Associations Between DNA Methylation Age Acceleration,Depressive Symptoms,and Cardiometabolic Traits in African American Mothers From the InterGEN Study

Abstract

Background:

African American women (AAW) have a high risk of both cardiometabolic (CM) illness and depressive symptoms. Depressive symptoms co-occur in individuals with CM illness at higher rates than the general population, and accelerated aging may explain this. In this secondary analysis, we examined associations between age acceleration; depressive symptoms; and CM traits (hypertension, diabetes mellitus [DM], and obesity) in a cohort of AAW.

Methods:

Genomic and clinical data from the InterGEN cohort (n = 227) were used. Age acceleration was based on the Horvath method of DNA methylation (DNAm) age estimation. Accordingly, DNAm age acceleration (DNAm AA) was defined as the residuals from a linear regression of DNAm age on chronological age. Spearman’s correlations, linear and logistic regression examined associations between DNAm AA, depressive symptoms, and CM traits.

Results:

DNAm AA did not associate with total depressive symptom scores. DNAm AA correlated with specific symptoms including self-disgust/self-hate (−0.13, 95% CI −0.26, −0.01); difficulty with making decisions (−0.15, 95% CI −0.28, −0.02); and worry over physical health (0.15, 95% CI 0.02, 0.28), but were not statistically significant after multiple comparison correction. DNAm AA associated with obesity (0.08, 95% CI 1.02, 1.16), hypertension (0.08, 95% CI 1.01, 1.17), and DM (0.20, 95% CI 1.09, 1.40), after adjustment for potential confounders.

Conclusions:

Associations between age acceleration and depressive symptoms may be highly nuanced and dependent on study design contexts. Factors other than age acceleration may explain the connection between depressive symptoms and CM traits. AAW with CM traits may be at increased risk of accelerated aging.

Keywords

Depressive symptoms DNA methylation African American women accelerated aging chronic medical conditions

Background

Cardiometabolic (CM) illnesses and depressive symptoms are increasingly common problems, each with profound personal and public health consequences.^1-4 African American women (AAW) bear a disproportionate CM burden, with higher rates of diabetes mellitus (DM), hypertension, and obesity—presenting at earlier ages—than other groups in the USA.^5,6 AAW also experience a greater proportion of depression that is considered chronic or severe compared to other racial/ethnic groups.^7-9 However, AAW are often poorly represented in studies aimed to elucidate biological correlates in CM illness and depressive symptoms.

Depressive presentations are heterogeneous and include an array of symptoms spanning affective, cognitive, and somatic domains. This heterogeneity combined with a lack of diagnostic biomarkers remains a barrier to the development of precision health strategies. Consequently, the National Institute of Mental Health’s (NIMH) Research Domain Criteria (RDoC) has shifted focus away from the sole use of diagnostic labels, encouraging examination of biomarkers in relation to continuous symptoms or domains.¹⁰ Depressive symptoms include depressed mood, loss of interest in previously enjoyed activities; disruption in sleep patterns; changes in appetite; loss of energy; poor concentration and recall; overwhelming feelings of guilt, hopelessness, or worthlessness; psychomotor slowing; and thoughts of death or suicide.¹¹ These symptoms not only diminish quality of life but are also associated with increased all-cause mortality, suggesting a quantitative impact on longevity.^12,13 Specific symptoms or symptom profiles (eg, atypical symptoms including hypersomnia, hyperphagia, interpersonal rejection sensitivity) may associate more or less strongly with CM or aging indicators than classic melancholic presentations,¹⁴ highlighting the strength of the RDoC approaches.

Adults with CM illness, when compared to those with other chronic conditions, are disproportionately affected by depressive symptoms—occurring in 55% of hypertension cases and 35% of DM cases.¹³ Further, 44% of adults with depression meet the criteria for obesity (ie, body mass index [BMI] ⩾30).¹⁵ In AAW specifically, depressive symptoms have been associated with increased blood pressure.¹⁶ Links between CM traits and depressive symptoms are characterized as bidirectional and mutually reinforcing through behavioral, biological, and genomic factors, but these connections are not well understood or sufficiently explored in AAW—a population most likely to be negatively affected.^15,17,18

A plausible and testable explanation for the cross-over between CM conditions—often associated with aging—and depressive symptoms is aging itself, or more specifically, accelerated aging.^19-21 Accelerated aging refers to being biologically older than expected for years lived,^19,22 and depressive symptoms are associated with several estimates of accelerated aging, including inflammatory cytokines, products of oxidative stress, and telomere length.^20,21,23 DNA methylation (DNAm) age is a reliable and parsimonious estimate of biological age based on patterns of DNAm and is highly predictive of all-cause mortality and independent of other aforementioned measures.^24,25 DNAm age is calculated and then contrasted with chronological age to determine the degree of DNAm age acceleration (DNAm AA).²⁵ Many DNAm age “clocks” (ie, algorithms) now exist, each with distinctions in tissue selection, training data populations, and DNAm sites selected.²⁶

Methodologies to calculate DNAm age are relatively new, and their application to exploring correlates of depressive symptoms and CM traits is nascent. Positive associations between DNAm AA and depression have been observed, but in these studies, AAW are not well represented and included CM indices are limited.^27,28 In a study of over 4000 adults, individuals with depression exhibited significantly greater DNAm AA than age-matched controls, an association partly mediated by BMI.²⁸ In another large case-control study, Han et al²⁷ demonstrated significant DNAm AA in individuals with depression and replicated this association in postmortem brain tissue. Pathway analysis of postmortem brain tissue, from a separate replication sample in this study, further demonstrated overlap of depression and several DNAm age CpG sites linked to neurogenesis, neuron differentiation, and regulation of neuron death, indicating several pathways relevant to depression were enriched in this DNAm indicator of cell aging.²⁷

Premature aging may be a plausible common factor in depressive symptoms and other stress-related and aging-related conditions including CM illness. DNAm AA has been shown to associate with hypertension, cardiovascular disease,^29,30 BMI,³¹ and blood glucose,³² but not consistently across diverse populations. This variability warrants further study, especially because links between DNAm AA and lifetime cumulative stress/early exposure to racial discrimination exist among African Americans,^33,34 positioning DNAm AA as a possible connection between stressors and clinical outcomes.

Some studies have demonstrated associations between DNAm AA and specific depressive symptoms—as opposed to composite scores or diagnostic labels alone.^35-37 Notably, in one of these studies, DNAm AA did not associate with total depressive symptom scores but did positively correlate with reduced positive affect,³⁵ underscoring the importance of considering symptoms rather than diagnostic labels or clinical cut-offs which yield heterogeneous cases.

Taken together, these findings suggest that examination of depressive symptoms and DNAm AA is enhanced when related factors such as specific symptoms and CM indices are considered. Such examinations may help to illuminate pathophysiological connections between highly comorbid presentations (ie, CM traits and depression) and result in a more nuanced description of depressive phenotypes to distinguish which depressive features in what context correlate to biological aging. To our knowledge, no studies to date have used DNAm age to examine associations between premature aging, depressive symptoms, and CM traits specifically in AAW. Here, we aim to address this knowledge gap by examining potential associations between DNAm AA, depressive symptoms, and CM traits in a population that bears a disproportionate risk. Specifically, we hypothesize that greater DNAm AA will associate with higher depressive symptom scores generally and especially in those with CM traits.

Methods

Study design and participants

The Intergenerational Impact of Genetic and Psychological Factors on Blood Pressure (InterGEN) study is a multidisciplinary, longitudinal investigation of genomic, environmental, and psychosocial factors influencing blood pressure among African American mother-child dyads.^38,39 The 250 dyads were recruited from early care and education (ECE) centers in Southwest and Central Connecticut, and data was collected from 2015 to 2020.^38,39 Eligible participants included women who were (i) English-speaking, (ii) ⩾21 years old, (iii) African American or Black (self-identified), (iv) mothers to their biological children (ages between 3 and 5 years), and (v) without evidence of psychiatric or cognitive impairment that could interfere with reliable reporting of information as determined by Mini-Mental Status Examination (MMSE).³⁹ Yale University IRB (approval #1311012986) approved procedures for this study.

Comprehensive study procedures have been previously described.^38,39 Briefly, trained InterGEN staff measured blood pressure, height, and weight following national guidelines and collected DNA samples (saliva) using Oragene Format tubes.³⁸ Audio self-assisted interviewing software was used to collect sociodemographic, health history, and psychosocial data.³⁹ Researchers collected clinical and psychosocial data every 6 months over 2 years. The present analysis includes data collected at the first visit, as DNAm data was only collected at the first time point.

Instruments and measures

Beck depression inventory

The Beck et al Depression Inventory (BDI-I) measured depressive symptoms spanning affective, cognitive, and somatic domains including sadness, guilt, psychomotor agitation, sleep disturbance, and appetite loss. The tool is highly reliable and valid for assessing both intensity of depressive symptoms and identifying undiagnosed persons who would meet criteria for a depressive disorder.⁴⁰ The BDI consists of 21 items that are individually scored from 0 to 3 according to severity. The sum of item scores represents the overall depressive symptom severity which ranges from normal ups and downs (0-10) and mild and moderate symptoms (11-30) to severe and extreme depression (over 30).⁴⁰ The BDI demonstrates high internal consistency in low-income African American outpatients and is well-validated in populations with chronic medical conditions.^40,41

Cardiometabolic traits

Health history and physiologic measures gauged the presence of CM traits. BMI (weight in kg/height in m²) was examined continuously and categorically according to the standards set by the World Health Organization. Accordingly, obesity was defined as BMI ⩾ 30 kg/m². Prior diagnoses (self-reported) of hypertension and DM were assessed via a health history questionnaire.³⁹

Sociodemographic

Sociodemographic variables included age, race, ethnicity, income, education, and employment. Although all participants identified as Black/AA, additional racial identities were assessed. The parent study categorized race as American Indian, Asian, Black or African American, Pacific Islander, White, and Other (participants select all that apply) and classified ethnicity as Hispanic (yes/no). Annual household income levels were categorized as less than $5000, $5000 to $9999, $10 000 to $14 9999, . . . through $100 000, and above. Highest level of education ordinal responses included less than high school, high school diploma or GED, some college, Associate’s degree, Bachelor’s degree, Master’s degree, and Doctorate. Employment (yes/no) was defined as any paid employment in the past 12 months including self-employment.

DNA methylation data and assessment of accelerated aging

Calculation of DNAm age in the cohort has previously been described.³¹

Briefly, the Illumina Infinium Methylation EPIC (850 K) BeadChip was used to perform the DNAm at over 850 000 CpG dinucleotides, providing near-exhaustive genomic coverage.³⁹ Confirmation of DNAm was achieved by methylation-specific PRC and bisulfite sequencing.³⁹ Fluorescent signals from methylated and unmethylated probes indicated methylation level at each CpG site (ß = max [M,0/|U| + |M| + 100]).^31,38 Sites with missing genotype cell rates over 10% were excluded using a detection P-value threshold (>.001) for each CpG site.³¹

DNAm age for each participant was calculated using the Horvath²⁵ multi-tissue epigenetic clock and DNAm data from saliva specimens.³¹ The Horvath clock was selected over other DNAm age clocks because of its reliability across tissue and cell types (including saliva), and up until the time of this analysis, other clocks were based on DNAm in blood.²⁶ It is worth noting that other blood-based clocks such as PhenoAge and GrimAge have performed well even in other tissues^42,43; however, the phenotypic characteristics and plasma proteins required for the calculation of these clocks were not available in this dataset. Further, as this was a secondary analysis of existing data generated from DNAm data, rather than the reanalysis of raw DNAm data, determination of biological age by other clocks was not feasible.

The Horvath epigenetic clock is based upon an elastic net regression analysis from over 8000 samples from 82 publicly available Illumina datasets. ß values from 353 CpG sites specified by the Horvath model were plugged into the following formula to calculate DNAm age.

D N A m A g e = i n v e r s e . F (b_{0} + b_{1} {C p G}_{1} + \dots . + b_{353} {C p G}_{353})

DNAm AA, an indicator of more rapid aging, was defined as the residual term when DNAm age is regressed on chronological age,^25,31 a distinction from ∆ Age (ie, the difference between DNAm age and chronological age by subtraction).

Statistical analysis

Power analysis estimated that 191 participants were needed to detect a minimal effect size (correlation coefficient = .2) at 80% power and an alpha level of .05. Effect size was estimated from prior studies that examined associations between DNAm AA and depressive symptoms showing an effect size range between 0.18 and 0.27.^27,35,44 The final sample included 227 African American mothers, exceeding the 191 participants necessary for 80% power. Individuals lacking depressive symptom data or saliva samples insufficient to calculate DNAm age (n = 23) were not eligible for inclusion in the present analysis.

