Biological aging predicts midlife cognitive decline, especially in those raised in poverty

Children from low-income families may be more vulnerable to age-related cognitive decline, according to a new twin study published in Aging. The researchers found that people who show signs of faster biological aging—measured using DNA from blood samples—tended to experience greater declines in IQ from childhood to midlife. This link between biological aging and cognitive decline was especially pronounced among those who grew up in disadvantaged socioeconomic conditions.

As people grow older, they tend to age at different biological rates. Some develop chronic illnesses or experience cognitive decline earlier than others, even if they are the same chronological age. To explore why this happens, researchers have turned to the concept of “epigenetic age,” which uses chemical markers in DNA—specifically, patterns of DNA methylation—to estimate how quickly the body is aging biologically.

DNA methylation refers to chemical changes that affect how genes are expressed without changing the underlying genetic code. These changes can be shaped by lifestyle, environmental exposures, and stress. Over the past decade, scientists have developed a set of predictive tools known as “epigenetic clocks,” which estimate a person’s biological age based on these methylation patterns. When a person’s biological age exceeds their chronological age, this is referred to as “epigenetic age acceleration,” and it has been associated with increased risk of disease, functional decline, and early mortality.

Although some studies have found that accelerated epigenetic aging is linked to lower cognitive functioning, the overall evidence has been inconsistent. Many earlier studies only looked at cognition at a single point in time, did not account for early-life environments, or used only one type of epigenetic clock.

The new study—led by Sophie Bell, a clinical psychology doctoral student at the University of Virginia; Eric Turkheimer, a professor of psychology at the University of Virginia; and Christopher Beam, an associate professor of psychology and gerontology at the University of Southern California—was designed to address these gaps. To this end, the researchers analyzed data from 287 participants in the Louisville Twin Study, a decades-long project that began tracking twins born between 1950 and 1997.

Free daily newsletter

“The Louisville Twin Study originally began as a study of childhood physical and cognitive development but has since become the longest-running twin study in the United States and a rare
resource for understanding lifespan aging,” explained Bell, the first author of the study.

“Over the past five years, twins who were followed through infancy and adolescence, now in middle adulthood, were invited back for comprehensive assessments of their cognitive, physical, and mental health. Middle adulthood is increasingly recognized as a window when modifiable factors may influence the earliest, preclinical stages of dementia. Yet very few studies have the data to connect midlife aging with the earliest developmental periods.”

“With the Louisville Twin Study, we saw a unique opportunity to study biological and cognitive aging in midlife while leveraging decades of rich developmental data. As we’ve studied aging across the lifespan, it’s become clear that we need better tools to capture variability in aging and risk for conditions like dementia and heart disease.”

“Epigenetic markers, specifically DNA methylation-based ‘epigenetic clocks,’ have emerged as promising tools because they can be measured at any point in life and have been linked to outcomes ranging from cognitive decline to cardiovascular disease, cancer, and mortality. We were especially excited about their potential in cognitive aging research, though we also noticed inconsistencies in how predictive they were of cognitive decline and in the algorithms researchers have chosen to use across studies.”

Participants in the Louisville Twin Study had undergone cognitive testing in childhood and again in midlife, around age 52 on average. DNA samples collected in midlife were used to estimate each person’s biological age using five different epigenetic clocks. These included both older algorithms focused on matching chronological age (first-generation clocks) and newer ones that incorporate additional health-related data (second-generation clocks).

The study also included measures of each participant’s childhood socioeconomic status, based on parental occupation and associated educational and income levels. This allowed the research team to test whether early-life disadvantage might interact with biological aging to influence changes in cognitive ability over time.

By using a twin design—especially comparisons between identical twins—the researchers were able to control for genetic differences and shared family environments. If one twin showed more epigenetic age acceleration and also experienced more cognitive decline, this would suggest a connection that cannot be explained by family background or genetics alone.

The five DNA methylation age measures were grouped into two factors through statistical analysis. The first group included the older, first-generation clocks (Horvath, Hannum, and Horvath Skin and Blood), while the second included the newer, second-generation clocks (GrimAge and PhenoAge), which are designed to capture risk of disease and mortality more directly.

The study’s main finding was that second-generation measures of biological aging were linked to greater declines in IQ from childhood to midlife. Importantly, this association held even after accounting for genetic and early environmental factors shared by twins. In other words, among identical twins raised in the same household, the one with faster epigenetic aging as measured by second-generation clocks tended to show a steeper drop in cognitive ability.

No such association was found for first-generation epigenetic clocks, suggesting that not all measures of biological age are equally informative for predicting cognitive decline.

“As we expected, the second generation DNA methylation clocks, PhenoAge and GrimAge, predicted midlife cognitive decline, while the first generation clocks did not,” Bell told PsyPost. “These second generation measures which incorporate biomarkers of physiological health have already shown improvements over first generation measures in predicting physical functioning, morbidity, and lifespan. Our findings suggest that cognitive decline is closely tied to broader age-related biological processes captured by DNA methylation changes.”

The effect of accelerated biological aging on cognitive decline was even stronger for individuals who grew up in low-income households. The researchers found that the relationship between biological aging and midlife cognitive ability depended in part on childhood socioeconomic status. Among twins raised in economically disadvantaged environments, the one with higher biological age also tended to show more cognitive decline, suggesting that early-life stress may amplify the long-term effects of accelerated aging.

“We found that the relationship between epigenetic aging and cognitive decline was stronger in twins who grew up in low socioeconomic status families, suggesting that early-life disadvantage may make individuals more vulnerable to the effects of broader biological aging on brain health,” Bell explained.

Smoking, a known risk factor for both aging and cognitive decline, was also strongly linked to accelerated epigenetic age in this sample. When the researchers included smoking as a covariate in their models, the association between biological aging and cognitive decline became weaker, though it remained in the same direction. This suggests that some of the observed effects may be partly explained by smoking-related methylation patterns, especially in the GrimAge clock, which was designed to capture lifetime tobacco exposure.

While the study has several strengths—including its longitudinal design, use of validated cognitive measures, and genetically informative twin sample—it also has some limitations. The sample was predominantly white and from a single geographic area, which may limit generalizability. The measure of childhood socioeconomic status, while based on historical census data, focused only on parental occupation and did not include other aspects such as income, education, or neighborhood quality.

In addition, the study did not include data on adult socioeconomic status or other life-course factors that might influence both epigenetic aging and cognitive outcomes. Although the twin design improves causal inference, it still cannot rule out all possible confounding factors. The cognitive assessments were focused on overall IQ, which captures general intellectual functioning but does not isolate specific domains like memory, attention, or processing speed.

Future research could benefit from larger and more diverse samples, including measures of adult SES, health behaviors, and domain-specific cognitive abilities. Studies might also explore whether interventions aimed at slowing biological aging—such as stress reduction, improved healthcare access, or lifestyle changes—could help protect against age-related cognitive decline, especially in those who experienced early adversity.

“One question that remains is how smoking fits into this picture,” Bell said. “We know that GrimAge is designed to capture the effects of lifetime smoking on the epigenome. When we accounted for smoking, the effect of DNA methylation age on cognitive decline was no longer statistically significant, yet smoking did not entirely explain the relationship either. Future studies are needed to disentangle the relationship between smoking, epigenetic aging, and cognitive health.”

The study, “Second generation DNA methylation age predicts cognitive change in midlife: the moderating role of childhood socioeconomic status,” was authored by Sophie A. Bell, Christopher R. Beam, Ebrahim Zandi, Alyssa Kam, Emily Andrews, Jonathan Becker, Deborah Finkel, Deborah W. Davis, and Eric Turkheimer.