Immigration Status and Breast Cancer Surgery Quality of Care Metrics: A Population-Level Analysis

INTRODUCTION

As immigrant women face challenges accessing health care, we hypothesized that immigration status would be associated with fewer women with breast cancer receiving surgery for curable disease, fewer undergoing breast conserving surgery (BCS), and longer wait time to surgery.

METHODS

A population-level retrospective cohort study, including women aged 18-70 years with Stage I-III breast cancer diagnosed between 2010 and 2016 in Ontario was conducted. Multivariable analysis was performed to assess odds of undergoing surgery, receiving BCS and wait time to surgery.

RESULTS

A total of 31,755 patients were included [26,253 (82.7%) Canadian-born and 5502 (17.3%) immigrant women]. Immigrant women were younger (mean age 51.6 vs. 56.1 years) and less often presented with Stage I/II disease (87.4% vs. 89.8%) (both p < .001). On multivariable analysis, there was no difference between immigrant women and Canadian-born women in odds of undergoing surgery [Stage I OR 0.93 (95% CI 0.79-1.11), Stage II 1.04 (0.89-1.22), Stage III 1.22 (0.94-1.57)], receiving BCS [Stage I 0.93 (0.82-1.05), Stage II 0.96 (0.86-1.07), Stage III 1.00 (0.83-1.22)], or wait time [Stage I 0.45 (-0.61-1.50), Stage II 0.33 (-0.86-1.52), Stage III 3.03 (-0.05-6.12)]. In exploratory analysis, new immigrants did not have surgery more than established immigrants (12.9% vs. 10.1%), and refugee women had longer wait time compared with economic-class immigrants (39.5 vs. 35.3 days).

CONCLUSIONS

We observed differences in measures of socioeconomic disadvantage and disease characteristics between immigrant and Canadian-born women with breast cancer. Upon adjusting for these factors, no differences emerged in rate of surgery, rate of BCS, and time to surgery. The lack of disparity suggests barriers to accessing basic components of breast cancer care may be mitigated by the universal healthcare system in Canada.

Stochastic Epigenetic Mutations: Reliable Detection and Associations with Cardiovascular Aging

Stochastic Epigenetic Mutations (SEMs) have been proposed as novel aging biomarkers that have the potential to capture heterogeneity in age-related DNA methylation (DNAme) changes. SEMs are defined as outlier methylation patterns at cytosine-guanine dinucleotide (CpG) sites, categorized as hypermethylated (hyperSEM) or hypomethylated (hypoSEM) relative to a reference. While individual SEMs are rarely consistent across subjects, the SEM load - the total number of SEMs - increases with age. However, given poor technical reliability of measurement for many DNA methylation sites, we posited that many outliers might represent technical noise. Our study of whole blood samples from 36 individuals, each measured twice, found that 23.3% of hypoSEM and 45.6% hyperSEM are not shared between replicates. This diminishes the reliability of SEM loads, where intraclass correlation coefficients are 0.96 for hypoSEM and 0.90 for hyperSEM. We linked SEM reliability to multiple factors, including blood cell type composition, probe beta-value statistics, and presence of SNPs. A machine learning approach, leveraging these factors, filtered unreliable SEMs, enhancing reliability in a separate dataset of technical replicates from 128 individuals. Analysis of the Framingham Heart Study confirmed previously reported SEM association with mortality and revealed novel connections to cardiovascular disease. We discover that associations with aging outcomes are primarily driven by hypoSEMs at baseline methylated probes and hyperSEMs at baseline unmethylated probes, which are the same subsets that demonstrate highest technical reliability. These aging associations are preserved after filtering out unreliable SEMs and are enhanced after adjusting for blood cell composition. Finally, we utilize these insights to formulate best practices for SEM detection and introduce a novel R package, SEMdetectR, which utilizes parallel programming for efficient SEM detection with comprehensive options for detection, filtering, and analysis.

Frailty Resilience Score: A Novel Measure of Frailty Resilience Associated With Protection From Frailty and Survival

Frailty is characterized by increased vulnerability to disability and high risk for mortality in older adults. Identification of factors that contribute to frailty resilience is an important step in the development of effective therapies that protect against frailty. First, a reliable quantification of frailty resilience is needed. We developed a novel measure of frailty resilience, the Frailty Resilience Score (FRS), that integrates frailty genetic risk, age, and sex. Application of FRS to the LonGenity cohort (n = 467, mean age 74.4) demonstrated its validity compared to phenotypic frailty and its utility as a reliable predictor of overall survival. In a multivariable-adjusted analysis, 1-standard deviation increase in FRS predicted a 38% reduction in the hazard of mortality, independent of baseline frailty (p < .001). Additionally, FRS was used to identify a proteomic profile of frailty resilience. FRS was shown to be a reliable measure of frailty resilience that can be applied to biological studies of resilience.

