How “Clocking in” Ages Us: An Examination of Work History,Work Quality,and Accelerated Epigenetic Aging in Older Adulthood

Abstract

Consistent employment, especially secure, high-wage employment, has well-documented associations with lower risk of later-life morbidity and mortality, and accelerated biological aging may underlie these associations. Although research evaluating employment and accelerated epigenetic aging is growing, questions remain about the implications of chronic work-related exposures for accelerated aging. This study uses longitudinal data from the Health and Retirement Study (HRS) and the 2016 HRS Venous Blood Study (n = 3,000) to evaluate how work-related experiences throughout midlife are associated with accelerated epigenetic aging. Results show that a history of not working for pay and a history of poor work quality (i.e., job insecurity, insufficient work hours, low wages) among workers in midlife are associated with accelerated epigenetic aging in later life. Symptoms of depression and health behaviors partially attenuate these associations. Overall, findings suggest that chronic work-related exposures are critical yet overlooked antecedents of accelerated aging.

Keywords

employment quality epigenetic aging HRS work history

Employment has a well-documented association with a range of health outcomes, such as chronic conditions, disability, and mortality risk (e.g., Brand 2015; Burgard and Lin 2013; Ross and Mirowsky 1995). Although substantial research documents the benefits of paid work for health (Brand 2015; Ross and Mirowsky 1995) and the pernicious effects of unstable and insecure work for health (Burgard and Lin 2013; Kalleberg 2018), we know less about how work-related experiences “get under the skin” to impact biological aging. This gap in the literature precludes understanding of how work-related exposures can impact aging processes. Although nascent research finds evidence of accelerated aging among workers in sales/clerical, service, and related occupations (Andrasfay and Crimmins 2023; Andrasfay et al. 2023; George et al. 2021; Schmitz et al. 2021) and among workers experiencing psychosocial job strain in European countries (Freni-Sterrantino, Fiorito, d’Errico, Robinson, et al. 2022; Freni-Sterrantino, Fiorito, d’Errico, Virtanen, et al. 2022), we know very little about the role of chronic work-related exposures in midlife (e.g., labor force participation, work characteristics) for accelerated aging despite the salience of chronic exposures for epigenetic aging processes (Simons et al. 2021).

The present study uses longitudinal data from the Health and Retirement Study (HRS; 1992–2016) and the 2016 HRS Venous Blood Study (VBS) to examine how work history (e.g., inconsistent employment) and work quality history (e.g., chronic job insecurity) in midlife are associated with accelerated epigenetic aging (PhenoAge, GrimAge, and DunedinPACE) among older adults. In this study, we situate work-related experiences as key social determinants of epigenetic aging in later life. Additionally, we examine whether associations between work experiences and epigenetic aging are sensitive to the inclusion of mental health (e.g., depressive symptoms) and health behaviors (e.g., smoking), which may be potential pathways linking work-related exposures with epigenetic aging.

This study has three main contributions to the literature on work and health. First, we extend prior research by examining two key aspects of working lives: duration of time working for pay and the quality of work throughout midlife. Specifically, we assess whether consistent paid work is linked to slower biological aging, whereas poor-quality jobs, especially long-term exposure to these jobs, may be linked to accelerated biological aging even among those with stable employment. This possibility highlights the need to consider work quality alongside employment status to better understand work’s varied effects on aging. Second, prior research provides novel insight into the role of occupation and some work characteristics for accelerated aging (e.g., Andrasfay and Crimmins 2023; Andrasfay et al. 2023; Freni-Sterrantino, Fiorito, d’Errico, Robinson, et al. 2022; Freni-Sterrantino, Fiorito, d’Errico, Virtanen, et al. 2022; George et al. 2021; Schmitz et al. 2021), yet this research focuses on work-related exposures at one point in time. We build on this research to consider how histories of employment and work quality in midlife can shape accelerated epigenetic aging in later life. As such, our study points to the importance of the chronicity of work experiences in midlife. Finally, the examination of mental health and health behaviors as risk factors that may partially attenuate linkages between work experiences and accelerated epigenetic aging informs theories of work and epigenetic aging and points to additional areas of potential intervention. Understanding how work-related experiences in midlife are associated with accelerated aging directly and through psychosocial pathways may provide insight into the ways that working lives shape indicators of healthy aging.

Background

Social Exposures, Epigenetic Aging, and Epigenetic Clocks

Recent methodological advances and biomarker data collection efforts in geroscience have led to the development of several epigenetic clocks. Broadly, epigenetic clocks are biomarkers of aging that are designed to predict age-related health outcomes, such as chronic conditions and disability, years or decades before their onset. Epigenetic clocks are based on DNA methylation changes across the epigenome that can influence gene expression, which can, in turn, influence the aging process (Belsky et al. 2020, 2022; Horvath 2013). Prior research has shown how epigenetic changes are strong predictors of age-related conditions such as cancers, cardiovascular disease, and Alzheimer’s disease (Brunet and Berger 2014; López-Otín et al. 2013). As such, biomarkers based on epigenetic changes may provide further insight into how and why health differences in later life exist across a variety of health conditions.

To understand biological risk differences for people of the same chronological age, researchers have turned to epigenetic clocks. “Accelerated aging” refers to an individual (or population group) who is epigenetically older than their chronological age. For example, a 65-year-old with an epigenetic age of 70 has five years of accelerated aging. Concomitantly, “decelerated aging” refers to people who are epigenetically younger than their chronological age. Accelerated epigenetic aging is of interest because it is linked to age-related health conditions decades prior to the onset of clinical symptoms (Chen et al. 2016; Faul et al. 2023; Hillary et al. 2020; Marioni et al. 2015; McCrory et al. 2021).

Epigenetic changes are also highly malleable to social and environmental exposures across life. Social science and public health researchers have used these measures to investigate how life course exposures are biologically embodied to influence later life health (Moffitt 2020). But not all exposures influence epigenetic changes equally. Research points to the timing and chronicity of exposures as two key aspects. Except for severe acute events, such as war or a cancer diagnosis, DNA methylation changes in adulthood tend to be more gradual and linked to chronic exposures (Simons et al. 2021). In fact, prior research has shown that long-term social exposures explained roughly 60% of variation between individual differences in accelerated epigenetic aging as measured by the DNA methylation-based clocks in the United States (Simons et al. 2021).

This study is grounded in a stress and life course perspective to understand how exposure to work-related characteristics throughout midlife is associated with accelerated epigenetic aging. The stress process model (Pearlin et al. 1981) posits that exposure to stressful events, such as unemployment or low-quality work, can initiate stress responses that undermine health and well-being. Importantly, stress can proliferate across the life course and accumulate over time to impact long-term patterns of health (Pearlin et al. 2005). For example, persistent or sporadic unemployment can lead to economic strain, family conflict, deleterious coping strategies (e.g., smoking), and mental health challenges, which, in turn, can contribute to dysregulation of biological systems and accelerated epigenetic aging. A stress and life course perspective demands attention to the duration of stress exposures, such as cumulative exposure to poor work quality throughout midlife. We build on prior research by situating this study within a stress and life course perspective and evaluating the duration and chronicity of labor force participation and job quality throughout midlife on accelerated biological aging among older adults.

Employment Histories and Epigenetic Aging

Employment is strongly linked to health, well-being, and mortality risk (Caputo, Pavalko, and Hardy 2020; Frech and Damaske 2012; Ross and Mirowsky 1995). Paid work provides socioeconomic resources, reduces economic strain, and enhances sense of purpose, social integration, and psychological well-being (Brand 2015; Burgard and Lin 2013; Ross and Mirowsky 1995). Whereas healthier individuals are more likely to work, studies confirm that employment’s benefits persist after accounting for selection (Klumb and Lampert 2004). Stable, consistent employment further amplifies these advantages (Caputo et al. 2020; Frech and Damaske 2012), contributing to better mental health, fewer functional limitations, and longer life (Caputo et al. 2020).

Based on this prior research, we hypothesize that consistent employment throughout midlife will be linked to slower epigenetic aging. We posit that work histories (e.g., consistent employment or consistent unemployment), rather than individual events of brief unemployment, will be strongly linked to accelerated epigenetic aging, especially during peak earning years in midlife. Moreover, although prior research documents the adverse effects of financial hardship for biological aging (e.g., Simons et al. 2016, 2021), we lack understanding of whether exclusion from the labor force, especially long-term exclusion, contributes to accelerated aging. Despite this possibility, prior research has not considered linkages between work histories and epigenetic aging, limiting our knowledge of how long-term patterns of work get under the skin to impact epigenetic aging.