Normality assumptions were tested with histograms, scatterplots, and Shapiro-Wilks testing. Since the questionnaires have a risk for social desirability bias, patterns of missingness (completely missing at random, missing at random, and not missing at random) were assessed.

Less than 10 observations were missing across sociodemographic and BDI items, and these participants were excluded from the corresponding analysis. Missing data from these questionnaires revealed a non-monotone pattern of missingness and no evidence of non-ignorable missingness as there were no significant associations between cases with missing data and the values of the other observed variables.

Spearman’s correlations and linear regression examined associations between DNAm AA and BDI scores in the total sample and subsamples with versus without one or more CM traits. Logistic regression examined associations between DNAm AA and categorical variables including CM traits. Variables that demonstrated significant association (ie, CM traits) with DNAm AA in unadjusted models were assessed again in models adjusted for smoking, income, education, marital status, ethnicity, and depression as covariates.^45,46 Wilcoxon signed-rank tests examined differences in BDI scores between participants with or without a CM trait and those with positive versus negative age acceleration. Chi-squared tests examined associations between age acceleration (binary) and categorical sociodemographic and clinical variables (eg, marital status, smoking). All analyses for this study were performed in R studio with R version 4.0.0. Descriptive and inferential statistics were performed utilizing [psych] and [hmisc] packages, and missingness analysis was conducted using the R packages [naniar], [finalfit], [ggplot], and [dplyr].^47-52

Results

Descriptive statistics

The sample for this study included 227 African American women with complete data for DNAm age and recorded depressive symptom data. Chronological age was correlated strongly with DNAm age (r = .918, P < .001) (Figure 1). Table 1 presents sample characteristics and DNAm AA results generated from regression analysis of DNAm age on chronological age. Overall, this sample demonstrated a younger median DNAm age (28.13 years) than chronological age (31.23). Over half the sample (54.19%) had at least 1 CM trait including hypertension (20.96%), DM (6.17%), or obesity as defined as a BMI of 30 or above (44.93%). Depressive symptom scores were heavily right-skewed (4, [IQR 1-11]), and 72.25% scored within a range indicative of normal ups and downs. A total of 52 participants identified as current smokers (22.91%).

Figure 1.

Linear regression calculating DNA methylation age acceleration. Significant correlation between chronological age and DNAm age in 227 African American women from the InterGEN study. The blue line is the regression line and the surrounding gray area indicates confidence interval. Points above or below the regression line indicate positive age acceleration (ie, individuals biologically older than chronological age) and age deceleration (ie, individuals biologically younger than chronological age) respectively. DNAm age acceleration refers to the residuals between each observation and the regression line.

Table 1.

Summary statistics of sample characteristics.

Characteristics	Median/N	(IQR)/(%)
Chronological age (years)	31.23	(27.32, 35.62)
DNA methylation age (years)	28.81	(24.36, 32.47)
∆ Age (years)	−3.19	(−5.93, −0.43)
DNAm AA (years)	−0.12	(−2.98, 2.58)
Age acceleration (binary)
No	175	(77.09%)
Yes	52	(22.91%)
Hispanic/Latina
No	208	(91.63%)
Yes	19	(8.37%)
Multiracial
No	221	(97.36%)
Yes	16	(7.05%)
Marital status
Married	54	(23.79%)
Single	149	(65.64%)
Other	24	(10.57%)
Education
Less than high school	13	(5.73%)
High school/GED	78	(34.36%)
Some college	76	(33.48%)
Associate’s degree	26	(11.45%)
Bachelor’s degree	25	(11.01%)
Master’s degree	7	(3.08%)
Doctorate	2	(0.88%)
Annual income
<$5000	51	(22.47%)
$5000-$9999	30	(13.22%)
$10 000-$14 999	26	(11.45%)
$15 000-$19 999	18	(7.93%)
$20 000-$24 999	22	(9.69%)
$25 000-$34 999	30	(13.22%)
$35 000-$49 999	27	(11.89%)
$50 000-$74 999	12	(5.29%)
$75 000-$99 999	7	(3.08%)
$100 000 or higher	4	(1.76%)
Employed
No	70	(30.84%)
Yes	155	(68.28%)
Depression score (total)	4	(1, 11)
Affective subscore	0	(0, 2.25)
Cognitive subscore	1	(0, 3)
Somatic subscore	2.5	(0, 5)
Possible MDD
No: BDI < 10	164	(72.25%)
Yes: BDI ⩾ 10	63	(27.75%)
Ever diagnosed HTN
No	181	(79.74%)
Yes	46	(20.26%)
Systolic BP, mmHg	113.33	(105.67, 120.00)
Diastolic BP, mmHg	72	(64.67, 79.33)
Ever diagnosed DM
No	213	(93.83%)
Yes	14	(6.17%)
BMI, kg/m²	28.67	(23.53, 34.21)
Obese (BMI ⩾ 30)
No	125	55.07%
Yes	102	(44.93%)
Any CM trait
No	104	(45.81%)
Yes	123	(54.19%)
Smoking (current)
No	175	(77.09%)
Yes	52	(22.91%)

Abbreviations: BMI, body mass index; BP, blood pressure; CM, cardiometabolic; DM, diabetes mellitus; HTN, hypertension; MDD, major depressive disorder.

DNA methylation age refers to the biological age calculated by the Horvath epigenetic clock; ∆ Age is the difference between DNA methylation age and chronological age; DNAm AA is (DNA methylation age acceleration) is the value of the residual of DNA methylation age regressed on chronological age in a linear model.

Spearman’s correlations indicated a significant positive association between DNAm AA and BMI (P-value <.001; Table 2). Contrary to our hypothesis, no significant associations were observed between overall depressive symptom severity (BDI total score) or symptom domain scores (affective, cognitive, and somatic depressive symptoms scores) and DNAm AA. No significant associations between depressive symptoms and DNAm AA were observed in subsamples with and without a CM trait (Table 3). In further exploratory analysis of BDI items (Supplemental Table 1), DNAm AA demonstrated negative correlations with specific depressive symptoms scores including disappointment, disgust, or hatred toward self (BDI item 7), frequency of crying (BDI item 10), and a positive association with BDI item 20, worry about physical health. However, the application of a Bonferroni correction reduced the alpha significance threshold to a value of P ⩽ .0024. No symptom remained significantly associated after this correction for multiple comparisons. In subsequent regression analysis (Table 4), none of the aforementioned symptoms significantly associated with DNAm AA.