Cellular allostatic load is linked to increased energy expenditure and accelerated biological aging

Stress triggers anticipatory physiological responses that promote survival, a phenomenon termed allostasis. However, the chronic activation of energy-dependent allostatic responses results in allostatic load, a dysregulated state that predicts functional decline, accelerates aging, and increases mortality in humans. The energetic cost and cellular basis for the damaging effects of allostatic load have not been defined. Here, by longitudinally profiling three unrelated primary human fibroblast lines across their lifespan, we find that chronic glucocorticoid exposure increases cellular energy expenditure by ∼60%, along with a metabolic shift from glycolysis to mitochondrial oxidative phosphorylation (OxPhos). This state of stress-induced hypermetabolism is linked to mtDNA instability, non-linearly affects age-related cytokines secretion, and accelerates cellular aging based on DNA methylation clocks, telomere shortening rate, and reduced lifespan. Pharmacologically normalizing OxPhos activity while further increasing energy expenditure exacerbates the accelerated aging phenotype, pointing to total energy expenditure as a potential driver of aging dynamics. Together, our findings define bioenergetic and multi-omic recalibrations of stress adaptation, underscoring increased energy expenditure and accelerated cellular aging as interrelated features of cellular allostatic load.

Systems Age: A single blood methylation test to quantify aging heterogeneity across 11 physiological systems

Individuals, organs, tissues, and cells age in diverse ways throughout the lifespan. Epigenetic clocks attempt to quantify differential aging between individuals, but they typically summarize aging as a single measure, ignoring within-person heterogeneity. Our aim was to develop novel systems-based methylation clocks that, when assessed in blood, capture aging in distinct physiological systems. We combined supervised and unsupervised machine learning methods to link DNA methylation, system-specific clinical chemistry and functional measures, and mortality risk. This yielded a panel of 11 system-specific scores- Heart, Lung, Kidney, Liver, Brain, Immune, Inflammatory, Blood, Musculoskeletal, Hormone, and Metabolic. Each system score predicted a wide variety of outcomes, aging phenotypes, and conditions specific to the respective system, and often did so more strongly than existing epigenetic clocks that report single global measures. We also combined the system scores into a composite Systems Age clock that is predictive of aging across physiological systems in an unbiased manner. Finally, we showed that the system scores clustered individuals into unique aging subtypes that had different patterns of age-related disease and decline. Overall, our biological systems based epigenetic framework captures aging in multiple physiological systems using a single blood draw and assay and may inform the development of more personalized clinical approaches for improving age-related quality of life.

OxPhos defects cause hypermetabolism and reduce lifespan in cells and in patients with mitochondrial diseases

Patients with primary mitochondrial oxidative phosphorylation (OxPhos) defects present with fatigue and multi-system disorders, are often lean, and die prematurely, but the mechanistic basis for this clinical picture remains unclear. By integrating data from 17 cohorts of patients with mitochondrial diseases (n = 690) we find evidence that these disorders increase resting energy expenditure, a state termed hypermetabolism. We examine this phenomenon longitudinally in patient-derived fibroblasts from multiple donors. Genetically or pharmacologically disrupting OxPhos approximately doubles cellular energy expenditure. This cell-autonomous state of hypermetabolism occurs despite near-normal OxPhos coupling efficiency, excluding uncoupling as a general mechanism. Instead, hypermetabolism is associated with mitochondrial DNA instability, activation of the integrated stress response (ISR), and increased extracellular secretion of age-related cytokines and metabokines including GDF15. In parallel, OxPhos defects accelerate telomere erosion and epigenetic aging per cell division, consistent with evidence that excess energy expenditure accelerates biological aging. To explore potential mechanisms for these effects, we generate a longitudinal RNASeq and DNA methylation resource dataset, which reveals conserved, energetically demanding, genome-wide recalibrations. Taken together, these findings highlight the need to understand how OxPhos defects influence the energetic cost of living, and the link between hypermetabolism and aging in cells and patients with mitochondrial diseases.

SYSTEM SPECIFIC AGING SCORES: A STATE OF THE ART AGING CLOCK BUILT USING AGING SCORES FROM DIFFERENT BODILY FUNCTIONS

Aging is a highly heterogeneous process at multiple levels. Different individuals, organs, tissues, and cell types are innately diverse and age in quantitatively different manners. Epigenetic clocks have been developed to capture overall degree of aging and typically report a single biological age value. However, single measures fail to provide insight into differential aging across organ systems. Our aim was to develop novel systems-specific methylation clocks, that when assessed in blood, capture distinct aging subtypes. We utilized three large human cohort studies and employed both supervised and unsupervised machine learning models by linking DNA methylation to lower dimensional vectors composed of system specific clinical chemistry and functional assays. In doing so, we were able to develop 11 unique system-specific scores–heart, lung, kidney, liver, brain, immune, inflammatory, hematopoietic, musculoskeletal, hormone, and metabolic. We observe that in independent data, the specific systems relate to meaningful outcomes–for instance the brain score is strongly associated with cognitive functioning; musculoskeletal score is strongly associated with physical functioning; and the lung score is strongly associated with lung cancer. Additionally, system scores and the composite systems clock outperforms presently available clocks in terms of associations with a wide variety of aging phenotypes and conditions. Overall, our biological systems based epigenetic clock outperforms presently available epigenetic aging clocks and provides meaningful insights into heterogeneity in aging.