Several possible pathways could link work histories with epigenetic aging. For instance, consistently not working for pay could contribute to economic hardship, which is associated with accelerated aging (Simons et al. 2016, 2020). Unemployment and intermittent employment may also activate psychosocial responses, such as health-harming health behaviors (e.g., diet, alcohol consumption, smoking, and physical activity), that contribute to accelerated epigenetic aging (Quach et al. 2017; Simons et al. 2022). Despite this possibility, some prior research finds limited support for health behaviors as mediators in associations between stressors and accelerated aging (Simons et al. 2021), although this research did not specifically examine employment histories. Mental health may also link work histories with epigenetic aging. Consistently not working for pay is associated with more depressive symptoms as individuals age (Caputo et al. 2020), and depression is associated with accelerated epigenetic aging (Han et al. 2018). In this study, we focus on depressive symptoms and health behaviors as two key psychosocial pathways linking work history in midlife with accelerated epigenetic aging.

Work Quality and Epigenetic Aging

Although paid work has numerous health benefits, evidence increasingly suggests that poor work quality can be harmful for health. Robust evidence demonstrates that unpredictable and insecure aspects of work, such as job insecurity and unpredictable work hours, are associated with a range of adverse health outcomes (for reviews, see Benach et al. 2014; De Witte, Pienaar, and De Cuyper 2016; Hajat et al. 2024; Kim and von dem Knesebeck 2016). A recent review article by Hajat et al. (2024) emphasizes that precarious work likely contributes to adverse health outcomes via material deprivation, psychosocial stress (e.g., physiological stress, psychological distress), and exposure to workplace hazards.

Similar to work history, chronic exposure to unstable and unpredictable working conditions can be particularly detrimental to health and well-being. For example, workers who experience consistent exposure to job insecurity, insufficient work hours, and low wages in midlife experience more chronic conditions and functional limitations (Donnelly 2022) and increased mortality risk (Kezios et al. 2023) in later life. Exposure to persistent job insecurity also has pernicious effects on psychological distress and self-rated health (Burgard, Brand, and House 2009; Burgard and Seelye 2017; Glavin 2015).

Despite the robust evidence linking poor work quality and health, much remains to be known about how specific aspects of work, especially when experienced over longer periods of time, may be associated with accelerated epigenetic aging. Recent research documents that workers in sales/clerical, service, and lower ranked occupations are more likely to experience accelerated aging relative to workers in managerial and professional occupations (Andrasfay and Crimmins 2023; Andrasfay et al. 2023; George et al. 2021; Schmitz et al. 2021). However, it is not clear why workers in these occupations experience accelerated aging, and evidence on specific characteristics of work is mixed. Indeed, some studies find that long working hours, psychosocial job strain, and physical job demands are associated with accelerated epigenetic aging (Freni-Sterrantino, Fiorito, d’Errico, Robinson, et al. 2022; Freni-Sterrantino, Fiorito, d’Errico, Virtanen, et al. 2022), whereas other studies do not find support for similar characteristics of work (Andrasfay and Crimmins 2023). These studies point to the need for additional research on the impacts of work quality for epigenetic aging, especially with attention to the chronicity of exposures.

In the present study, we focus on a history of exposure to job insecurity, insufficient work hours, and low wages. Job insecurity (both objective and perceived) has well-documented effects on numerous health outcomes (e.g., Benach et al. 2014; Burgard et al. 2009; De Witte et al. 2016). Moreover, insufficient work hours may reflect a combination of low wages, part-time work, and/or a lack of schedule control, all of which are dimensions of precarious work (Kalleberg 2018; Kalleberg, Reskin, and Hudson 2000; Lambert, Fugiel, and Henly 2014) that have adverse effects on health (Donnelly and Schoenbachler 2021; Kalleberg 2018; Kezios et al. 2023). Although long working hours may have adverse effects on outcomes such as mental health and sleep (Bannai and Tamakoshi 2014), they are not inherently indicative of precarious work. In some cases, long hours are characteristics of high-status, high-wage professions, such as surgeons or attorneys, where workers have greater autonomy and job security. As such, we focus on characteristics of work that are pervasive in the contemporary labor market, often perceived as stressful for workers, and potentially associated with accelerated epigenetic aging when experienced throughout midlife.

We also hypothesize that exposure to several characteristics of poor work quality will take a cumulative toll on accelerated epigenetic aging; as such, we consider a summative scale of characteristics of work throughout midlife in addition to the individual characteristics of work. Considering multiple aspects of work quality can capture the multidimensional nature of work, provide information about the accumulation of dimensions of work, and allow scholars to examine the impact of differing levels of adverse work exposures on health outcomes (Hajat et at. 2024).

Stress may be a key explanation connecting job insecurity, insufficient work hours, and low wages with accelerated epigenetic aging. Indeed, a recent study found that workers in low control, precarious jobs had higher levels of stress (perceived and biomarker indicated [i.e., high-sensitivity C-reactive protein]) compared to workers in more secure and stable jobs (Venechuk 2024). Therefore, we focus on two common stress responses as key pathways linking work quality with epigenetic aging in this study: mental health and health behaviors and behavior-related risks. Stress stemming from work conditions, such as job insecurity, may lead to health-harming behaviors (e.g., smoking) that contribute to accelerated epigenetic aging (Klopack et al. 2022). Unstable and insecure work can also undermine mental health (e.g., Benach et al. 2014; Burgard and Seelye 2017; Glavin 2015; Hajat et al. 2024; Macmillan and Shanahan 2022), which increases the risk of accelerated epigenetic aging (Han et al. 2018). Although mental health and health behaviors may be salient pathways linking work quality with epigenetic aging, this possibility has not been tested in prior research. We fill a gap in the literature by examining how a history of poor job quality (i.e., job insecurity, insufficient work hours) in midlife is associated with accelerated aging due, in part, to depressive symptoms and health behaviors.

Present Study

Although prior studies have documented associations between cumulative exposure to midlife work conditions and health outcomes broadly (Donnelly 2022) and have begun to explore links between work characteristics at one point in time and epigenetic aging (Andrasfay and Crimmins 2023), our study makes several novel contributions. First, we leverage longitudinal data to capture cumulative patterns of work history and work quality throughout midlife rather than relying on single-point assessments. Second, we explicitly examine potential pathways (i.e., mental health and health behaviors), drawing on stress process and life course frameworks. Finally, by applying this theoretical framework to epigenetic aging outcomes, we contribute to emerging research connecting social exposures to biological aging processes. We analyze nationally representative longitudinal data from the HRS to test the following hypotheses:

Hypothesis 1a: A history of not working for pay in midlife will be associated with accelerated epigenetic aging.

Hypothesis 1b: Associations between work history in midlife and accelerated epigenetic aging will be attenuated, in part, by mental health and health behaviors.

Hypothesis 2a: A history of poor work quality in midlife will be associated with accelerated epigenetic aging.

Hypothesis 2b: Associations between poor work quality in midlife and accelerated epigenetic aging will be attenuated, in part, by mental health and health behaviors.

Hypothesis 3a: Exposure to multiple characteristics of poor work quality in midlife will be associated with accelerated epigenetic aging.

Hypothesis 3b: Associations between overall poor work quality in midlife and accelerated epigenetic aging will be attenuated, in part, by mental health and health behaviors.

Data And Methods

Data

This study used data from the Health and Retirement Study (HRS; 1992–2016), a nationally representative biennial, longitudinal survey of adults 51+ years in the United States. Measures of epigenetic aging came from the 2016 Venous Blood Study (VBS), which is representative of adults 56+ years. Participants who completed an in-person interview in 2016, were not living in a nursing home, and did not require a proxy were asked to participate in the VBS (N = 9,443). DNAm assays were obtained for a subsample of VBS participants (N = 4,018), from which epigenetic clocks were estimated. Measures related to work, mental health, health behaviors, and covariates come from the core HRS data. We excluded respondents without valid VBS weights of DNAm and epigenetic clocks and respondents from the AHEAD (born < 1924) and CODA (born 1924–1930) cohorts who entered the HRS at older ages and were more likely to be out of the labor force during the study. We further excluded 315 respondents who were retired in every wave during the HRS observation period, resulting in N = 3,000 respondents. Missing data in the work history sample was minimal, with approximately 2% missing on covariates.¹ The most commonly missing variables included weight status (n = 31), drinking status (n = 14), and smoking status (n = 12). However, to maximize the sample size, we used multiple imputation with chained equations to impute missing data. We conducted 10 imputations and include all variables described in the following section in the model.

The analysis of work quality excluded an additional 414 individuals who never worked for pay during the HRS observation period, leading to slightly smaller sample of 2,586 adults. In the analysis of work quality, we again used multiple imputation with chained equations (described previously) to impute missing data on covariates and measures of work quality. Missing data on some work quality measures were more common, with approximately 12% of respondents missing data on perceived job insecurity, 20% missing data on insufficient work hours, and 2% missing data on low wages.

Measures

Epigenetic aging

We examined three measures of accelerated epigenetic aging using clocks based on principal components (PC): PC PhenoAge, PC GrimAge, and DunedinPACE. These clocks predict age-related health outcomes, such as heart disease, diabetes, and cognitive decline (Levine et al. 2018; Lu et al. 2019; McCrory et al. 2021), and changes in the pace of aging across multiple systems (Belsky et al. 2022). PC versions of epigenetic clocks account for unstable probes, reducing noise and improving reliability in measuring biological age (Higgens-Chen et al. 2022). Main results are shown for PC PhenoAge; results for PC GrimAge and DunedinPACE are included in the supplemental tables in the online version of the article and are consistent with PC PhenoAge.

To measure accelerated aging for PC GrimAge and PC PhenoAge, we regressed each epigenetic clock on chronological age and took the residuals. Positive values indicated accelerated aging, and negative values indicated decelerated aging (Levine et al. 2018; Lu et al. 2019). For example, a value of .5 for epigenetic age acceleration would indicate half a year of accelerated aging.

Work history in midlife

We examined the proportion of waves that the respondent was working for pay during the HRS (1992–2014). In each wave, respondents reported whether they were working for pay (1 = not working for pay). We assessed the proportion of waves that respondents were not working for pay during the HRS observation period (hereafter, “not working for pay”) by dividing the number of waves workers were not working for pay by the number of nonmissing waves. With this approach, we created a measure ranging from 0 to 1 (0 = always working for pay, 1 = never worked for pay).

The HRS adds new (younger) cohorts to the study every six years, resulting in different observation periods across cohorts. Therefore, the measure of work history is more likely to capture exit from the labor force via retirement for the older birth cohorts, compromising the comparability of this measure across cohorts in the HRS. As such, we coded waves when respondents reported being retired as missing. The measure of work history, then, assessed the proportion of nonretired, nonmissing waves when respondents were not working for pay in midlife. Results were substantively similar when retired waves are coded as 1 (i.e., not working for pay).

History of poor work quality in midlife

We examined work quality during the HRS using four measures of poor work quality: short job tenure, high job insecurity, insufficient work hours, and low wages. We examined employer tenure as an objective measure of job insecurity and perceived job insecurity as a subjective assessment of job insecurity. The HRS asked respondents how many years they have worked in their current job, and we created a dichotomous indicator of short job tenure by classifying two years or less as short job tenure (the lowest quintile of the distribution), with a job tenure of three or more years as the reference group. For perceived job insecurity, respondents answered the question “What are the chances that you will lose your job during the next year?” by providing a number between 0 (absolutely no chance) and 100 (absolutely certain). We created a dichotomous measure so that a value of 1 indicated high job insecurity (50% chance or greater that they will lose their job; the highest quintile of the distribution) and 0 indicated little job insecurity (less than 50% chance).

We focused on the lowest (job tenure) and highest (job insecurity) quintiles of the distribution for three main reasons. First, focusing on the worst quintile aligns with other cutoffs for “bad” work qualities, such as the cutoff demarcating low-wage work (Kalleberg et al. 2000). Second, the worst quintile is a commonly used threshold in health disparities research (Gill et al. 2022) because the adverse effects may be concentrated among the most socially disadvantaged (e.g., worst quintile of job insecurity). Finally, a dichotomous measure allowed us to examine consistent exposures (i.e., the proportion of waves) rather than potentially less informative measures (e.g., based on average scores). As a supplemental test, we also assessed (1) the proportion of waves with above-average job insecurity/below-average job tenure (i.e., a lower threshold) and (2) average scores of job insecurity/job tenure throughout midlife (i.e., preserving the continuous measures). Associations with accelerated epigenetic aging were weaker with the above/below-average measures and not significant with the average scores, suggesting that the adverse health effects of job insecurity are primarily concentrated among the most insecure workers.

Respondents were also asked whether they would like to increase the number of hours they work each week (with earnings increasing proportionally). An affirmative response to this question indicated that the respondent experienced insufficient work hours and was coded as 1. Finally, respondents reported their hourly wages. In line with prior research on low-wage work (Kalleberg et al. 2000), we classified the lowest quintile of the distribution as low-wage work (coded as 1). To measure the history of each characteristic of work, we assessed the proportion of nonmissing waves that the respondent endorsed the specific characteristic of work (e.g., high job insecurity or low wages) during the HRS (1992–2014). Thus, each measure ranged from 0 to 1, where 0 indicated that the respondent never experienced the characteristic of work throughout the study period and 1 indicated that the respondent experienced the characteristic of work in every nonmissing wave.

To capture the multidimensional nature of work and consider the accumulation of dimensions of work, we summed the four dichotomous indicators to create an index of poor work quality ranging from 0 to 4, where higher values indicated an endorsement of more characteristics of history of poor work quality. Then, to consider work quality history during the HRS study period, we took the average score of the poor work quality index for all waves (1992–2014).

Mental health and health behaviors/behavior-related risks

For the measure of mental health, depressive symptoms were measured with an eight-item Center for Epidemiologic Studies Depression (CES-D) scale (Steffick 2000), where higher scores indicated higher levels of distress (range = 0–8). The CES-D has been used in hundreds of studies in clinical and nonclinical populations to measure depressive symptoms and demonstrates high reliability and strong evidence of construct validity (Steffick 2000). We dichotomized this measure using a cutoff of 4+, which has been validated to have high sensitivity and high specificity to identify respondents who are likely at risk of depression (Steffick 2000).

Behavioral factors included drinking status and smoking status; we also considered weight status as another health behavior-related risk but acknowledge the multiple individual and contextual factors that may influence weight (e.g., genetic predisposition, stress, sleep, green space, etc.). Based on guidelines of alcohol consumption among older adults established by the National Institute of Alcohol Abuse and Alcoholism (Lin, Guerrieri, and Moore 2011), drinking was coded with three categories: nondrinker (reference), light/moderate drinker (one to seven beverages per week), and heavy drinker (eight or more beverages per week). Smoking status included never smoked (reference), former smoker, and current smoker. For weight status, body mass index was assessed by self-reported height and weight, and this continuous measure was collapsed into four categories: underweight, normal weight (reference), overweight, and obese. These measures were assessed in the 2016 interview.

Covariates

We accounted for covariates including age (in years), gender (1 = female), race-ethnicity (non-Hispanic White [reference], non-Hispanic Black, Hispanic, and other racial-ethnic identity), marital status (currently married/partnered [reference], divorced/separated, widowed, never married), and educational attainment (less than high school [reference], high school diploma/GED [General Educational Development], some college, college degree or more). Because respondents who entered the HRS after 1992 had a shorter observation period, we included a variable for HRS birth cohort (HRS [reference], war babies, early baby boomers, mid baby boomers). We also included an indicator of whether the respondent’s health limited the kind or amount of paid work they could do at their baseline interview (1 = yes) to account for some of the selection of healthier people into work.

Analytic Approach

To examine the associations between work history and work quality history and epigenetic age in later life, the present study used ordinary least squares regression to estimate a series of models. Beginning with work history, models first regressed epigenetic accelerated aging (assessed in 2016) on work history (1992–2014), including covariates for age, gender, race-ethnicity, birth cohort, and marital status. In a second model, we added educational attainment and whether health limits work at baseline to account for some of the selection of healthier and higher educated adults into work. Then, models subsequently added symptoms of depression and health behaviors and behavior-related risk separately (Models 3 and 4) before including all measures jointly in a final model (Model 5). We took a similar approach using each indicator of work quality history. Conceptually, Model 2 adjusted for potential confounding, and Models 3 to 5 adjusted for potential downstream pathways.

This analytic strategy used a series of adjusted linear regression models to examine how the association between work history/work quality and accelerated epigenetic aging changed across models. Although this approach is widely applied in observational aging research, it does not formally identify causal direct or indirect effects. As such, estimates may be biased in the presence of time-varying confounding or interactions between the exposure and potential mediators. We also tested the equality of the coefficients across models (Mize, Doan, and Long 2019) and examined gender-stratified results. All analyses include respondent-level weights for DNAm participants.

Results

Descriptive Results

Table 1 presents the descriptive results for the two analytic samples: the whole sample (N = 3,000) and the sample that worked for pay in at least one wave (n = 2,586). As a point of comparison, we also present the descriptive statistics for respondents who did not work for pay during the HRS and were excluded from the work quality analysis (n = 414). When considering work history, Table 1 shows that respondents were not working for pay in a fifth (20%) of their nonmissing, nonretired waves during the HRS study period. Not working for pay includes several labor force statuses, such as unemployed, not working due to a disability, and not in the labor force for other reasons (e.g., homemaker). Table 1 also presents descriptive information on accelerated aging scores and all covariates. The average PC PhenoAge acceleration of −.55 indicates that respondents, on average, had an epigenetic age that was half a year younger than their chronological age (i.e., decelerated epigenetic aging).

Table 1.

Weighted Descriptive Statistics for Analytic Samples (Health and Retirement Study; N = 3,000).

	Whole Sample(N = 3,000)		Did not Work for Pay in HRS(n = 414)		Worked for Pay in at Least 1 Wave(n = 2,586)
	M or %	SD	M or %	SD	M or %	SD
PC PhenoAge score	−.55	7.91	.73	8.92	−.73	7.73
PC GrimAge score	−.23	4.87	.10	5.04	−.27	4.84
PC DunedinPACE score	1.01	.18	1.05	.20	1.01	.18
History of not working for pay (range: 0–1)	.20	.41
History of short job tenure (range: 0–1)					.22	.35
History of high job insecurity (range: 0–1)					.19	.40
History of insufficient work hours (range: 0–1)					.26	.48
History of low-wage work (range: 0–1)					.17	.36
Mean poor work quality index (range: 0–4)					.61	.72
Age	67.18	8.75	67.74	9.70	67.11	8.61
Female	54.15		73.79		51.40
Race-ethnicity
Non-Hispanic White	78.16		62.93		80.29
Non-Hispanic Black	8.87		14.02		5.15
Hispanic	9.27		17.05		8.18
Non-Hispanic other race-ethnicity	3.70		6.00		3.38
Birth cohort
HRS	24.43		27.01		24.07
War babies	21.27		18.14		21.71
Early baby boomers	25.70		25.10		25.78
Mid baby boomers	28.61		29.75		28.45
Marital status
Married/partnered	67.89		56.96		69.45
Separated/divorced	12.09		13.87		11.83
Widowed	12.49		19.26		11.52
Never married	7.53		9.90		7.20
Educational attainment
Less than high school	11.05		27.67		8.71
High school diploma/GED	31.01		33.28		30.70
Some college	27.58		23.21		28.18
College degree+	30.36		15.84		32.40
Health limits work (at baseline)	14.73		49.75		9.82
Symptoms of depression	13.05		29.94		10.68
Smoking status
Never smoker	44.99		47.61		44.64
Former smoker	44.37		35.71		45.59
Current smoker	10.63		16.69		9.77
Drinking status
Nondrinker	55.06		75.16		52.22
Light/moderate drinker	32.38		18.76		34.28
Heavy drinker	12.56		6.08		13.51
Weight status
Underweight	1.14		.83		1.18
Normal weight	23.74		28.14		23.17
Overweight	37.33		26.09		38.85
Obese	37.79		44.93		36.80

Note: PC = principal component; GED = General Educational Development; HRS = Health and Retirement Study.

Characteristics of history of poor work quality in midlife were somewhat common in the sample. Among respondents who worked in at least one wave during the study, workers experienced short job tenure in 22% of their nonmissing, working waves; high job insecurity in 19% of waves; insufficient work hours in 26% of waves; and low-wage work in 17% of waves, on average. However, the average score on a poor work quality index throughout midlife was .61 (range = 0–4) indicating that respondents did not often experience multiple indicators of poor work quality simultaneously and/or consistently throughout this period. Respondents who worked for pay in at least one wave during the HRS study period had an epigenetic age that was over half a year (.73) younger than their chronological age based on PC PhenoAge.

Respondents who did not work for pay during the HRS (excluding respondents who were retired in every wave) were much more likely to experience accelerated epigenetic aging (e.g., .73 years older than chronological age based on PC PhenoAge). These respondents were also more likely to experience a health limitation at baseline, experience symptoms of depression, and be a member of a marginalized or minoritized group.

To provide more information about the measures of work history and work quality during the HRS by birth cohort, Supplement A in the online version of the article presents the descriptive means, and Supplement B in the online version of the article illustrates the distribution of each work measure by birth cohort.

Work History in Midlife and Epigenetic Aging

Table 2 presents the results for epigenetic accelerated aging (PC PhenoAge) regressed on history of not working for pay in midlife. Results for PC Grim Age and DunedinPACE are shown in Supplement C in the online version of the article. Model 1 of Table 2 shows that consistently not working for pay is positively associated with accelerated epigenetic aging such that respondents who never worked for pay in midlife had 1.76 years of accelerated epigenetic age relative to respondents who worked for pay in every year during this period net of sociodemographic covariates (p < .001). Although attenuated by 45% with the addition of educational attainment and baseline health limitations in Model 2, this association remains statistically significant (p < .05). Taken together, Models 1 and 2 provide support for Hypothesis 1a.

Table 2.

Accelerated Aging (PC PhenoAge) Regressed on Work History in Midlife (Health and Retirement Study; N = 3,000).

	Model 1	Model 2	Model 3	Model 4	Model 5
History of not working for pay (range: 0–1)	1.76*** (.44)	.96* (.46)	.82⁺ (.45)	.72(.45)	.59(.44)
Female	−2.15*** (.31)	−2.16*** (.31)	−2.17*** (.31)	−2.22*** (.31)	−2.23*** (.31)
Age	−.05(.04)	−.04(.04)	−.04(.04)	−.03(.04)	−.02(.04)
NH Black (reference: NH White)	3.13*** (.49)	2.88*** (.51)	2.85*** (.50)	2.67*** (.51)	2.64*** (.50)
Hispanic (reference: NH White)	.47(.42)	−.04(.48)	−.08(.48)	−.00(.48)	−.04(.48)
NH other race (reference: NH White)	−.07(.59)	−.18(.57)	−.20(.57)	−.08(.56)	−.11(.56)
War babies cohort (reference: HRS)	−.73(.50)	−.64(.50)	−.64(.50)	−.73(.50)	−.73(.50)
Early baby boomer cohort (reference: HRS)	−1.34* (.60)	−1.07⁺ (.60)	−1.00⁺ (.60)	−1.06⁺ (.58)	−.99⁺ (.58)
Mid baby boomer cohort (reference: HRS)	−1.94* (.77)	−1.69* (.76)	−1.63* (.77)	−1.60* (.74)	−1.55* (.74)
Divorced (reference: married)	.41(.47)	.23(.46)	.11(.46)	−.02(.44)	−.13(.45)
Widowed (reference: married)	.71(.46)	.51(.46)	.49(.46)	.39(.45)	.37(.45)
Never married (reference: married)	.85(.75)	.83(.73)	.72(.74)	(.72)	.57(.73)
High school/GED (reference: LTHS)		−.03(.44)	.06(.44)	.10(.43)	.18(.43)
Some college (reference: LTHS)		−.69(.48)	−.62(.48)	−.38(.47)	−.32(.47)
College+ (reference: LTHS)		−2.04*** (.49)	−1.95*** (.49)	−1.33** (.49)	−1.25* (.49)
Health limits work at baseline		1.23** (.44)	1.11* (.44)	.90* (.44)	.78⁺ (.43)
Symptoms of depression			1.04* (.49)		.96* (.48)
Former smoker (reference: never)				.34(.30)	.35(.30)
Current smoker (reference: never)				1.20** (.46)	1.19** (.46)
Light/moderate drinker (reference: nondrinker)				−1.31*** (.31)	−1.30*** (.31)
Heavy drinker (reference: nondrinker)				−1.30** (.42)	−1.32** (.42)
Underweight (reference: normal weight)				.37(.95)	.37(.94)
Overweight (reference: normal weight)				−.09(.38)	−.08(.38)
Obese (reference: normal weight)				1.91*** (.41)	1.90*** (.41)
Constant	4.00(3.09)	4.41(3.07)	3.89(3.08)	2.87(2.93)	2.41(2.94)

Note: Standard errors are in parentheses. NH = non-Hispanic; HRS = Health and Retirement Study; LTHS = less than high school; GED = General Educational Development.

p < .10, *p < .05, **p < .01, ***p < .001.

Depression and health behaviors/weight status are included in Models 3 and 4. The inclusion of depression in Model 3 attenuates the association between work history and accelerated epigenetic aging by 15%, and the inclusion of health behaviors (i.e., smoking and drinking status) and weight status in Model 4 attenuates the association by 25%. When all variables are jointly included in Model 5, the association between work history and accelerated epigenetic aging is attenuated by 39%. The association between work history and accelerated aging is no longer statistically significant in Models 4 and 5. Postestimation tests of equality indicate that the attenuation with the inclusion of depression is marginally significant and that the attenuation with health behaviors/weight status and all variables is statistically significant (p < .05, p < .01, respectively). Results in Models 3 to 5 of Table 2 provide some support for Hypothesis 1b.

Figure 1 presents the predicted accelerated epigenetic aging scores for respondents who worked for pay consistently, worked for pay in half of their waves, and never worked for pay during the HRS study period. The dark gray bars present predicted scores based on estimates in Model 1 of Table 2 (adjusting for age, gender, race-ethnicity, birth cohort, and marital status) and show stark differences in accelerated epigenetic aging scores across midlife work history. These differences are notably attenuated in fully adjusted models (indicated by the gray bars in Figure 1; based on estimates from Model 5 of Table 2).

Figure 1.

Work History in Midlife and Accelerated Epigenetic Aging (Health and Retirement Study; N = 3,000).

History of Poor Work Quality in Midlife and Epigenetic Aging

Table 3 presents the results for PC PhenoAge for history of poor work quality, and results for PC Grim Age and DunedinPACE are shown in Supplement D in the online version of the article. Model 1 demonstrates that a history of short job tenure (p < .10), high job insecurity (p < .05), insufficient work hours (p < .01), and low wages (p < .01) was positively associated with accelerated epigenetic aging. Respondents who experienced high perceived job insecurity throughout the study period had 1.54 years of accelerated epigenetic age. With the inclusion of educational attainment and baseline health limitations in Model 2, the association is slightly attenuated but is marginally significant for short job tenure and high job insecurity and remains statistically significant for insufficient work hours and low hourly wages. Overall, Models 1 and 2 generally provide support for Hypothesis 2a.

Table 3.

Accelerated Aging (PC PhenoAge) Regressed on Work Quality History in Midlife (Health and Retirement Study; n = 2,586).

	Model 1	Model 2	Model 3	Model 4	Model 5
History of short job tenure in midlife (range: 0–1)	1.06⁺ (.56)	.94⁺ (.56)	.89(.56)	.90(.56)	.86(.56)
History of high job insecurity in midlife (range: 0–1)	1.54* (.62)	1.15⁺ (.61)	1.09⁺ (.62)	.89(.62)	.84(.62)
History of insufficient work hours in midlife (range: 0–1)	1.39** (.46)	1.09* (.46)	1.05* (.46)	.77⁺ (.46)	.75(.46)
History of low hourly wages in midlife (range: 0–1)	1.62** (.53)	1.10* (.54)	1.01⁺ (.54)	.84(.53)	.76(.54)
Poor work quality index in midlife (range: 0–4)	1.32*** (.29)	1.06*** (.30)	1.02*** (.30)	.86** (.29)	.83** (.30)
Controls: age, gender, race-ethnicity, birth cohort, marital status	X	X	X	X	X
Controls: educational attainment, health limits work		X	X	X	X
Controls: symptoms of depression			X		X
Controls: smoking status, drinking status, weight status				X	X

Note: Standard errors are in parentheses. Work quality items measure the proportion of waves during the Health and Retirement Study period that workers experienced each characteristic (range: 0–1). Work quality index is a summative scale of each item averaged across all waves (range: 0–4). Each work measure is included in a separate model. Full tables with covariates are available on request. PC = principal component.

p < .10, *p < .05, **p < .01, ***p < .001.

Associations between poor work quality and accelerated aging were attenuated by 4% (insufficient work hours) to 8% (low hourly wages) with the inclusion of depression, 4% (short job tenure) to 29% (insufficient works hours) with the inclusion of health behaviors/weight status, and 9% (short job tenure) to 31% (insufficient work hours) in the fully adjusted final model. Associations are no longer statistically significant in the final model. Postestimation tests of equality indicate that the attenuation with the inclusion of depression is not statistically significant for any measure of work quality. The attenuation with health behaviors/weight status and all covariates is statistically significant for all measures except short job tenure. Results in Models 3 to 5 of Table 3 provide partial support for Hypothesis 2b; health behaviors and weight status tend to partially explain the association between poor work quality history in midlife and accelerated epigenetic aging.

Figure 2 shows predicted accelerated epigenetic aging scores for respondents who never experienced low hourly wages, experienced low wages in half of their waves, and experienced low wages in every (nonmissing) wave throughout the HRS study period. Estimates come from Model 1 (dark gray bars) and Model 5 (light gray bars) of Table 3. Overall, Figure 2 demonstrates higher predicted accelerated epigenetic aging for respondents who experienced low hourly wages for a longer duration in midlife. These differences were attenuated to nonsignificance in the fully adjusted model (light gray bars).

Figure 2.

Low Wages throughout Midlife and Accelerated Epigenetic Aging (Health and Retirement Study; n = 2,586).

Lastly, we considered an index of poor work quality based on the four characteristics of work (PC PhenoAge in Table 3; PC GrimAge and DunedinPACE in Supplement D in the online version of the article). Overall, poor work quality throughout midlife is positively associated with accelerated epigenetic aging. Each additional endorsement of poor work quality throughout midlife (index range = 0–4) is associated with 1.32 years accelerated epigenetic age (p < .001). This association was attenuated by 20% with the inclusion of educational attainment and health limitations, 4% with the inclusion of depression, 19% with the inclusion of health behaviors/weight status, and 22% in the fully adjusted final model. Postestimation tests of equality indicate that the attenuation with health behaviors/weight status and all variables is statistically significant (p < .001). Notably, the association of poor work quality history throughout midlife with accelerated epigenetic aging remained statistically significant in the final model (p < .01). Overall, results document that endorsing multiple indicators of poor job quality throughout midlife is associated with accelerated epigenetic aging (support for Hypothesis 3a) and that health behaviors/weight status partially explain the association between work quality history in midlife and accelerated epigenetic aging (partial support for Hypothesis 3b).

Sensitivity Test: Gender-Stratified Models

As a sensitivity test, we examine whether results differ for women and for men (Supplements E–G in the online version of the article). Starting with work history (Supplement E in the online version of the article), a history of not working for pay in midlife is associated with accelerated epigenetic aging for women (e.g., PC PhenoAge; coefficient = 2.07; p < .001) and men (e.g., PC PhenoAge; coefficient = 1.82; p < .10). However, the inclusion of all covariates attenuated this focal association somewhat more for men (81%) compared to women (42%); the association remains marginally statistically significant for women.

For work quality history (Supplement F in the online version of the article), results differ somewhat depending on the epigenetic aging measure. For PC PhenoAge, the association between poor work quality in midlife and accelerated epigenetic aging tends to be stronger for women compared to men in base models (Model 1). For PC GrimAge, associations tend to be slightly stronger for men compared to women in base models (Model 1). Results for DunedinPACE are similar for women and men. For all outcomes, results look more similar by gender in models controlling for all covariates (i.e., Model 5).