Table 2.

Associations between DNAm AA and depressive symptoms and cardiometabolic indicators as determined by Spearman’s correlations in the sample of African American mothers from the InterGEN study (n = 227).

Variable	DNAm AA
Depression score	−0.02
Affective subscore	−0.13^†
Cognitive subscore	−0.05
Somatic subscore	0.02
Disappointment, disgust, or hatred of oneself	−0.13*
Frequency of crying	−0.08
Difficulty making decisions	−0.15*
Worry/preoccupation with physical health	0.15*
Systolic BP	0.12
Diastolic BP	0.12^†
BMI	0.25***

P < .05. ***P < .001. ^†P ⩽ .06.

Table 3.

Results from Spearman’s correlations between age acceleration variables, depressive symptoms, and cardiometabolic indicators in the CM trait subsample (n = 123) and no CM trait subsample (n = 104).

Variable	CM trait (n = 123)	No CM trait (n = 104)
Depression score	0.03	−0.11
Disappointment, disgust, or hatred of oneself	−0.13	−0.10
Frequency of crying	−0.03	−0.11
Difficulty making decisions	−0.16	−0.16
Worry/preoccupation with physical health	0.16^†	0.04

P < .05. **P < .01. ***P < .001. ^†P ⩽ .06.

Table 4.

Results from linear regression of DNAm AA and depressive symptoms.

Variable	ß	(95% CI)
Depression score	−.01	(−0.53, 1.19)
Affective subscore	−.19	(−0.45, 0.064)
Disappointment, disgust, or hatred of oneself	.33	(−0.53, 1.19)
Frequency of crying	−.43	(1.23, 0.39)
Difficulty making decisions	−1.01^†	(−2.16, 0.19)
Worry/preoccupation with physical health	.73^†	(−0.07, 1.54)

^†P ⩽ .06.

Logistic regression analysis demonstrated significant positive associations between DNAm AA and CM traits (Table 5)—specifically hypertension, DM, and obesity, in unadjusted and adjusted models which included smoking and sociodemographic factors as covariates. Wilcoxon signed-rank tests detected no significant differences between depressive symptom severity scores between subsamples with or without at least 1 CM trait (binary) or subsamples divided according to positive versus negative DNAm AA (binary). DNAm AA was not significantly associated with depression clinical cut-offs (BDI ⩾ 10), or sociodemographic variables (ie, marital status, multiracial status, ethnicity, education, income, or employment) in Chi-squared analysis.

Table 5.

Results from logistic regression models of DNAm AA and cardiometabolic traits.

Cardiometabolic trait	Unadjusted		Adjusted model**
Cardiometabolic trait	ß	(95% CI)	ß	(95% CI)
Hypertension	.08	(1.00, 1.17)	.08	(1.01, 1.17)
Diabetes	.18	(1.01, 1.35)	.20	(1.09, 1.40)
Obesity	.07	(1.01, 1.15)	.08	(1.02, 1.16)

The adjusted model includes smoking, income, education, marital status, ethnicity, and total depression score as covariates.

Discussion

Refuting our initial hypothesis, no association between DNAm AA—as determined by the Horvath epigenetic clock—and total depressive symptoms was detected among the AAW sampled from the InterGEN study. There was also no association between DNAm AA and specific depressive symptom domain subscores nor were there associations between DNAm AA and depressive symptoms in subsamples with versus without CM traits. Irrespective of depressive symptoms, significantly higher DNAm AA was associated with participants with DM, hypertension, and obesity. In this sample, we observed negative DNAm AA (ie, participants were biologically younger than chronological age), an unexpected finding in light of the known risk factors typical of the target population. These findings, however, are consistent with a prior analysis in the InterGEN cohort³¹ and with reported slower rates of biological aging among AAW compared to men and White women.⁵³ The possibility exists that AAW may not have been well represented in the training data used to generate the estimate of biological age, and concern for racial bias has been raised in similar DNAm age clocks as many methylation sites included in these algorithms are related to genetic ancestry as well as aging.⁵⁴ Alternatively, this finding could represent resiliency—perhaps more characteristic of healthier self-selected volunteers than the target population in general—warranting further exploration of this paradoxical discrepancy between biological and chronological age.

Consistency of findings with similar studies

Overall depressive symptom severity

Although several lines of evidence suggest that depression and depressive symptoms are associated with age-dependent molecular markers and age-related pathologies,^22,27,28,34 we observed no evidence of association between depressive symptoms and DNAm AA, similar to studies prior studies with more representation of African Americans.^33,55 Crawford et al⁵⁵ did not detect an association between participants with a history of depression and Horvath-estimated DNAm AA; however, they did observe significantly decreased telomere lengths in these individuals. In contrast, Whalley et al²⁸ reported significant DNAm AA with a small effect size (0.20 year) estimated by the Horvath (but not Hannum) clock among subjects with depression in a study of 1219 depression cases and 3833 healthy controls. Transcriptomic and age-dependent gene expression methods have detected positive associations between age acceleration and depressive symptoms.^22,34 Collectively, these findings suggest that the associations between age acceleration and depressive symptoms may be highly nuanced and susceptible to reduced detectability in certain samples and study design contexts. This nuance is plausible given that the aforementioned studies that detected associations based their DNAm analysis on blood or brain tissue, a distinction from this study which used saliva samples. Distinctions in sample characteristics between this study and prior studies may have further reduced comparability and/or detectability of any effect in this sample of AAW. For example, all but two^33,34 of the aforementioned studies included samples of predominately White males and females—a large proportion of whom had depressive scores near or exceeding the clinical threshold for diagnosis of major depressive disorder.^27,28,35,44 In contrast, the InterGEN cohort is comprised exclusively of Black/African American mothers with predominately low levels of depression symptoms, typical of self-selected volunteers.

Specific depressive symptoms

Although no significant association was found concerning total depressive symptom severity scores and DNAm AA, the findings of this secondary analysis do not preclude the possibility of depressive symptom associations with biological aging. Another study that examined DNAm AA in a sample of 3720 men and women found that although DNAm AA (Horvath) did not significantly associate with total depressive symptom scores, increased DNAm AA was associated with reduced positive affect, even after adjusting for multiple comparisons,³⁵ in contrast to the current study.

