Data for this study was taken from the National Longitudinal Study of Adolescent to Adult Health (Add Health) cohort, a nationally representative sample followed up for 25 years, beginning in adolescence (1994-1994) and with the 5th and final wave of interviews conducted in 2016-18, from which DNA methylation (DNAm) analysis was performed from a collected blood sample. A total of 4237 participants had complete data available for analysis, including DNAm, socioeconomic status, ethnicity, BMI, and tobacco and alcohol use. The 16 epigenetic clocks were characterized as either 1st, 2nd, and 3rd generation clocks, which were designed to approximate chronological age, capture differences in aging that relate to disease and mortality risk, and to approximate the pace of aging as it relates to physical and cognitive declines, respectively.
One takeaway was that the various clocks had a moderate degree of correlation between epigenetic age and chronological age, as well as only a moderate degree of correlation with each other. This second finding is not surprising, as they each are calibrated differently, and past studies have found that they have varying degrees of disease and mortality predictive value. For example, a study published in Clinical Epigenetics that included nearly 10,000 participants found that while some clocks had better predictive value for type 2 diabetes and heart disease, others were more predictive of COPD, lung cancer, and all-cause mortality. It may be that each clock may capture distinct aspects of aging or biological function.
Perhaps a more important takeaway is that even in this relatively young population, the 2nd and 3rd generation clocks seemed to capture biological aging in association with social and lifestyle factors known to be linked to mortality/morbidity. For example, the GrimAge (2nd gen), PCGrimAge (2nd gen modified version of GrimAge), and DunedinPACE (3rd gen) clocks predicted more rapid aging among people with little or no college education compared to people with a college degree (ranging from 0.8 to 6.5 years), as well as people living at or near poverty level vs. people with incomes greater than $100,000 per year. The lifestyle factor with the largest impact on biological age was obesity; severe obesity was linked to more rapid aging in 15 of the 16 clocks, with a range of approximately 1 to 6 years added. Weekly exercise was also associated with aging, with a lack of exercise marked by more rapid aging in 9 of 16 clocks. Alcohol and tobacco use had mixed findings, though generally, former smokers had accelerated aging in 2nd and 3rd generation clocks. The GrimAgeAA, for example, predicted that current smokers had advanced 7.16 years more than never smokers, one of the largest accelerants. Keeping in mind that the participants at the time of testing were at a mean age of 38.4, an additional 7.16 years is quite substantial.
This study suggests that at least some of the available methylation clocks are capturing biological changes at a somewhat early age, offering the possibility that the pace of aging (not just a prediction of chronological age) could be both monitored and at least possibly targeted for improvement. It also appears that the later generation clocks may have some ability to capture accelerated aging quite early in life. For example, the DunedinPoAm, developed by using an analysis of the rate of change in the integrity of organ-systems over time, assessed the rate of change of children enrolled in the Texas Twin Project. A study published in the journal Pediatrics found that socioeconomically disadvantaged children (mean age 12 to 13) had a faster pace of aging, lending weight to the theory that early life adversity may accelerate age-related declines, marked (or mediated by) epigenetic changes. It’s worth noting that this same biological clock predicted both overall and cardiovascular mortality in a nationally representative sample of US adults over age 50.
It may also be difficult to disentangle what elements specifically are driving an accelerated pace of change. For example, acute exposure to air pollution (PM2.5) has been associated with methylation changes in several biological clocks, even with short-term exposure. A methylome-wide association study using data from subpopulations of the Women's Health Initiative and the Atherosclerosis Risk in Communities studies found that discrete locations within DNA (Cytosine-phosphate-Guanine (CpG) sites) were associated with particulate matter exposure of various sizes, and those sites were linked to neurological, pulmonary, endocrine, and cardiovascular disease-related genes. A number of other studies have found that specific diets and nutrients, such as polyphenols, are inversely associated with biological aging; distinguishing causation from correlation will be difficult but quite important to help determine the optimal interventions to prevent accelerated aging.
Related Biotics Research Products: