Longevity Briefs: How Well Do Biological Ageing Clocks Predict Age-Related Frailty?

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:

How do we actually measure how quickly someone is ageing? If we want to know whether a treatment slows ageing down, we could give it to a group of people and observe them over multiple decades to see if they developed fewer age-related diseases or lived longer than expected. Unfortunately, this approach is simply too expensive and lengthy to be practical, and so the solution that most researchers opt for is to use DNA methylation clocks (a form of epigenetic clock). These are algorithms designed to estimate biological age based on the presence of molecular ‘tags’ on the DNA molecule called methyl groups. Methylation is an epigenetic modification that alters the way the DNA is read, and is thought to play a role in the ageing process by suppressing important genes or increasing the activity of other genes in an undesirable way.

An increasing number of services offer methylation age measurements so that people can know their epigenetic age and try to lower it. While there’s nothing wrong with this, it is often incorrectly stated as fact that a reduction in epigenetic age means that ageing has been reversed or that health has been improved. While treatments specifically aimed at reversing epigenetic age in animals do appear to slow ageing, all we can really say about human epigenetic age is that it correlates with age-related decline. How well does it correlate? Researchers are still in the process of answering that question.

The discovery:

In this study, researchers conducted a systematic review and meta-analysis – essentially a large-scale analysis combining data from many existing studies examining the relationship between epigenetic clocks and age-related frailty. They narrowed their search down to 12 studies that met their criteria for quality and could provide suitable data. These studies involved a total of 28,325 participants, with a median age of 65.2 years. They included both cross-sectional studies (where participants’ epigenetic age and frailty were assessed as a snap-shot in time) and longitudinal studies, in which participants were followed up over an extended period to assess health outcomes.

The researchers aimed to examine the associations between various different DNA methylation-based biological age metrics and frailty. Frailty was usually measured either with the frailty index (which is based on the number of accumulated health deficits) and the frailty phenotype (which is based on measurements of specific physical functions like grip strength). The biological age metrics used were:

• DNA methylation age: The age predicted by methylation clocks. Comparisons between individuals control for actual chronological age.
• Age deviation: The difference between methylation age and chronological age.
• Epigenetic-age acceleration (EAA): A more sophisticated statistical approach that aims to estimate how much faster or slower a person is ageing by estimating what their epigenetic age should be, then comparing it to the actual value.

The study found that while overall DNA methylation age and age deviation showed no clear association with frailty, EAA was associated with significantly higher frailty rates for some epigenetic clocks. The epigenetic clocks showing statistically significant associations were the Hannum EAA, PhenoAge EAA, GrimAge EAA, DunedinPoAm and DunedinPACE. These clocks differ from each other in that they measure different methylation sites within the DNA and are sometimes ‘trained’ on other factors besides methylation, such as blood markers, which can make them better or worse at predicting certain types of outcome. Most of these clocks are so-called second or third generation clocks, essentially meaning that they are more modern and sophisticated. Most of the earlier epigenetic clocks did not show significant association with frailty.

Furthermore, when researchers looked only at longitudinal studies and not cross-sectional studies, only GrimAge EAA was significantly associated with frailty. GrimAge is distinct from other epigenetic clocks due to its focus on predicting mortality risk in addition to biological age.

The implications:

The results of this study could be described as mixed. Since different epigenetic clocks are designed to focus on different factors when making their biological age predictions, they can be better or worse at predicting different outcomes such as frailty. This research suggests that more sophisticated methods for estimating biological age can produce results that correlate well with frailty. This still doesn’t necessarily mean that epigenetic changes cause ageing and frailty, or that having a lower epigentic age means you are ageing more slowly. It does however suggest that epigenetic age acceleration could be a useful indicator for risk of frailty, which could then prompt interventions aimed at preventing it.