Several of the specific depressive symptoms that demonstrated a correlation (both positive and negative), before correction for multiple comparisons, with DNAm AA in the current study (ie, self-hate/disgust, difficulty with decision making, and worry/preoccupation with health), have been identified in relation to age acceleration in prior studies.^56-58 The concept of self-hate may have particular relevance to the population of AAW,⁵⁹ but this has yet to be fully explored in relation to aging. Self-hate/disgust and difficulty with decision-making are prominent depressive symptoms among persons with chronic conditions, including CM illnesses,^57,58 which makes the negative correlation between DNAm AA and self-hate/ self-disgust, observed in this sample unexpected and worthy of validation and additional follow-up. Increased health worry, anxiety, and/or cumulative stress had been associated with increased DNAm AA,³³ and limitations in cognitive functioning, including decision-making difficulty, have demonstrated associations with increases in Horvath-estimated DNAm AA.^60,61 The initial findings of the current study that these symptoms varied in correspondence with DNAm AA are not inconsistent with the notion of heterogeneity in depression; however, multiple comparison correction yielded non-significant findings, which may be resultant in part from reduced detectability in a sample whose overall depression scores were low.

Age acceleration and cardiometabolic illness

With respect to age acceleration associations in CM illness, the findings of this study are consistent with prior analysis showing that DNAm AA associates with hypertension,^31,62 but in contrast with earlier studies less representative of AAW,^29,30 suggesting that this finding is particularly salient for this population. Increased DNAm AA has also been demonstrated in cardiovascular disease and hypertensive target organ damage.³⁰ Therefore, detectability of age acceleration in cardiac disease may be threshold dependent, relative to disease stage, and variable across samples.

The results of this analysis are consistent with findings from a prior analysis by the InterGEN study team that demonstrated a strong positive relationship between BMI and obesity-related traits and DNAm AA.³¹ The current study adds to the growing body of literature concerning the relationship between DM and accelerated aging.²⁰ Among women with DM, increased Horvath-estimated DNAm AA has been associated with increases in BMI, waist circumference, and fasting blood glucose,³² and recently GrimAge accelerated aging has demonstrated associations with fasting insulin in African Americans⁶². The current secondary analysis further demonstrates significant associations between DM/obesity and DNAm AA, specifically in AAW, a noteworthy finding considering the population’s high degree of CM risk.

Other factors

Although we examined depressive symptoms in relation to DNAm AA in this study, additional factors including discrimination and racism could also play a role in DNAm AA. Mounting evidence confirms the damaging effects of perceived discrimination and other components of racism on health,^63,64 particularly CM health in AAW.^65-67 In line with the notion of biological weathering^68,69—described as the early deterioration of health as a physiologic result of socioeconomic disadvantage—several studies show associations between measures of racial discrimination and socioeconomic stress with accelerated aging and DNA methylation among AAW.^70-73 Further, experiences of discrimination may also affect children of AAW, through effects on parenting stress, highlighting the intergenerational impact of discrimination on health.^71,74 Future studies should consider this major stressor among AAW when examining DNAm AA.

Study strengths and limitations

Although a target sample size was determined through power analysis centered on the small to moderate effect size reported in prior studies,^27,28,35,44 this study may have been underpowered to detect small, but meaningful, associations between depressive symptom severity scores and DNAm AA. Generalizability may be affected by the relationship between fertility and biological aging as well as the role that menopause plays in accelerating biological aging.^75,76 Because the participants in this sample were all mothers with a median chronological age under 35 years and low BDI scores, the results of this study should not be generalized to significantly depressed, nulliparous, or post-menopausal women. Currently, there remains a dearth of available datasets with both quality genomic data and highly valid depressive symptom measures with samples of AAW—even fewer with data reflecting medical comorbidities. This lack of information underscores the uniqueness and importance of the InterGEN cohort data, which provided a rare, yet urgently needed, opportunity to examine accelerated aging and depressive symptoms in an understudied population with disproportionate risk. Because of the cross-sectional nature of the study design, there can be no distinction of cause from consequence concerning the noted correlates of DNAm AA.

This study has several notable strengths. First, the rigor and richness of the parent study data allowed for the exploration of the primary research question in a population with disproportionate risk. In addition to the use of a well-validated measure of depressive symptoms, this study also examined domains of depressive symptoms including cognitive, somatic, and affective domains and specific depressive symptoms, in light of the heterogeneity of the phenotype. Additionally, a standard, accessible, and multi-tissue DNAm marker of epigenetic aging was used, enhancing the replicability of these findings for future studies. The current analysis adds to the nascent literature demonstrating associations between DM/obesity and DNAm AA among AAW, a group with disproportionately high rates of morbidity and mortality compared to other groups, including African American men.

Implications for clinical practice and future research

DNA methylation patterns, unlike gene sequences, can be influenced by many environmental behavioral, and lifestyle factors (eg, diet, smoking, stress) throughout the life course. Determinants such as diets high in plants and lean meats and physical activity have been linked to slower epigenetic aging, confirming current wisdom surrounding the relationship between lifestyle and longevity.⁷⁷ Patient education regarding these modifiable risk factors should remain a mainstay of prevention and management for a variety of age-related conditions. The finding that DNAm AA is strongly associated with DM and obesity in this sample underscores the importance and urgency of diabetes and prediabetes prevention and management among AAW. Clinicians must redouble their health promotion and screening efforts—starting at younger ages in this highest-risk population. These evidence-based efforts include monitoring for the development of DM annually, referring patients with prediabetes to intensive lifestyle behavior change programs modeled after the Diabetes Prevention Program that include nutritional counseling, and encouraging weekly targets of at least 150 minutes of moderate-intense physical activity.⁷⁸ Currently, metformin therapy is recommended in those with prediabetes, especially those over age 60 or with BMI ⩾ 35.⁷⁸ In light of both its anti-diabetes and potential anti-aging effects,^79,80 metformin regimens could conceivably be justified for use earlier within the course of disease development in this population, although further research is needed. Diabetes prevention efforts must also be supported at the systems level, through initiatives aimed at improving access and affordability of care and increasing high-quality patient-centered care environments equipped with coordinated care teams adept at assessing and addressing psychological concerns and social determinants of health (eg, food insecurity and homelessness) known to have detrimental effects on diabetes outcomes.⁸¹

There are several important implications for future studies, specifically the need for larger samples that include wider age ranges, a full array of severity of depressive symptoms, and women across the adult lifespan. Additional studies with larger samples of AAW, particularly genomic studies, which address considerations outlined by the American Heart Association,⁸² are needed to better understand factors related to DNAm AA in AAW. Focus on samples of women with 1 CM illness, and increasing distinctions between groups with and without depressive symptoms may address the need for increased power in future studies by reducing variance and increasing the magnitude of group differences in depressive symptoms. Longitudinal designs could offer an opportunity to assess temporal relationships between epigenetic age acceleration and clinical outcomes. The additional inclusion of environmental, sociodemographic, and stress-related variables, particularly lifetime/cumulative stress and early childhood experiences—as these have been significant in prior studies examining DNAm AA—would further enhance this objective.^33,83,84 Future studies should also consider the use of more than one method of biological age estimation, as several have been developed and further refined since the completion of this study.