The history of poor work quality index in midlife results (Supplement G in the online version of the article) differed by epigenetic measure. In Model 1, associations between the history of poor work quality in midlife are stronger for women compared to men for PC PhenoAge, stronger for men compared to women for PC GrimAge, and generally similar across gender for DunedinPACE. In fully adjusted models, the focal associations between the history of poor work quality and accelerated aging are fairly similar for women and for men.

Discussion

Working for pay is beneficial for health and well-being (e.g., Caputo et al. 2020; Frech and Damaske 2012; Ross and Mirowsky 1995), although working in insecure and unstable jobs can erode mental and physical health and increase mortality risk (for reviews, see Benach et al. 2014; De Witte et al. 2016; Hajat et al. 2024; Kim and von dem Knesebeck 2016). Much remains to be known about how work-related exposures, especially repeated exposures, can contribute to accelerated epigenetic aging—a “hallmark of aging” that is highly predictive of the development of age-related conditions, such as cancers, cardiovascular disease, and Alzheimer’s disease, as well as premature mortality (Brunet and Berger 2014; López-Otín et al. 2013). Although recent research points to accelerated aging for workers in specific occupations, such as sales and service occupations (Andrasfay and Crimmins 2023; Andrasfay et al. 2023; George et al. 2021; Schmitz et al. 2021), the present study examines the consequences of labor force participation and job quality throughout midlife for accelerated epigenetic aging in later life. We discuss three key themes from the findings.

The first theme considers the adverse effects of inconsistent employment histories throughout midlife for accelerated epigenetic aging among older adults. We found that consistently not working for pay throughout the HRS study period was associated with accelerated epigenetic age. This finding builds on prior research on the benefits of stable and consistent employment for outcomes such as mental health, functional limitations, and mortality risk (Caputo et al. 2020; Frech and Damaske 2012). We extend this literature by showing that the adverse effects of not working for pay throughout midlife become biologically embodied in ways that will increase the risk of chronic conditions and mortality in later life. Findings from the present study point to the salience of chronic or repeated exposures. Although our measure could capture consistent, chronic exposure to limited labor force participation or intermittent yet repeated exposure to not working for pay (i.e., cumulative exposure), repeated reports over time likely reflect sustained or recurrent experiences of work-related adversity or an economic instability related to work opportunities. Such patterns are conceptually consistent with chronic exposures, which tend to be more consequential for biological aging than isolated or acute events (Simons et al., 2021). Accordingly, this study underscores the importance of examining work histories in midlife to understand how sustained employment conditions become biologically embedded.

The second theme emphasizes the importance of considering the quality of work in addition to labor force participation. Findings from this study indicate that consistent exposure to job insecurity, insufficient work hours, and low wages throughout midlife are associated with accelerated epigenetic aging. We build on prior research and fill a gap in the literature on epigenetic aging by pointing to the importance of the quality of work throughout midlife. Although prior research documents accelerated aging among workers in sales and service occupations (Andrasfay and Crimmins 2023; Andrasfay et al. 2023; George et al. 2021; Schmitz et al. 2021) and some studies find evidence that long working hours, psychosocial job strain, and physical job demands are associated with accelerated epigenetic aging (Freni-Sterrantino, Fiorito, d’Errico, Robinson, et al. 2022; Freni-Sterrantino, Fiorito, d’Errico, Virtanen, et al. 2022), we build on this research by documenting the potential adverse effects of chronic exposure to certain characteristics of work in midlife. That is, the observed health effects of unfavorable work conditions may be evidence of the toll that poor work quality takes on the body, as measured by accelerated epigenetic aging, especially when adverse work experiences unfold over longer periods of time. Moreover, we find that simultaneous exposure to multiple characteristics of poor work quality throughout midlife increases the risk of accelerated epigenetic aging, pointing to the importance of multiple, repeated exposures throughout midlife.

Finally, we considered mental health and health behaviors/weight status as key pathways linking work history and work quality history throughout midlife with accelerated epigenetic aging. Findings indicate that these measures jointly explain 39% of the association between work history and accelerated aging and 9% (for short job tenure) to 31% (for insufficient work hours) of the association between work quality and accelerated epigenetic aging. We hypothesized that consistently not working for pay or consistently experiencing poor work quality would be a source of stress for workers that activated psychological and behavioral responses that, in turn, contributed to accelerated epigenetic aging. Notably, symptoms of depression did not explain much of the associations between work history/work quality and accelerated epigenetic aging. Mental health may not be as salient a pathway as health behaviors and other untested pathways (e.g., relationship strain), or other dimensions of well-being (e.g., loneliness, anxiety) may be more prominent—an avenue for future research. At the same time, health behaviors and weight status attenuated more of the focal associations (except short job tenure). As such, health behaviors may be potentially modifiable risk factors that can be targeted to address work-related health inequities. Although we find initial evidence of possible pathways linking chronic work-related exposures and accelerated epigenetic aging, future research should test formal causal mediation frameworks.

This study provides important new insight into linkages between work-related exposures in midlife and epigenetic aging in later life using longitudinal data that partly reduces bias from mortality selection and reverse causation. However, limitations should be noted. First, because the HRS samples adults over age 50, we lack information about work histories earlier in the life course, which limits our ability to assess the full duration and timing of exposure to labor force participation across the life course. Future research incorporating life course employment histories could help clarify whether the timing and continuity of work-related stressors differentially contribute to epigenetic age acceleration. Second, given the adverse health consequences of unemployment and precarious work, people who are unemployed or exposed to precarious work from a young age have a greater chance of being too sick to participate in the study or may be deceased by midlife. Thus, participation in the study and mortality selection may make our results more conservative. To partly address this concern, we accounted for whether health limited the respondent’s ability to work in their baseline interview. Nevertheless, findings must be interpreted with this consideration in mind. Third, although the HRS asks respondents about job security, work hours, and wages, it lacks information about other dimensions of work that have adverse consequences for health and may contribute to accelerated aging, such as temporary work contracts, the (in)flexibility of work schedules, and involuntary part-time employment. Fourth, the measure of depressive symptoms focuses on symptoms within the past week, which may not fully capture experiences of and fluctuations in mental health throughout midlife. Finally, epigenetic age measures are available at only one point in time (2016), precluding an examination of when epigenetic changes first occur in the life course and changes in epigenetic aging over time in response to work-related exposures.

Overall, the present study documents how work history and work quality history in midlife are important antecedents of accelerated epigenetic aging. Findings shed light on how work-related exposures can “get under the skin” to impact epigenetic aging, especially when those exposures are chronic. These findings are concerning in light of the increase in insecure and unstable work in the United States (e.g., Kalleberg 2018; Lambert et al. 2014). That is, future cohorts of older adults are more likely to be exposed to these health-harming work exposures and to be exposed for a longer period of the life course. Therefore, exposure to inconsistent employment and “bad jobs” will remain a threat for individuals, their families, and society more broadly.

Supplemental Material

sj-docx-1-hsb-10.1177_00221465261432569 – Supplemental material for How “Clocking in” Ages Us: An Examination of Work History, Work Quality, and Accelerated Epigenetic Aging in Older Adulthood

Supplemental material, sj-docx-1-hsb-10.1177_00221465261432569 for How “Clocking in” Ages Us: An Examination of Work History, Work Quality, and Accelerated Epigenetic Aging in Older Adulthood by Mateo P. Farina, Rachel Donnelly and Jessica D. Faul in Journal of Health and Social Behavior

Footnotes

Authors’ Note

Mateo P. Farina and Rachel Donnelly contributed equally.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD042849), the National Institute on Aging (R00AG076964), and the Center on Aging and Population Sciences (P30AG066614), University of Texas at Austin. The Health and Retirement Study is supported by the National Institutes of Aging (U01AG009740). The funders had no role in the design, analysis, interpretation, or decision to publish this study.

ORCID iDs

Mateo P. Farina

Rachel Donnelly

Notes

Supplemental Material

Supplements A to G are available in the online version of the article.

Author Biographies

Mateo P. Farina is an assistant professor in human development and family sciences at The University of Texas at Austin, is affiliated with the Population Research Center, and is associate director of the Center for Aging and Population Sciences. His research examines how social and economic conditions across the life course shape cognitive and biological aging. His work has been published in Demography, The Journals of Gerontology: Series B, and Brain, Behavior, Immunity – Health.

Rachel Donnelly is assistant professor of sociology at Vanderbilt University. Her research primarily focuses on the individual and contextual determinants of mental and physical health across the life course. In recent research supported by the National Institutes of Health, she examines how stressors and state policies jointly shape mental health outcomes, including inequities in mental health. Her research has been published in journals such as Journal of Health and Social Behavior, Social Forces, and Demography.

Jessica D. Faul is a research associate professor at the Survey Research Center in the Institute for Social Research at the University of Michigan and is the current co-director of the Health and Retirement Study. Her research examines socioeconomic predictors of health and health disparities across the life course, integrating biomarker and genetic data to study aging, cognitive health, and biosocial pathways using longitudinal methods. Her work has been published in outlets such as PNAS, The Journals of Gerontology: Series B, and Clinical Epigenetics.

References

Andrasfay

Theresa

Crimmins

Eileen

. 2023. “Occupational Characteristics and Epigenetic Aging among Older Adults in the United States.” Epigenetics 18(1):2218763. doi:10.1080/15592294.2023.2218763.

Andrasfay

Theresa

Kim

Jung Ki

Ailshire

Jennifer A.

Crimmins

Eileen

. 2023. “Aging on the Job? The Association between Occupational Characteristics and Accelerated Biological Aging.” The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences 78(7):1236–45.

Bannai

Akira

Tamakoshi

Akiko

. 2014. “The Association between Long Working Hours and Heath: A Systematic Review of Epidemiological Evidence.” Scandinavian Journal of Work, Environment, & Health 40(1):5–18.

Belsky

Daniel W.

Caspi

Avshalom

Corcoran

David L.

Sugden

Karen

Poulton

Richie

Arseneault

Louise

Baccarelli

Andrea

, et al. 2020. “Quantification of the Pace of Biological Aging in Humans through a Blood Test, the DunedinPoAm DNA Methylation Algorithm.” eLife 9:e54870. doi:10.7554/eLife.54870.

Belsky

Daniel W.

Caspi

Avshalom

Corcoran

David L.

Sugden

Karen

Poulton

Richie

Arseneault

Louise

Baccarelli

Andrea

, et al. 2022. “DunedinPACE, a DNA Methylation Biomarker of the Pace of Aging.” elife 11:e73420. doi:10.7554/eLife.73420.

Benach

Joan

Vives

Alejandra

Amable

Marcelo

Vanroelen

Christophe

Tarafa

Gemma

Muntaner

Carles

. 2014. “Precarious Employment: Understanding an Emerging Social Determinant of Health.” Annual Review of Public Health 35:229–53.

Brand

Jennie E.

2015. “The Far-Reaching Impact of Job Loss and Unemployment.” Annual Review of Sociology 41:359–75.

Brunet

Anne

Berger

Shelley L.

2014. “Epigenetics of Aging and Aging-Related Disease.” The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences 69(1):S17–20.

Burgard

Sarah A.

Brand

Jennie E.

House

James S.

2009. “Perceived Job Insecurity and Worker Health in the United States.” Social Science & Medicine 69(5):777–85.

10.

Burgard

Sarah A.

Lin

Katherine Y.

2013. “Bad Jobs, Bad Health? How Work and Working Conditions Contribute to Health Disparities.” American Behavioral Scientist 57(8):1105–27.

11.

Burgard

Sarah A.

Seelye

Sarah

. 2017. “Histories of Perceived Job Insecurity and Psychological Distress among Older U.S. Adults.” Society and Mental Health 7(1):21–35.

12.

Caputo

Jennifer

Pavalko

Eliza K.

Hardy

Melissa A.

2020. “Midlife Work and Women’s Long-Term Health and Mortality.” Demography 57(1):373–402.

13.

Chen

Brian H.

Marioni

Riccardo E.

Colicino

Elena

Peters

Marjolein J.

Ward-Caviness

Cavin K.

Tsai

Pei-Chien

Roetker

Nicholas S.

, et al. 2016. “DNA Methylation-Based Measures of Biological Age: Meta-analysis Predicting Time to Death.” Aging 8(9):1844–59.

14.

De Witte

Hans

Pienaar

Jaco

De Cuyper

Nele

. 2016. “Review of 30 Years of Longitudinal Studies on the Association between Job Insecurity and Health and Well-Being: Is There Causal Evidence?” Australian Psychologist 51(1):18–31.

15.

Donnelly

Rachel

. 2022. “Precarious Work in Midlife: Long-Term Implications for the Health and Mortality of Women and Men.” Journal of Health and Social Behavior 63(1):142–58.

16.

Donnelly

Rachel

Schoenbachler

Adam K.

2021. “Part-Time Work and Health in the United States: The Role of State Policies.” SSM - Population Health 15:100891. doi:10.1016/j.ssmph.2021.100891.

17.

Faul

Jessica D.

Kim

Jung Ki

Levine

Morgan E.

Thyagarajan

Bharat

Weir

David R.

Crimmins

Eileen M.

2023. “Epigenetic-Based Age Aceleration in a Representative Sample of Older Americans: Associations with Aging-Related Morbidity and Mortality.” Proceedings of the National Academies of Science 120(9):e2215840120. doi:10.1073/pnas.2215840120.

18.

Frech

Adrianne

Damaske

Sarah

. 2012. “The Relationships between Mothers’ Work Pathways and Physical and Mental Health.” Journal of Health and Social Behavior 53(4):396–412.

19.

Freni-Sterrantino

Anna

Fiorito

Giovanni

d’Errico

Angelo

Robinson

Oliver

Virtanen

Marianna

Ala-Mursula

Leena

Järvelin

Marjo-Riitta

Ronkainen

Justiina

Vineis

Paolo

. 2022. “Work-Related Stress and Well-Being in Association with Epigenetic Age Acceleration: A Northern Finland Birth Cohort 1966 Study.” Aging 14(3):1128–56.

20.

Freni-Sterrantino

Anna

Fiorito

Giovanni

d’Errico

Angelo

Virtanen

Marianna

Ala-Mursula

Leena

Järvelin

Marjo-Riitta

Vineis

Paolo

Robinson

Oliver

. 2022. “Association between Work Characteristics and Epigenetic Age Acceleration: Cross-Sectional Results from U.K.—Understanding Society Study.” Aging 14(19):7752–73.

21.

George

Anitha

Hardy

Rebecca

Fernandez

Juan Castillo

Kelly

Yvonne

Maddock

Jane

. 2021. “Life Course Socioeconomic Position and DNA Methylation Age Acceleration in Mid-life.” Journal of Epidemiology and Community Health 75(11):1084–90.

22.

Gill

Thomas M.

Wyk

Bren Vander

Leo-Summers

Linda

Murphy

Terrence E.

Becher

Robert D.

2022. “Population-Based Estimates of 1-Year Mortality after Major Surgery among Community-Living Older U.S. Adults.” JAMA Surgery 157(12):e225155. doi:10.1001/jamasurg.2022.5155.

23.

Glavin

Paul

. 2015. “Perceived Job Insecurity and Health: Do Duration and Timing Matter?” The Sociological Quarterly 56(2):300–328.

24.

Hajat

Anjum

Andrea

Sarah B.

Oddo

Vanessa M.

Winkler

Megan R.

Ahonen

Emily Q.

2024. “Ramifications of Precarious Employment for Health and Health Inequity: Emerging Trends from the Americas.” Annual Review of Public Health 45:235–51.

25.

Han

Laura K. M.

Aghajani

Moji

Clark

Shaunna L.

Chan

Robin F.

Hattab

Mohammad W.

Shabalin

Andrey A.

Zhao

Min

, et al. 2018. “Epigenetic Aging in Major Depressive Disorder.” American Journal of Psychiatry 175(8):774–82.

26.

Higgens-Chen

Albert T.

Thrush

Kyra L.

Wang

Yunzhang

Minteer

Christopher J.

Kuo

Pei-Lun

Wang

Meng

Niimi

Peter

, et al. 2022. “A Computational Solution for Bolstering Reliability of Epigenetic Clocks: Implications for Clinical Trials and Longitudinal Tracking.” Nature Aging 2:644–61.

27.

Hillary

Robert

Stevenson

Anna

McCartney

Daniel

Campbell

Archie

Walker

Rosie M.

Howard

David M.

Ritchie

Craig W.