The Role of Epigenetic Clocks in Explaining Educational Inequalities in Mortality: A Multicohort Study and Meta-analysis

Educational inequalities in all-cause mortality have been observed for decades. However, the underlying biological mechanisms are not well known. We aimed to assess the role of DNA methylation changes in blood captured by epigenetic clocks in explaining these inequalities. Data were from 8 prospective population-based cohort studies, representing 13 021 participants. First, educational inequalities and their portion explained by Horvath DNAmAge, Hannum DNAmAge, DNAmPhenoAge, and DNAmGrimAge epigenetic clocks were assessed in each cohort via counterfactual-based mediation models, on both absolute (hazard difference) and relative (hazard ratio) scales, and by sex. Second, estimates from each cohort were pooled through a random effect meta-analysis model. Men with low education had excess mortality from all causes of 57 deaths per 10 000 person-years (95% confidence interval [CI]: 38, 76) compared with their more advantaged counterparts. For women, the excess mortality was 4 deaths per 10 000 person-years (95% CI: -11, 19). On the relative scale, educational inequalities corresponded to hazard ratios of 1.33 (95% CI: 1.12, 1.57) for men and 1.15 (95% CI: 0.96, 1.37) for women. DNAmGrimAge accounted for the largest proportion, approximately 50%, of the educational inequalities for men, while the proportion was negligible for women. Most of this mediation was explained by differential effects of unhealthy lifestyles and morbidities of the World Health Organization (WHO) risk factors for premature mortality. These results support DNA methylation-based epigenetic aging as a signature of educational inequalities in life expectancy emphasizing the need for policies to address the unequal social distribution of these WHO risk factors.

Reconciliation in human adults: a video-assisted naturalistic observational study of post conflict conciliatory behaviour in interpersonal aggression

Reconciliation is an aspect of conflict resolution, with similar behavioural patterns documented in non-human primates, human children, and human adults of non-Western, non-industrialized cultures. Reconciliation amongst adults of industrialized societies has rarely been studied. We observed naturally occurring conflicts between adults, captured by public security cameras in England. Reconciliation was found in one-quarter of all conflicts and was more prevalent in milder conflicts. Reconciliation typically occurred spontaneously between opponents — and was found within friendship groups and across stranger groups. Reconciliation between opponents also appeared to be stimulated by peers, law enforcement, or shared objects. In some instances, reconciliation extended beyond the initial conflict dyad toward victimized third-party peacemakers. These findings add to growing cross-cultural and cross-species evidence demonstrating the presence and function of post-conflict reconciliation. We extend the repertoire of reconciliatory behaviour and introduce five common features of reconciliation that are central to the study of adult peacemaking.

The Socioeconomic Gradient in Epigenetic Ageing Clocks: Evidence from the Multi-Ethnic Study of Atherosclerosis and the Health and Retirement Study

Epigenetic clocks have been widely used to predict disease risk in multiple tissues or cells. Their success as a measure of biological ageing has prompted research on the connection between epigenetic pathways of ageing and the socioeconomic gradient in health and mortality. However, studies examining social correlates of epigenetic ageing have yielded inconsistent results. We conducted a comprehensive, comparative analysis of associations between various dimensions of socioeconomic status (SES) (education, income, wealth, occupation, neighbourhood environment, and childhood SES) and eight epigenetic clocks in two well-powered US ageing studies: The Multi-Ethnic Study of Atherosclerosis (MESA) (n = 1,211) and the Health and Retirement Study (HRS) (n = 4,018). In both studies, we found robust associations between SES measures in adulthood and the GrimAge and DunedinPoAm clocks (Bonferroni-corrected p-value < 0.01). In the HRS, significant associations with the Levine and Yang clocks were also evident. These associations were only partially mediated by smoking, alcohol consumption, and obesity, which suggests that differences in health behaviours alone cannot explain the SES gradient in epigenetic ageing in older adults. Further analyses revealed concurrent associations between polygenic risk for accelerated intrinsic epigenetic ageing, SES, and the Levine clock, indicating that genetic risk and social disadvantage may contribute additively to faster biological aging.

Evidence of accelerated epigenetic aging of breast tissues in patients with breast cancer is driven by CpGs associated with polycomb-related genes

Purpose

Age is one of the strongest risk factors for the development of breast cancer, however, the underlying etiology linking age and breast cancer remains unclear. We have previously observed links between epigenetic aging signatures in breast/tumor tissue and breast cancer risk/prevalence. However, these DNA methylation-based aging biomarkers capture diverse epigenetic phenomena and it is not known to what degree they relate to breast cancer risk, and/or progression.

Methods

Using six epigenetic clocks, we analyzed whether they distinguish normal breast tissue adjacent to tumor (cases) vs normal breast tissue from healthy controls (controls).

Results

The Levine (p = 0.0037) and Yang clocks (p = 0.023) showed significant epigenetic age acceleration in cases vs controls in breast tissue. We observed that much of the difference between cases and controls is driven by CpGs associated with polycomb-related genes. Thus, we developed a new score utilizing only CpGs associated with polycomb-related genes and demonstrated that it robustly captured epigenetic age acceleration in cases vs controls (p = 0.00012). Finally, we tested whether this same signal could be seen in peripheral blood. We observed no difference in cases vs. controls and no correlation between matched tissue/blood samples, suggesting that peripheral blood is not a good surrogate marker for epigenetic age acceleration.

Conclusions

Moving forward, it will be critical for studies to elucidate whether epigenetic age acceleration in breast tissue precedes breast cancer diagnosis and whether methylation changes at CpGs associated with polycomb-related genes can be used to assess the risk of developing breast cancer among unaffected individuals.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13148-022-01249-z.

Extending human healthspan and longevity: a symposium report

For many years, it was believed that the aging process was inevitable and that age-related diseases could not be prevented or reversed. The geroscience hypothesis, however, posits that aging is, in fact, malleable and, by targeting the hallmarks of biological aging, it is indeed possible to alleviate age-related diseases and dysfunction and extend longevity. This field of geroscience thus aims to prevent the development of multiple disorders with age, thereby extending healthspan, with the reduction of morbidity toward the end of life. Experts in the field have made remarkable advancements in understanding the mechanisms underlying biological aging and identified ways to target aging pathways using both novel agents and repurposed therapies. While geroscience researchers currently face significant barriers in bringing therapies through clinical development, proof-of-concept studies, as well as early-stage clinical trials, are underway to assess the feasibility of drug evaluation and lay a regulatory foundation for future FDA approvals in the future.

Longitudinal Profiling in Phenotypic Metric of Aging: Insights From the Baltimore Longitudinal Study of Aging

It remains challenging to quantify the pace of aging across lifespan due to lack of comprehensive longitudinal measurements across wide range of age. In Baltimore Longitudinal Study of Aging, we have measured the longitudinal trajectories of more than 30 phenotypes across four pre-identified domain - body composition, energy regulation, homeostatic mechanisms and neurodegeneration/neuroplasticity, among participants with age between 20+ and 90+. We implemented a two-stage approach to summarize the longitudinal trajectories of these phenotypes across four domains into a summarized score. We demonstrated that higher summarized score (denoting for slower longitudinal phenotypic decline) is associated with slower decline in both cognitive and physical functions, across different stages of adulthood. Our results imply that deep longitudinal profiling contains rich information and may potentially replace diseases as an early endpoint in trials targeting at aging. Further, understanding the underpinning of longitudinal phenotypic trajectories may provide clues to the biological mechanisms of aging.

Vincent Cristofalo “Rising Star”Award: DNA Methylation Landscapes in Aging

The epigenetic code can be thought of as the operating system of the cell. It controls the most basic and critical cellular processes including differentiation, replication, metabolism, and signaling. Yet, with age, the epigenetic landscape is remodeled, bringing about widespread consequences for cellular and tissue identity, integrity, and functioning. But, what if like computer programmers, we could discover how to recode or restore the original program? The revolutionary discoveries by Yamanaka and Takahashi suggests this may be possible. While early experiments showed that Yamanaka factors could be used to convert somatic cells into induced pluripotent stem cells, more recent work by us and others have shown that signatures of epigenetic aging are also wiped clean during this process. What’s more, epigenetic age reversal appears to take place early in the process and thus can be achieved without the cell _needing to dedifferentiate. Building off of this discovery, our lab is combining novel experiments and advanced bioinformatic techniques to decipher the epigenetic code and determine how it is remodeled during aging, development, and reprogramming. In our recent work, we have made advancements in mapping the epigenetic alterations observed in aging and linking them to both cellular processes and disease etiology. We have identified specific age changes in mouse and human cells that reflect mitotic history, cellular senescence, oxidative damage, and mitochondrial dysfunction. We have also demonstrated that these changes inform differences in organismal lifespan and/or disease etiology at the tissue level. Overall, this work has sweeping implications for our basic understanding of epigenetic aging and reprogramming, and will help provide the foundation for potent therapeutics that extend healthspan and lifespan.

The Arc of Astrocyte Aging: Insights From scRNAseq

There is an urgent need to increase our understanding of brain aging and its role in neurodegeneration. While, evidence suggest that many hallmarks of aging, including epigenetic alterations and cellular senescence may be implicated in dementia, studying these and other progressive molecular changes in the brain remains extremely challenging. We asked whether something as simple as artificially aging cells in culture could recapitulate the changes that occur during organismal aging. To test this, we passaged human primary astrocytes and performed single-cell RNA sequencing (scRNAseq) of cells at passages 2-10. We observe that the sequential passaging—that terminates with a cluster of senescent cells—can be captured by manifolds and used to quantify a pseudo-time measure of progressive transcriptional changes. We identify genes underlying this transition and apply this signature of in vitro astrocyte passaging to scRNAseq from human and mouse brain aging studies, demonstrating associations with aging and neuropathology.

A DNAmCULTURE Epigenetic Fingerprint Recapitulates Physiological Aging

Aging elicits dramatic changes to DNA methylation (DNAm), however the causes and consequences of such alterations to the epigenome remain unclear. The utility of biomarkers of aging based on DNAm patterns would be greatly enhanced if in vitro models existed that recapitulated physiological phenotypes such that modulation could garnish mechanistic insights. Using DNAm from serially passaged mouse embryonic fibroblasts, we developed a marker of culture aging and asked if culture phenotypes, like exhaustive replication, are epigenetically analogous to physiological aging. Our measure, termed DNAmCULTURE, accurately estimated passage number and was shown to strongly increase with age when examined in multiple tissues. Furthermore, we observed epigenetic alterations indicative of early cultured cells in animals undergoing caloric restriction and in lung and kidney fibroblasts re-programmed to iPSCs. This study identifies culture-derived alterations to the methylome as physiologically relevant and implicates culture aging as an important feature in known epigenetic aging phenomena.