Conclusions

This study aimed to address several substantial gaps in the current literature surrounding depressive symptoms and accelerated aging in AAW with or at risk for CM traits. Despite the noted limitations, this secondary analysis adds to the literature by posing this question in an understudied population with disproportionate risk and approaching inquiry with a focus on symptoms rather than psychiatric diagnostic labels. The notion that epigenetic aging estimated by DNAm markers may be associated with specific depressive symptoms, rather than total depressive symptom scores reinforces the need for the RDoC approach, as suggested by NIMH, to examine potential biomarkers alongside continuous symptoms or domains—as opposed to sole reliance on diagnostic criteria that frequently yields heterogeneous cases and lacks biological validity.¹⁰ This study also demonstrated links between CM traits and accelerated aging, independent of depressive symptoms. Further research with larger, well-defined samples is warranted to clarify these findings and better inform clinical practice.

Supplemental Material

sj-docx-1-gae-10.1177_25168657221109781 – Supplemental material for Associations Between DNA Methylation Age Acceleration, Depressive Symptoms, and Cardiometabolic Traits in African American Mothers From the InterGEN Study

Supplemental material, sj-docx-1-gae-10.1177_25168657221109781 for Associations Between DNA Methylation Age Acceleration, Depressive Symptoms, and Cardiometabolic Traits in African American Mothers From the InterGEN Study by Nicole Beaulieu Perez, Allison A Vorderstrasse, Gary Yu, Gail D’Eramo Melkus, Fay Wright, Stephen D Ginsberg, Cindy A Crusto, Yan V Sun and Jacquelyn Y Taylor in Epigenetics Insights

Footnotes

Declaration of Conflicting Interests:

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding:

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The InterGEN study was funded by the National Institute of Nursing Research of the National Institutes of Health (R01NR013520). Nicole Beaulieu Perez held a predoctoral position at NYU Grossman School of Medicine, Clinical Translational Science Institute funded by the National Institutes of Health (TL1TR001447). This work was also supported by a research grant from Sigma Theta Tau International Honor Society of Nursing.

Authors Contributions

NBP- conceptualization, formal analysis, visualization, writing original draft, writing review and editing, project administration.

GEM- conceptualization, methodology, supervision, writing review and editing.

AAW- conceptualization, methodology, supervision, writing review and editing.

FW- methodology, writing review and editing.SG- methodology, writing review and editing.

GY- conceptualization, methodology, formal analysis, writing review and editing.

YVS- formal analysis, writing review and editing.

CAC- funding acquisition, writing review and editing.

JYT- conceptualization, methodology, funding acquisition, writing review and editing.

ORCID iD

Jacquelyn Y Taylor

Supplemental Material

Supplemental material for this article is available online.

References

Blüher

. Obesity: global epidemiology and pathogenesis. Nat Rev Endocrinol. 2019;15:288-298.

Bommer

Sagalova

Heesemann

, et al. Global economic burden of diabetes in adults: projections from 2015 to 2030. Diabetes Care. 2018;41:963-970.

Forouzanfar

Liu

Roth

, et al. Global burden of hypertension and systolic blood pressure of at least 110 to 115 mm Hg, 1990–2015. JAMA. 2017;317:165-182.

World Health Organization. Depression and Other Common Mental Disorders: Global Health Estimates. World Health Organization. 2017. Accessed March 12, 2019. https://apps.who.int/iris/handle/10665/254610

Centers for Disease Control (CDC). Prevalence of diagnosed diabetes. Published August 7, 2020. Accessed August 5, 2021. http://www.cdc.gov/diabetes/data/statistics-report/diagnosed-diabetes.html

Hales

. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief. 2017;288:1-8.

Perez

Lanier

Squires

. Inequities along the depression care cascade in African American women: an integrative review. Issues Ment Health Nurs. 2021;42:720-729.

Williams

González

Neighbors

, et al. Prevalence and distribution of major depressive disorder in African Americans, Caribbean Blacks, and non-Hispanic whites: results from the National Survey of American Life. Arch Gen Psychiatry. 2007;64:305-315.

Sohail

Bailey

Richie

. Misconceptions of depression in African Americans. Front Psychiatry. 2014;5:65.

10.

Katahira

Yamashita

. A theoretical framework for evaluating psychiatric research strategies. Comput Psychiatr. 2017;1:184-207.

11.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-5. 5th ed. American Psychiatric Association; 2013.

12.

Sullivan

O’Connor

Feeney

, et al. Depression predicts all-cause mortality: epidemiological evaluation from the ACCORD HRQL substudy. Diabetes Care. 2012;35:1708-1715.

13.

Wallace

Zhao

Misra

Sambamoorthi

. The humanistic and economic burden associated with anxiety and depression among adults with comorbid diabetes and hypertension. J Diabetes Res. 2018;2018:4842520.

14.

Milaneschi

Lamers

Berk

Penninx

BWJH

. Depression heterogeneity and its biological underpinnings: toward immunometabolic depression. Biol Psychiatry. 2020;88:369-380.

15.

Milaneschi

Simmons

van Rossum

EFC

Penninx

. Depression and obesity: evidence of shared biological mechanisms. Mol Psychiatry. 2019;24:18-33.

16.

Gao

Barcelona

DeWan

, et al. Depressive symptoms and blood pressure in African American women: a secondary analysis from the intergenerational impact of genetic and psychological factors on blood pressure study. J Cardiovasc Nurs. Published online March 16, 2021. doi:10.1097/JCN.0000000000000800

17.