, et al. 2020. “Epigenetic Measures of Ageing Predict the Prevalence and Incidence of Leading Causes of Death and Disease Burden.” Clinical Epigenetics 12(1):115. doi:10.1186/s13148-020-00905-6.

28.

Horvath

Steve

. 2013. “DNA Methylation Age of Human Tissues and Cell Types.” Genome Biology 14(10):R115. doi:10.1186/gb-2013-14-10-r115.

29.

Kalleberg

Arne L.

2018. Precarious Lives: Job Insecurity and Well-Being in Rich Democracies. London, UK: Polity Press.

30.

Kalleberg

Arne L.

Reskin

Barbara F.

Hudson

Ken

. 2000. “Bad Jobs in America: Standard and Nonstandard Employment Relations and Job Quality in the United States.” American Sociological Review 65(2):256–78.

31.

Kezios

Katrina L.

Peiyi

Calonico

Sebastian

Hazzouri

Adina Zeki Al

. 2023. “History of Low Hourly Wage and All-Cause Mortality among Middle-Aged Workers.” JAMA 329(7):561–73.

32.

Kim

Tae Jun

von dem Knesebeck

Olaf

. 2016. “Perceived Job Insecurity, Unemployment and Depressive Symptoms: A Systematic Review and Meta-analysis of Prospective Observational Studies.” International Archives of Occupational and Environmental Health 89(4):561–73.

33.

Klopack

Eric T.

Carroll

Judith E.

Cole

Steve W.

Seeman

Teresa E.

Crimmins

Eileen M.

2022. “Lifetime Exposure to Smoking, Epigenetic Aging, and Morbidity and Mortality in Older Adults.” Clinical Epigenetics 14:72. doi:10.1186/s13148-022-01286-8.

34.

Klumb

Petra L.

Lampert

Thomas

. 2004. “Women, Work, and Well-Being 1950–2000: A Review and Methodological Critique.” Social Science & Medicine 58(6):1007–24.

35.

Lambert

Susan J.

Fugiel

Peter J.

Henly

Julia R.

2014. Precarious Work Schedules among Early-Career Employees in the United States: A National Snapshot. Chicago, IL: University of Chicago Employment Instability, Family Well-being, and Social Policy Network. https://voices.uchicago.edu/einet/files/2018/05/lambert.fugiel.henly_.precarious_work_schedules.august2014_0-298fz5i.pdf.

36.

Levine

Morgan E.

Ake T.

Quach

Austin

Chen

Brian H.

Assimes

Themistocles L.

Bandinelli

Stefania

Hou

Lifang

, et al. 2018. “An Epigenetic Biomarker of Aging for Lifespan and Healthspan.” Aging 10(4):573–91.

37.

Lin

James C.

Guerrieri

Joy Gioia

Moore

Alison A.

2011. “Drinking Patterns and the Development of Functional Limitations in Older Adults: Longitudinal Analyses of the Health and Retirement Survey.” Journal of Aging and Health 23(5):806–21.

38.

López-Otín

Maria A. Blasco

Partridge

Linda

Serrano

Manuel

Kroemer

Guido

. 2013. “The Hallmarks of Aging.” Cell 153(6):1194–217.

39.

Ake T.

Quach

Austin

Wilson

James G.

Reiner

Alex P.

Aviv

Abraham

Raj

Kenneth

Hou

Lifang

, et al. 2019. “DNA Methylation GrimAge Strongly Predicts Lifespan and Healthspan.” Aging 11(2):303–27.

40.

Macmillan

Ross

Shanahan

Michael J.

2022. “Explaining the Occupational Structure of Depressive Symptoms: Precarious Work and Social Marginality across European Countries.” Journal of Health and Social Behavior 63(3):446–69.

41.

Marioni

Riccardo E.

Shah

Sonia

McRae

Alan F.

Chen

Brian H.

Colicino

Elena

Harris

Sarah E.

Gibson

Jude

, et al. 2015. “DNA Methylation Age of Blood Predicts All-Cause Mortality in Later Life.” Genome Biology 16:25. doi:10.1186/s13059-015-0584-6.

42.

McCrory

Cathal

Fiorito

Giovanni

Hernandez

Belinda

Polidoro

Silvia

O’Halloran

Aisling M.

Hever

Ann

Cheallaigh

Cliona Ni

, et al. 2021. “GrimAge Outperforms Other Epigenetic Clocks in the Prediction of Age-Related Clinical Phenotypes and All-Cause Mortality.” The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences 76(5):741–49.

43.

Mize

Trenton D.

Doan

Long

J. Scott

. 2019. “A General Framework for Comparing Predictions and Marginal Effects across Models.” Sociological Methodology 49(1):152–89.

44.

Moffitt

Terrie E.

2020. “Behavioral and Social Research to Accelerate the Geroscience Translation Agenda.” Ageing Research Reviews 63:101146. doi:10.1016/j.arr.2020.101146.

45.

Pearlin

Leonard I.

Menaghan

Elizabeth G.

Lieberman

Morton A.

Mullan

Joseph T.

1981. “The Stress Process.” Journal of Health and Social Behavior 22(4):337–56.

46.

Pearlin

Leonard I.

Schieman

Scott

Fazio

Elena M.

Meersman

Stephen C.

2005. “Stress, Health, and the Life Course: Some Conceptual Perspectives.” Journal of Health and Social Behavior 46(2):205–19.

47.

Quach

Austin

Levine

Morgan E.

Tanaka

Toshiko

Ake T.

Chen

Brian H.

Ferrucci

Luigi

Ritz

Beate

, et al. 2017. “Epigenetic Clock Analysis of Diet, Exercise, Education, and Lifestyle Factors.” Aging 9(2):419–46.

48.

Ross

Catherine E.

Mirowsky

John

. 1995. “Does Employment Affect Health? ” Journal of Health and Social Behavior 36(3):230–43.

49.

Schmitz

Lauren L.

Zhao

Wei

Ratliff

Scott M.

Goodwin

Julia

Miao

Jiacheng

Qiongshi

Guo

Xiuqing

, et al. 2021. “The Socioeconomic Gradient in Epigenetic Ageing Clocks: Evidence from the Multi-ethnic Study of Atherosclerosis and the Health and Retirement Study.” Epigenetics 17(6):589–611.

50.

Simons

Ronald L.

Lei

Man-Kit

Beach

Steven S. R.

Philibert

Robert

Cutrona

Carolyn

Gibbons

Frederick

Barr

Ashley

. 2016. “Economic Hardship and Biological Weathering: The Epigenetics of Aging in a U.S. Sample of Black Women.” Social Science & Medicine 150:192–200.

51.

Simons

Ronald L.

Lei

Man-Kit

Klopack

Eric

Beach

Steven R. H.

Gibbons

Frederick X.

Philibert

Robert A.

2020. “The Effects of Social Adversity, Discrimination, and Health Risk Behaviors on the Accelerated Aging of African Americans: Further Support for the Weathering Hypothesis.” Social Science & Medicine 282:113169. doi:10.1016/j.socscimed.2020.113169.

52.

Simons

Ronald L.

Lei

Man-Kit

Klopach

Eric

Berg

Mark

Zhang

Yue

Beach

Steven S. R.

2021. “(Re)Setting Epigenetic Clocks: An Important Avenue whereby Social Conditions Become Biologically Embedded across the Life Course.” Journal of Health and Social Behavior 62(3):436–53.

53.

Simons

Ronald L.

Ong

Mei Ling

Lei

Man-Kit

Klopach

Eric

Berg

Mark

Zhang

Yue

Philibert

Robert

Gibbons

Frederick X.

Beach

Steven R. H.

2022. “Shifts in Lifestyle and Socioeconomic Circumstances Predict Change—for Better or Worse—in Speed of Epigenetic Aging: A Study of Middle-Aged Black Women.” Social Science & Medicine 307:115175. doi:10.1016/j.socscimed.2022.1151.

54.

Steffick

Diane E.

2000. Documentation of Affective Functioning Measures in the Health and Retirement Study. Ann Arbor: University of Michigan. https://hrs.isr.umich.edu/publications/biblio/5411.

55.

Venechuk

Grace

. 2024. “Peeking under the Hood of Job Stress: How Men and Women’s Stress Levels Vary by Typologies of Job Quality and Family Composition.” Journal of Health and Social Behavior 65(2):200–220.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.47 MB