Biological Aging Predicts Vulnerability to COVID-19 Severity in UK Biobank Participants

The Veterans Health Administration (VA) is increasingly purchasing long-term care for eligible Veterans from non-VA, community nursing homes (CNHs). Antipsychotics present safety risks for older adults, but it is unknown how the prevalent use of antipsychotics at CNHs influences whether newly admitted Veterans will initiate antipsychotic therapy. This study used 2013-2016 VA data, Medicare claims, Nursing Home Compare, and Minimum Data Set (MDS) assessments. We identified 10,531 long-stay CNH episodes for Veterans not prescribed antipsychotics 6 months before CNH admission. We categorized Veterans by whether, 12 months before admission, they were diagnosed with FDA-approved indications (including schizophrenia, Tourette’s syndrome, Huntington’s disease) for antipsychotic use. The exposure was the proportion of all CNH residents prescribed antipsychotics in the quarter preceding a Veteran’s admission. Using adjusted logistic regression, we analyzed two outcomes measured using MDS assessments collected ~100 days after CNH admission: 1) new antipsychotic use, and 2) new diagnosis for an FDA-approved indication among Veterans without these conditions at admission. Among antipsychotic-naïve Veterans admitted to CNHs, 7,924 (75.2%) lacked an antipsychotic indication. Prevalent antipsychotic use in CNHs ranged 0%-10.9% (quintile 1) and 25.7%-91.4% (quintile 5). The odds of initiating antipsychotic use increased with higher CNH antipsychotic use rates (OR=2.52, 95% CI:2.05-3.10, quintile 5 vs. 1), as did the odds of acquiring a new indication (OR=2.08, 95% CI:1.27-3.40, quintile 5 vs. 1). Provider practices may be influencing new diagnoses indicating antipsychotic use at CNHs with high antipsychotic use. It may be important for VA to consider antipsychotic use when contracting with CNHs.

Age-Associated Epigenetic Alterations, Somatic Mutations, and Their Crosstalk in Alzheimer’s Disease

Aging is the major risk factor for Alzheimer’s Disease (AD), and as life expectancy increases, neurodegeneration will continue to afflict an ever-increasing proportion of the population. While numerous theories are attempting to explain the drivers behind AD pathology, what unites them is the observation that AD is reliably associated with a progressive buildup of age-related molecular changes. Because of the varying clinical presentations of AD in patients with similar genetic backgrounds, it has been postulated that epigenetics may be implicated in its etiology. Building on our prior work showing that AD pathology is linked to alterations in age-related DNA CpG methylation (DNAme) across various brain regions, we use state-of-the-art machine learning approaches to identify patterns of molecular damage in postmortem brain samples. We show that alterations in DNAme are associated with accelerated biological aging, AD, and the APOE e4 genotype, which is a major risk factor for AD. We also demonstrate that these associations are present in the PFC but not cerebellum -- in line with the current understanding of AD progression in the brain. Finally, we perform whole-exome sequencing and protein mass spectrometry on the same brain samples to test our hypothesis as to whether AD-associated alterations of DNAme are linked with the accumulation of somatic mutations that affect the structural and binding properties of protein epigenetic regulators.

Epigenetic Trajectories of Aging and Reprogramming

The epigenetic landscape is remodeled with age, bringing about widespread consequences for cell function. With the revolutionary discoveries by Yamanaka and Takahashi, as well as those that built on this work, the transcription factors Oct4, Sox2, KLF4, and C-Myc (OSKM) can be expressed in a variety of cells, including fibroblasts, to make iPSCs. Once cells are reprogrammed, they show an erasure of epigenetic remodeling, suggesting an avenue to reverse aging. It has been recently shown that ectopic expression of three factors, OSK, can restore vision in mouse glaucoma model and reduces epigenetic age. It is not known the path epigenetic remodeling takes or whether all three factors, OSK, are required to remodel the epigenetic landscape. We hypothesize that during reprogramming, cells will reverse along a similar path they took during aging and eventually reverse along that path they took during differentiation. Alternatively, it may also be possible that cells take entirely new paths to reach a state of partial reprogramming or pluripotency. We used DNA methylation and RNA-seq as a multi-omics approach to map the trajectories cells make during aging, differentiation, and reprogramming. In human fibroblasts and hepatocytes, we tested the three-factor OSK mix, as well as pairwise factors OS, OK, and SK and individual Oct4, Sox2, and KLF4 for their effect on cell trajectories. This study provides a dynamic model for epigenetic changes in aging, differentiation, and reprogramming and highlights barriers and bottlenecks throughout the process.

A Computational Solution to Bolster Epigenetic Clock Reliability for Clinical Trials and Longitudinal Tracking

Epigenetic clocks are widely used aging biomarkers, but they are calculated from methylation data for individual CpGs that can be surprisingly unreliable. We report that technical noise causes six major epigenetic clocks to deviate by 3 to 9 years between replicates. We present a novel computational solution: we perform principal component analysis followed by biological age prediction using principal components, extracting shared age-related changes across CpGs while ignoring noise from individual CpGs. Our novel principal-component versions of six clocks show agreement between most technical replicates within 1 year, and increased stability in short- and long-term longitudinal studies. This requires only one additional step compared to traditional clocks, does not require prior knowledge of CpG reliabilities, and can improve the reliability of any existing or future epigenetic biomarker. The extremely high reliability of principal component epigenetic clocks makes them particularly useful for personalized medicine and clinical trials evaluating novel aging interventions.

