Longevity Briefs: A Faster ‘Ageing Clock’ Is Associated With Mortality

Longevity briefs provides a short summary of novel research in biology, medicine, or biotechnology that caught the attention of our researchers in Oxford, due to its potential to improve our health, wellbeing, and longevity.

The problem:

Not everyone ages at the same pace. Currently, the best way we have of estimating how rapidly someone is ageing is to use an epigenetic clock. Epigenetic clocks are algorithms that look at modifications to the DNA molecule, usually DNA methylation (the addition of molecular ‘tags’ to the DNA called methyl groups). These modifications alter how the genetic code is read without changing the code itself. Since they occur in a predictable pattern throughout life, they can be used to estimate someone’s age by comparing their epigenetic modifications to the population average. For example, an epigenetic clock may tell you that you have epigenetic modifications expected of a typical of a 45 year-old – you have an ‘epigenetic age’ of 45. If your actual, chronological age is 50, this could imply that you are ageing more slowly than the population average. Some companies now provide services where they will measure your epigenetic age so that you can find out if you are ageing faster or more slowly than average.

The problem with using epigenetic clocks in this way is that even though greater epigenetic age has been found to correlate with poorer health and greater mortality at the population level, there are still questions about what epigenetic age actually means for individual health. As an analogy, we might be able to predict the remaining lifespan of a car by counting the scratches on its surface, but scratches don’t cause a car to break down, nor will buffing them out extend its road life. There may also be outliers in which, for whatever reason, the scratches don’t correlate well with age. A better strategy might be to look at how quickly change is occurring from one measurement to the next in the same vehicle. This still doesn’t tell us that what we are measuring is meaningful, but it may be a more reliable indicator of how quickly wear-and-tear is building up. Here, researchers apply the same logic to epigenetic age.

The discovery:

Researchers looked at data from 699 participants in an Italian cohort study, aged 63 on average at the start of the study and who were followed for up to 24 years. All participants had DNA methylation data available at 2-3 time points over the course of the study. This allowed researchers to calculate epigenetic age using 7 different epigenetic clock, ranging from the original ‘first generation’ clocks (Hannum and Horvath clocks) to newer second- and third-generation clocks (DNAmPhenoAge, DNAmGrimAge, DunedinPOAm_38, and DunedinPACE). These clocks are all designed to measure epigenetic age based on DNA methylation, but have been ‘tuned’ to emphasise different methylation sites in an attempt to produce estimates that are increasingly relevant to human health.

The main finding was that a greater increase in epigenetic clock readout from one measurement to the next was associated with a significantly higher risk of death for 5 out of 7 clocks, independent of starting epigenetic age. In other words, regardless of what epigenetic age was at the start of the study, a more rapidly increasing epigenetic age was associated with greater risk of death. When researchers then combined baseline epigenetic age with change over time, the relationship with mortality became even stronger. Second and third-generation clocks outperformed first generation clocks, which is unsurprising as these clocks were specifically designed produce estimates that correlate with mortality.

The implications:

These findings suggest that the way epigenetic age measurements change over time is predictive of mortality and should be incorporated together with snapshots of epigenetic age to produce estimates that are more relevant to human health. If someone’s epigenetic clock is accelerating rapidly, it could signal that their health is deteriorating in ways that predict mortality, even if their baseline biological age wasn’t alarming.

This still doesn’t answer one of the most problematic questions concerning epigenetic age: if an intervention lowers epigenetic age in humans, will this actually translate to improved health and reduced mortality? Until this question is answered, the practice of using epigenetic clocks as a way to measure the success of an anti-ageing intervention will remain somewhat dubious. Even so, epigenetic clocks may still have value for predicting future health and informing lifestyle interventions. Anything that makes their predictions more accurate can only be a good thing.