Stable Jobs Don’t Shield Black Workers From Faster Biological Aging, DNA Study Finds

A decades-long look at American work lives has delivered one of the most unsettling findings yet in the science of aging: holding a steady, full-time job for thirty years does not protect Black workers from the biological wear that accelerates aging. In a new study published in SSM – Population Health, researchers led by Wen-Hua Lai of the University of Michigan linked detailed employment histories from the Health and Retirement Study to DNA methylation measurements in more than 1,800 older adults, and discovered that racial gaps in biological aging persist even among people whose working lives look nearly identical on paper. The results challenge a comfortable assumption that employment stability is an equalizer, and instead point to what happens inside jobs, not just whether people have them, as the crucial variable in who ages faster.

The team reconstructed each participant’s work history from age 18 to 50, converting retrospective reports of full-time work, part-time work, caregiving, unemployment, and disability leave into a 33-year sequence of annual states. Using a technique called sequence analysis, which treats an entire life course as an ordered pattern rather than a snapshot, they identified four distinct employment trajectories. Roughly half of respondents followed a Stable Full-Time path, entering continuous full-time work early and staying attached through midlife. A smaller group, about 8 percent, experienced Late Workforce Entry, remaining out of paid work into their late twenties or thirties. About 30 percent cycled through Multiple Job Disruptions, and roughly 11 percent drifted into Long-Term Non-Employment, withdrawing from the labor force from their late twenties onward with little re-entry before age 50.

The trajectories themselves were clearly patterned by race and ethnicity. White adults were the most likely to follow the stable full-time path, at 52 percent, compared with 45 percent of Black adults and 39 percent of Hispanic adults. Black and Hispanic respondents were roughly twice as likely as White respondents to enter the workforce late, with adjusted models showing odds ratios of 2.26 and 2.45 respectively. Notably, once the researchers controlled for education, nativity, childhood health, and other factors, the racial differences concentrated almost entirely in delayed entry rather than in mid-career disruption or chronic joblessness. Nativity mattered too: U.S.-born adults had only about a quarter the odds of late entry compared with foreign-born adults, suggesting that migration timing and credentialing barriers partly shape Hispanic workers’ delayed starts.

To measure biological aging, the team turned to epigenetic clocks, mathematical models that read chemical tags called methyl groups at specific sites across the genome to estimate how fast the body is wearing down. They applied five widely used clocks: Horvath and Hannum, first-generation models trained simply to predict chronological age, and PhenoAge, GrimAge, and DunedinPACE, second-generation clocks designed to capture health-relevant aging, physiological dysregulation, and mortality risk. Epigenetic age acceleration, or EAA, is the gap between the clock’s estimate and a person’s actual age. Positive values signal that someone’s body is aging faster than the calendar says it should, a pattern previously linked to frailty, cardiovascular disease, cognitive decline, and earlier death, often before any clinical diagnosis appears.

The racial contrasts were striking and clock-dependent. Black adults showed significantly faster acceleration than White adults on DunedinPACE and GrimAge, the two clocks most tightly tied to stress-related biological wear and mortality, with standardized differences of roughly a quarter to a third of a standard deviation. Yet on the Hannum clock, Black adults actually showed slower acceleration, and Hispanic adults showed marginally slower Horvath acceleration. This is not a contradiction but a lesson in measurement: first-generation clocks mostly track chronological time, while second-generation clocks index the cumulative physiological toll that the weathering hypothesis describes, the idea that chronic exposure to economic insecurity, discrimination, and occupational hazard becomes embedded in neuroendocrine, inflammatory, and immune systems over decades.

The study’s central surprise came when the researchers asked whether employment trajectories themselves predicted epigenetic aging. They largely did not. After adjusting for covariates, none of the three non-stable trajectories showed a significant association with acceleration on any of the five clocks, relative to stable full-time work. Nor did employment trajectories significantly moderate racial differences in aging; joint statistical tests of all race-by-trajectory interactions were non-significant across every clock. In other words, the shape of a person’s work life, measured purely as timing, continuity, and attachment, carried little independent signal about how fast their cells were aging once broader social circumstances were taken into account.