Deep Learning Methods Capture Non-Linear Brain Aging Patterns Underlying Alzheimer’s Disease and Resilience

The current era of multi-omics data collection has enabled researchers to obtain exceptionally comprehensive profiling of disease subjects. However, exceptionally high dimensionality can ultimately be an obstacle to biological insight. Previously, we presented a method in which penalized regression of methylation principal components reduces noise and improves prediction of age, disease, and Alzheimer’s Disease (AD) pathophysiology. However, strictly linear methods may overly simplify the complex epigenetic aging landscape. We hypothesized that non-linear deep learning methods could identify molecular signatures that better reflect individual resilience to AD. Through the use of an autoencoder to represent high dimensional methylation array data, and supplemental machine learning methods, we connect latent nonlinear representations of the brain to aging, resilience, and indications of AD. In particular, resultant age-predicting representations of methylation were correlated with enrichment of methylation regions and biological pathways. Contextualized within AD pathology, this work provides valuable, ongoing insight into resilience in AD.

Systems aging clock: A novel epigenetic aging clock modeled from organ & bodily function based mortality indices

A diverse array of aging clocks, derived from a variety of omics data and clinical biomarkers, have been developed to describe aging and predict age-related disease. As such, these biomarkers are particularly applicable for use in observational studies, basic science and clinical trials focused on tackling biological aging. However, ongoing research suggests significant heterogeneity in aging, with deterioration and disease occurring in different organ systems or functional domains at various rates across individuals. Existing aging clocks only measure heterogeneity in the degree of aging, not in the manner of aging (e.g. different organ systems or functional domains). We hypothesize these unique trajectories exist and that they can be captured using a systems based approach. In our work, using clinical chemistry biomarkers from participants in the Health and Retirement Study (HRS), Framingham Heart study (FHS) and Women’s Health Initiative (WHI) , we modeled unique epigenetic aging trajectories from distinct groups of biological processes (such as Immune function, metabolic function, hepatic function, cardiac function, renal function and more). Interestingly, these biological system specific scores when combined gave an aging clock with superior mortality prediction than any published aging clock. We further validate the system aging scores and aging clock in different clinical studies to show the added advantage of such a measure, such as the fact that people with similar epigenetic age may have very different system scores. Overall, this method introduces the potential for quantitative and multi-dimensional, personalized aging scores that are indicative of an individual’s disease and disorder risk.

SnapShot: The Regulatory Genome

Enhancers switch genes on and off in response to a variety of intrinsic and external cellular signals. They are the cornerstone of gene regulation and the most pervasive constituents of the regulatory genome. Sequence polymorphisms in enhancer DNAs are a major source of population diversity and predilection to disease. To view this SnapShot, open or download the PDF.

Aging Clocks

While chronological age is arguably the strongest risk factor for death, disease, and disability, same-aged individuals remain heterogeneous in their susceptibilities to these various outcomes. One explanation is that chronological age is an imperfect proxy of the degree of biological aging an individual has undergone. Thus, defining measurable estimates of ‘biological age’ (in contrast to chronological age) has become a major initiative in Geroscience research. Such biomarkers of aging, or ‘aging clocks’ will 1) help identify underlying mechanisms of aging, 2) enable identification of at-risk individuals prior to disease onset, and 3) provide outcomes to assess efficacy of interventions. In this session, I will describe the various aging clocks, how they were developed, and what they track. I will also describe how aging clocks can facilitate research both within and outside of the biological sciences.

A Panel of DNA Methylation and Proteomic Biomarkers for Specific Aging Pathways

Most aging biomarkers such as DNA methylation and proteomic clocks have focused on measuring overall “biological age,” a single number that predicts age-related morbidity and mortality better than absolute chronological age. While intuitive and interpretable, this single biological age number does not account for the possibility that different individuals may preferentially experience aging in different molecular and cellular pathways, and therefore does not suggest personalized aging interventions. We reasoned that a panel of biomarkers each capturing specific aging pathways, such as mitochondrial dysfunction or cellular senescence, may capture the heterogeneity of aging better than existing composite measures. To address this, we employed weighted gene co-expression network analysis to cluster tissue-specific transcriptomes and the serum proteome into specific modules with distinct biological functions and characterized how these modules change with age. We trained DNA methylation proxies of these functional modules that we then applied to independent validation data to identify associations with age-related morbidity and mortality. Clustering analysis using the DNA methylation biomarkers showed that different individuals show distinct patterns of aging. These pathway-specific biomarkers will elucidate how different aging mechanisms interact with each other to produce the larger phenomenon of aging, and for evaluating novel therapeutics targeting specific hallmarks of aging.

CpG Methylation in Aging: Trajectories of Individual Sites

Age-related changes in methylation in a set of genomic CpGs have been shown to form a kind of molecular clocks of aging – DNA methylation (DNAm) clocks. These markers are usually based on a small set of CpGs in every case, but 1) they rarely overlap between different clocks and 2) they are interchangeable, meaning that one can remove all clock sites from a data set and make a new clock of similar precision selecting a new set from the remaining sites. Nonetheless, only a fraction of CpG sites would be suitable for DNAm clocks. We performed an extensive analysis of all CpG sites aging behavior. Previous studies were focused on identifying positions where changes in DNAm correlate with age, but in this case, some of CpGs where DNAm changes occur in a non-linear way can be overlooked. We assessed the aging trajectory of every CpG, clustered CpGs by the type of aging behavior and applied a machine learning approach to construct a new kind of DNAm clocks based on the DNAm of these clusters. Since every cluster is composed of multiple CpGs, it makes this marker resistant to a common problem of missing data. Using blood, brain, skin, colon and liver samples we were also able to investigate tissue specificity of CpGs trajectories.