Kahl

Stapel

Frieling

. Link between depression and cardiovascular diseases due to epigenomics and proteomics: focus on energy metabolism. Prog Neuropsychopharmacol Biol Psychiatry. 2019;89:146-157.

18.

Lyra

Silva

Lam

Soares

Munoz

Milev

De Felice

. Insulin resistance as a shared pathogenic mechanism between depression and type 2 diabetes. Front Psychiatry. 2019;10:57.

19.

Wolkowitz

Reus

Mellon

. Of sound mind and body: depression, disease, and accelerated aging. Dialogues Clin Neurosci. 2011;13:25-39.

20.

Morrison

Logue

Guetta

, et al. Investigation of bidirectional longitudinal associations between advanced epigenetic age and peripheral biomarkers of inflammation and metabolic syndrome. Aging. 2019;11:3487-3504.

21.

Révész

Milaneschi

Verhoeven

Penninx

BWJH

. Telomere length as a marker of cellular aging is associated with prevalence and progression of metabolic syndrome. None. 2014;99:4607-4615.

22.

Sibille

. Molecular aging of the brain, neuroplasticity, and vulnerability to depression and other brain-related disorders. Dialogues Clin Neurosci. 2013;15:53-65.

23.

Luca

Calandra

. Accelerated aging in major depression: the role of nitro-oxidative stress. Oxid Med Cell Longev. 2013;2013:230797

24.

Marioni

Shah

McRae

, et al. DNA methylation age of blood predicts all-cause mortality in later life. Genome Biol. 2015;16:16.

25.

Horvath

. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14:3156.

26.

Gillespie

Hardy

Anderson

. Patterns of DNA methylation as an indicator of biological aging: state of the science and future directions in precision health promotion. Nurs Outlook. 2019;67:337-344.

27.

Han

LKM

Aghajani

Clark

, et al. Epigenetic aging in major depressive disorder. Am J Psychiatry. 2018;175:774-782.

28.

Whalley

Gibson

Marioni

, et al. Accelerated epigenetic ageing in major depressive disorder. bioRxiv. Published online October 28, 2017. doi:10.1101/210666

29.

Horvath

Gurven

Levine

, et al. An epigenetic clock analysis of race/ethnicity, sex, and coronary heart disease. Genome Biol. 2016;17:171.

30.

Smith

Raisky

Ratliff

, et al. Intrinsic and extrinsic epigenetic age acceleration are associated with hypertensive target organ damage in older African Americans. BMC Med Genomics. 2019;12:141.

31.

Wang

Hardy

, et al. Association of obesity with DNA methylation age acceleration in African American mothers from the InterGEN Study. Int J Mol Sci. 2019;20:4273.

32.

Grant

Jafari

Hou

, et al. A longitudinal study of DNA methylation as a potential mediator of age-related diabetes risk. GeroScience. 2017;39:475-489.

33.

Zannas

Arloth

Carrillo-Roa

, et al. Lifetime stress accelerates epigenetic aging in an urban, African American cohort: relevance of glucocorticoid signaling. Genome Biol. 2018;19:266.

34.

Carter

Ong

Simons

Gibbons

Lei

Beach

SRH

. The effect of early discrimination on accelerated aging among African Americans. Health Psychol. 2019;38:1010-1013.

35.

Beydoun

Hossain

Chitrala

, et al. Association between epigenetic age acceleration and depressive symptoms in a prospective cohort study of urban-dwelling adults. J Affect Disord. 2019;257:64-73.

36.

Carroll

Irwin

Levine

, et al. Epigenetic aging and immune senescence in women with insomnia symptoms: findings from the Women’s Health Initiative Study. Biol Psychiatry. 2017;81:136-144.

37.

Cruz-Almeida

Sinha

Rani

Huo

Fillingim

Foster

. Epigenetic aging is associated with clinical and experimental pain in community-dwelling older adults. Mol Pain. 2019;15:1744806919871819.

38.

Taylor

Wright

Crusto

Sun

. The intergenerational impact of genetic and psychological factors on blood pressure (InterGEN) study: design and methods for complex DNA analysis. Biol Res Nurs. 2016;18:521-530.

39.

Crusto

Barcelona

Mendoza

Connell

Sun

Taylor

. The intergenerational impact of genetic and psychological factors on blood pressure study (InterGEN): design and methods for recruitment and psychological measures. Nurs Res. 2016;65:331-338.

40.

Beck

Steer

Carbin

. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clin Psychol Rev. 1988;8:77-100.

41.

Grothe

Dutton

Jones

Bodenlos

Ancona

Brantley

. Validation of the Beck Depression Inventory-II in a low-income African American sample of medical outpatients. Psychol Assess. 2005;17:110-114.

42.

Levine

Quach

, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10:573-591.

43.

Quach

Wilson

, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging. 2019;11:303-327.

44.

Vyas

Hazra

Chang

, et al. Pilot study of DNA methylation, molecular aging markers and measures of health and well-being in aging. Transl Psychiatry. 2019;9:118.

45.

Lim

Nzegwu

Wright

. The impact of psychosocial stress from life trauma and racial discrimination on epigenetic aging—a systematic review. Biol Res Nurs. 2022;24:202-215.

46.

Simons

Lei

Beach

SRH

, et al. Economic hardship and biological weathering: the epigenetics of aging in a U.S. sample of black women. Soc Sci Med. 2016;150:192-200.

47.

R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; 2021. https://www.R-project.org

48.

Wickman

Francois

Henry

Muller

. Dplyr: A Grammar of Data Manipulation. R Package Version 1.0.8; 2022. https://CRAN.R-project.org/package=dplyr

49.

Harrel

Jr.

Hmisc: Harrell Miscellaneous. R Package Version 4.6-0; 2021. https://CRAN.R-project.org/package=Hmisc

50.

Tierney

Cook

McBain

Fay

. Nanair: Data Structures, Summaries, and Visualizations for Missing Data. R Package Version 0.6.0; 2020. http://CRAN.R-project.org/package=naniar

51.

Revelle

. Psych: Procedures and Personality and Psychological Research. Northwestern University; 2019. https://cran.r-project.org/web/packages/psych/index.html

52.

Harrison

Drake

Ots

. Finalfit: Quickly Create Elegant Regression Results Tables and Plots When Modelling. R Package Version 1.0.2; 2020. https://cran.r-project.org/web/packages/finalfit/index.html

53.