But the within-trajectory comparisons told a different and more consequential story. Among people who had followed the Stable Full-Time path, the very trajectory usually treated as protective, Black adults still showed higher DunedinPACE acceleration than White adults, and the groups differed significantly on three of the five clocks. Stable employment, the analysis showed, was not the same experience across racial lines. Supplemental data revealed that White workers in this trajectory were far more often in professional and managerial occupations, about 38 percent, while Black workers were concentrated in service and sales roles, about 53 percent. The same thirty-year label concealed radically different exposures to job control, hazards, benefits, discrimination, and strain. Even after adjusting for broad occupational categories, Black-White differences on DunedinPACE and GrimAge persisted, implying that occupational sorting explains only part of the racialized conditions embedded within apparently similar careers.

A parallel pattern emerged within the Long-Term Non-Employment group, where Black adults again showed faster GrimAge and DunedinPACE acceleration than White adults. The composition of this group differed sharply by race: White respondents were overwhelmingly female, at 85 percent, and had more often left professional, managerial, or office positions, a profile consistent with resourced exits into homemaking or caregiving. Black respondents were more concentrated in service and production work and included a larger share of men, suggesting exits more often tied to job loss, displacement, disability, or constrained labor-market options. The researchers also found that the reasons behind non-employment in the Multiple Job Disruption group diverged by race, with White adults’ non-working years most often attributed to caregiving, Black adults’ to unemployment, and Hispanic adults’ to disability or medical leave. The same trajectory, in short, can encode very different biographies of stress and security.

The authors are careful about causality. Epigenetic age was measured at a single point in 2016, employment histories were reconstructed retrospectively and may carry recall error, and the DNA methylation subsample was non-random, with survival to biomarker collection potentially selecting healthier respondents and muting true racial differences. Small subgroup cells, particularly within the non-employment trajectory, mean some comparisons are descriptive rather than definitive. The trajectories also captured the timing and continuity of work but not wages, job security, workplace discrimination, or physical demands, all of which prior research has linked to accelerated aging. Adverse work conditions such as low job control, night shifts, and long hours have each been associated with roughly 1.4 to 2.1 years of accelerated epigenetic aging in earlier studies.

Even with those caveats, the implications are hard to ignore. The findings suggest that racial disparities in biological aging do not arise mainly because Black and Hispanic workers are sorted into unstable employment pathways; they arise, at least in substantial part, because the conditions within similar pathways are unequal, and because the health returns to stable work itself are diminished for Black Americans. That reframes the policy conversation: simply promoting employment stability or longer careers will not close aging gaps if the jobs available to racialized workers remain lower-paid, less protected, and more hazardous. What gets under the skin, this study argues, is not the absence of work but the unequal character of the work people do, and the cumulative weight of disadvantage that follows them from their first job to their last.

Subject of Research: Racial and ethnic disparities in life-course employment trajectories and epigenetic age acceleration among older U.S. adults

Article Title: Employment Trajectories and Biological Aging: Racial and Ethnic Disparities in Work Instability and Epigenetic Age Acceleration

Article References: Lai, W.-H., Lin, T.-S., Cho, T.-C., Meier, H. C., Ammous, F., & Leggett, A. N. (2026). Employment Trajectories and Biological Aging: Racial and Ethnic Disparities in Work Instability and Epigenetic Age Acceleration. SSM – Population Health, Article 101973. https://doi.org/10.1016/j.ssmph.2026.101973

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101973

Keywords: epigenetic clocks, biological aging, employment trajectories, health disparities, weathering hypothesis, DNA methylation, Health and Retirement Study, GrimAge, DunedinPACE, racial inequality, labor market, life course