Epigenetic Signatures of Cell States in Aging

Epigenetic clocks based on DNA methylation (DNAm) show striking age correlations and predict various outcomes. Patterns of DNAm also reflect critical mechanisms in differentiation and proliferation. As such, an outstanding question is whether part of the signal epigenetic clocks are capturing represent shifts in the proportions of somatic stem cells, senescence cells, and/or tumorigenic cells. Here, we assembled various methylation datasets that captured relevant phenomena, including pluripotent stem cells, differentiation, senescence, and cancer, and performed weighted network analysis to cluster and compare DNAm modules. We find overlapping clusters between in vitro samples and in vivo tissue samples, suggesting that cell-level phenomena like cell replication, senescence, and cancer intersect with age-related epigenetic signatures. While the effects of aging manifest at multiple systems levels, from the genome to clinical phenotypes, these analyses may help provide insight to the contribution of cell phenotype dynamics to the general aging phenomenon.

A rat epigenetic clock recapitulates phenotypic aging and co-localizes with heterochromatin

Robust biomarkers of aging have been developed from DNA methylation in humans and more recently, in mice. This study aimed to generate a novel epigenetic clock in rats-a model with unique physical, physiological, and biochemical advantages-by incorporating behavioral data, unsupervised machine learning, and network analysis to identify epigenetic signals that not only track with age, but also relates to phenotypic aging. Reduced representation bisulfite sequencing (RRBS) data was used to train an epigenetic age (DNAmAge) measure in Fischer 344 CDF (F344) rats. This measure correlated with age at (r = 0.93) in an independent sample, and related to physical functioning (p=5.9e-3), after adjusting for age and cell counts. DNAmAge was also found to correlate with age in male C57BL/6 mice (r = 0.79), and was decreased in response to caloric restriction. Our signatures driven by CpGs in intergenic regions that showed substantial overlap with H3K9me3, H3K27me3, and E2F1 transcriptional factor binding.

Genetic associations for two biological age measures point to distinct aging phenotypes

Biological age measures outperform chronological age in predicting various aging outcomes, yet little is known regarding genetic predisposition. We performed genome-wide association scans of two age-adjusted biological age measures (PhenoAgeAcceleration and BioAgeAcceleration), estimated from clinical biochemistry markers ^1,2 in European-descent participants from UK Biobank. The strongest signals were found in the APOE gene, tagged by the two major protein-coding SNPs, PhenoAgeAccel-rs429358 (APOE e4 determinant) (p=1.50 × 10 ^-72 ); BioAgeAccel-rs7412 (APOE e2 determinant) (p=3.16 × 10 ^-60 ). Interestingly, we observed inverse APOE e2 and e4 associations and unique pathway enrichments when comparing the two biological age measures. Genes associated with BioAgeAccel were enriched in lipid related pathways, while genes associated with PhenoAgeAccel showed enrichment for immune system, cell function, and carbohydrate homeostasis pathways, suggesting the two measures capture different aging domains. Our study reaffirms that aging patterns are heterogenous across individuals, and the manner in which a person ages may be partly attributed to genetic predisposition.

A roadmap to build a phenotypic metric of ageing: insights from the Baltimore Longitudinal Study of Aging

Over the past three decades, considerable effort has been dedicated to quantifying the pace of ageing yet identifying the most essential metrics of ageing remains challenging due to lack of comprehensive measurements and heterogeneity of the ageing processes. Most of the previously proposed metrics of ageing have been emerged from cross-sectional associations with chronological age and predictive accuracy of mortality, thus lacking a conceptual model of functional or phenotypic domains. Further, such models may be biased by selective attrition and are unable to address underlying biological constructs contributing to functional markers of age-related decline. Using longitudinal data from the Baltimore Longitudinal Study of Aging (BLSA), we propose a conceptual framework to identify metrics of ageing that may capture the hierarchical and temporal relationships between functional ageing, phenotypic ageing and biological ageing based on four hypothesized domains: body composition, energy regulation, homeostatic mechanisms and neurodegeneration/neuroplasticity. We explored the longitudinal trajectories of key variables within these phenotypes using linear mixed-effects models and more than 10 years of data. Understanding the longitudinal trajectories across these domains in the BLSA provides a reference for researchers, informs future refinement of the phenotypic ageing framework and establishes a solid foundation for future models of biological ageing.

Vasomotor Symptoms and Accelerated Epigenetic Aging in the Women's Health Initiative (WHI)

PURPOSE

The hallmark menopausal symptom, vasomotor symptoms (VMS), has been linked to adverse health indicators. However, the relationship between VMS and biological aging has not been tested. We examined associations between menopausal VMS and biological aging as assessed by 2 DNA methylation-based epigenetic aging indicators previously linked to poor health outcomes.