Tajuddin

Hernandez

Chen

, et al. Novel age-associated DNA methylation changes and epigenetic age acceleration in middle-aged African Americans and whites. Clin Epigenetics. 2019;11:119.

54.

Philibert

Beach

SRH

Lei

, et al. Array-based epigenetic aging indices may be racially biased. Genes. 2020;11:685.

55.

Crawford

Craig

Mansell

, et al. DNA methylation and inflammation marker profiles associated with a history of depression. Hum Mol Genet. 2018;27:2840-2850.

56.

Chae

Epel

Nuru-Jeter

, et al. Discrimination, mental health, and leukocyte telomere length among African American men. Psychoneuroendocrinology. 2016;63:10-16.

57.

Francoeur

. Symptom profiles of subsyndromal depression in disease clusters of diabetes, excess weight, and progressive cerebrovascular conditions: a promising new type of finding from a reliable innovation to estimate exhaustively specified multiple indicators-multiple causes (MIMIC) models. Diabetes Metab Syndr Obes. 2016;9:391-416.

58.

Abbey

Toner

Garfinkel

Kennedy

Kaplan

. Self-report symptoms that predict major depression in patients with prominent physical symptoms. Int J Psychiatr Med. 1990;20:247-258.

59.

Walton

Shepard Payne

. Missing the mark: cultural expressions of depressive symptoms among African-American women and men. Soc Work Ment Health. 2016;14:637-657.

60.

Levine

Bennett

Horvath

. Epigenetic age of the pre-frontal cortex is associated with neuritic plaques, amyloid load, and Alzheimer’s disease related cognitive functioning. Aging. 2015;7:1198-1211.

61.

Marioni

Shah

McRae

, et al. The epigenetic clock is correlated with physical and cognitive fitness in the Lothian birth cohort 1936. Int J Epidemiol. 2015;44:1388-1396.

62.

Ammous

Zhao

Ratliff

, et al. Epigenetic age acceleration is associated with cardiometabolic risk factors and clinical cardiovascular disease risk scores in African Americans. Clin Epigenetics. 2021;13:55.

63.

Williams

Mohammed

. Discrimination and racial disparities in health: evidence and needed research. J Behav Med. 2009;32:20-47.

64.

Williams

. Race, socioeconomic status, and health the added effects of racism and discrimination. Ann N Y Acad Sci. 1999;896:173-188.

65.

Wright

Lim

Sales

, et al. The influence of discrimination and coping style on blood pressure among Black/African American women in the InterGEN study. Health Equity. 2020;4:272-279.

66.

Basile Ibrahim

Barcelona

Condon

Crusto

Taylor

. The association between neighborhood social vulnerability and cardiovascular health risk among Black/African American women in the InterGEN study. Nurs Res. 2021;70:S3-S12.

67.

Millender

Barile

Bagneris

, et al. Associations between social determinants of health, perceived discrimination, and body mass index on symptoms of depression among young African American mothers. Arch Psychiatr Nurs. 2021;35:94-101.

68.

Geronimus

. The weathering hypothesis and the health of African-American women and infants: evidence and speculations. Ethn Dis. 1992;2:207-221.

69.

Geronimus

Hicken

Pearson

Seashols

Brown

Cruz

. Do US Black women experience stress-related accelerated biological aging?: a novel theory and first population-based test of black-white differences in telomere length. Hum Nat. 2010;21:19-38.

70.

Brown

Hui

Huang

, et al. Association between stress and coping with DNA methylation of blood pressure-related genes among African American women. Chronic Stress. 2019;3:1-9. doi:10.1177/2470547019879088

71.

Wright

Huang

Hui

, et al. Parenting stress and DNA methylation among African Americans in the InterGEN study. J Clin Transl Sci. 2017;1:328-333.

72.

Barcelona de Mendoza

Huang

Crusto

Sun

Taylor

. Perceived racial discrimination and DNA methylation among African American women in the InterGEN study. Biol Res Nurs. 2018;20:145-152.

73.

McCrory

Fiorito

Ni Cheallaigh

, et al. How does socio-economic position (SEP) get biologically embedded? A comparison of allostatic load and the epigenetic clock(s). Psychoneuroendocrinology. 2019;104:64-73.

74.

Condon

Barcelona

Ibrahim

Crusto

Taylor

. Racial discrimination, mental health, and parenting among African American mothers of preschool-aged children. J Am Acad Child Adolesc Psychiatry. 2022;61:402-412.

75.

Levine

Chen

, et al. Menopause accelerates biological aging. Proc Natl Acad Sci U S A. 2016;113:9327-9332.

76.

Morin

Tao

Marin

, et al. DNA methylation-based age prediction and telomere length in white blood cells and cumulus cells of infertile women with normal or poor response to ovarian stimulation. Aging. 2018;10:3761-3773.

77.

Quach

Levine

Tanaka

, et al. Epigenetic clock analysis of diet, exercise, education, and lifestyle factors. Aging. 2017;9:419-446.

78.

American Diabetes Association. 3. Prevention or delay of type 2 diabetes: standards of medical care in diabetes—2021. Diabetes Care. 2021;44:S34-S39.

79.

Kessing

Rytgaard

Ekstrøm

Knop

Berk

Gerds

. Antidiabetes agents and incident depression: a nationwide population-based study. Diabetes Care. 2020;43:3050-3060.

80.

Barzilai

Crandall

Kritchevsky

Espeland

. Metformin as a tool to target aging. Cell Metab. 2016;23:1060-1065.

81.

American Diabetes Association. Improving care and promoting health in populations: standards of medical care in diabetes-2021. Diabetes Care. 2021;44:S7-S14.

82.

Mudd-Martin

Cirino

Barcelona

, et al. Considerations for cardiovascular genetic and genomic research with marginalized racial and ethnic groups and indigenous peoples: a scientific statement from the American Heart Association. Circ Genom Precis Med. 2021;14:e000084.

83.

Tang

Howe

Suderman

Relton

Crawford

Houtepen

. Adverse childhood experiences, DNA methylation age acceleration, and cortisol in UK children: a prospective population-based cohort study. Clin Epigenetics. 2020;12:55.

84.

Austin

Chen

Ross

, et al. Early-life socioeconomic disadvantage, not current, predicts accelerated epigenetic aging of monocytes. Psychoneuroendocrinology. 2018;97:131-134.

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

0.03 MB