimAge, DunedinPoAm and Zhang) and a signature of inflammaging were considered.

Associations were assessed using linear regression, and mortality hazard ratios (HR) were estimated using Cox regression. Cross-sectionally, most inflammation-related markers were associated with epigenetic ageing measures, although with generally modest effect sizes and explaining altogether between 1% and 11% of their variation. Prospectively, baseline inflammation-related markers were not, or only weakly, associated with epigenetic ageing after 11 years of follow-up. Epigenetic ageing and inflammaging were strongly and independently associated with mortality, e.g. inflammaging: HR=1.41, which was only slightly attenuated after adjustment for four epigenetic ageing measures: HR=1.35. Although cross-sectionally associated with epigenetic ageing, inflammation-related markers accounted for a modest proportion of its variation. Inflammaging and epigenetic ageing are essentially non-overlapping markers of biological ageing and may be used jointly to predict mortality.


Is it surprising that an epigenetic clock constructed by selecting DNA methylation instances, a thousand or less out of many hundreds of thousands, that reflect calendar age is relatively insensitive to inflammation condition? I don't think so. If you want a "clock" that reflects inflammation condition, you could use the same selection algorithm to select DNA methylation instances that correlate with inflammation. However, such a clock would be less well correlated with age.

Posted by: John G. Cramer at August 11th, 2022 7:26 PM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.