METHODS

Participants were members of the Women's Health Initiative Observational Study integrative genomics substudy (N = 1206) who had both ovaries and were not taking hormone therapy. Relationships between VMS at enrollment (presence, severity) or VMS timing groups (no VMS: not at menopause onset nor at study enrollment; early VMS: at menopause onset but not at enrollment; persistent VMS: at menopause onset and study enrollment; and late VMS: at enrollment but not at menopause onset) and epigenetic clock indicators predictive of physical aging and early death (DNAm PhenoAge, DNAm GrimAge) were tested in linear regression models adjusting for age, race/ethnicity, hysterectomy, education, body mass index, smoking, and, in additional models, sleep disturbance.

RESULTS

Women were on average 65 years of age at enrollment. Severe hot flashes at enrollment were associated with higher DNAm PhenoAge [relative to no hot flashes: B (SE) = 2.79 (1.27), P = 0.028, multivariable]. Further, late-occurring VMS were associated with both higher DNAm PhenoAge [B (SE) = 2.15 (0.84), P = 0.011] and DNAm GrimAge [B (SE) = 1.09 (0.42), P = 0.010, multivariable] relative to no VMS.

MAIN CONCLUSIONS

Among postmenopausal women, severe or late-occurring VMS were associated with accelerated epigenetic age, controlling for chronological age. Postmenopausal women with severe or late-occurring VMS may have greater underlying epigenetic aging.

Abstract P2-09-02: Evidence of accelerated epigenetic aging of breast tissues in patients with breast cancer compared to women without cancer

Background: Age is the largest risk factor for breast cancer development. Faster biological aging, particularly in breast tissue, may predispose women to higher risk of developing breast cancer. Recently, a set of promising biomarkers of aging, called “epigenetic clocks” (also referred to as “epigenetic age”) have been developed to estimate aging rates in tissues and cells. Using one of the original epigenetic clocks based on DNA methylation patterns, we previously demonstrated that normal breast tissue from luminal breast cancer patients exhibited accelerated aging compared to patients without breast cancer. Here we test whether incorporation of nearly a dozen epigenetic clocks, each capturing slightly different biological aging phenomena, can consistently distinguish normal breast tissue from cancer patients versus healthy controls. We also examined concordance between epigenetic age assessment from paired breast tissue and peripheral blood mononuclear cells.

Methods: We examined 282 specimens including DNA from 94 normal breast tissues (41 from patients with breast cancer, and 53 from non-cancer controls) and 188 blood samples (97 from breast cancer, and 91 from controls). DNA methylation was assessed via the Illumina HumanMethylation450k and HumanMethylationEPIC BeadChip for both tissue and blood samples, and 11 epigenetic clocks were calculated following published methods. Associations of these clocks with breast cancer were examined using logistic regression models and Receiver Operating Characteristics (ROC) curves to determine predictive power in distinguishing samples from cancer patients versus healthy controls.

Results: After accounting for chronological age, we observed increases in most epigenetic clocks for cancer patients compared to normal samples from controls, which indicates accelerated epigenetic aging (e.g., Zhang clock in tissue, 0.55 vs. -0.42; in blood, 0.10 vs. -0.20, all P&lt;0.05). When including age, Zhang clock showed the best predictive power of breast cancer in tissues (Area Under The Curve [AUC] =0.79, 95% Confidence Interval [CI]: 0.69, 0.87) and including all clocks further improved the predictive power (AUC=0.89, 95% CI: 0.83, 0.95). However, all clocks exhibited moderate predictive powers in blood (all AUC&lt;0.7), even when combining them together (AUC=0.70, 95% CI: 0.62, 0.77). In 80 participants with both tissue and blood samples, we observed highest concordance for Hannum clock (bicor=0.43) and lowest concordance for Zhang clock (bicor =-0.19).

Conclusions: Epigenetic aging-based biomarkers distinguish non-tumor breast tissues of patients with breast cancer from breast tissues of healthy controls. These findings raise the possibility that epigenetic markers assessed on needle biopsies of the breast might identify individuals at risk for developing breast cancer.

Citation Format: Erin Hofstatter, Morgan Levine, Zuyun Liu, Tess O'Meara, Disha Dalela, Lajos Pusztai. Evidence of accelerated epigenetic aging of breast tissues in patients with breast cancer compared to women without cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-09-02.

TRAJECTORIES OF PHENOTYPIC MARKERS OF AGING AS PRECURSORS TO FUNCTIONAL CHANGE

Delineating trajectories of aging phenotypes is essential to understanding mechanisms of clinical disease and disability. We investigated longitudinal changes in measures of body composition, energy expenditure, and brain volumes in >900 participants (age 67.0 (IQR: 57-77) years, 48.1% male) of the Baltimore Longitudinal Study of Aging using mixed effects regression models. Computed tomography-derived thigh muscle cross-sectional area declined 754.2 cm2 per decade at age 60 years (p<0.001) and 1294.3 cm2 at 75 years (p<0.001). Energy reserves, defined as a ratio of energy-cost-to-energy-capacity measured using indirect calorimetry, decreased 11.2% per decade at 60 years (p<0.001), and 16.8% at 75 years (p<0.001). MRI-derived measures of total brain volumes declined 41.6 cm3 per decade at 60 years (p<0.001) and 44.9 cm3 at 75 years (p<0.001). Linking these findings to biological and clinical measures of aging may contribute to more accurate assessment of phenotypic age.

Abstract P3-05-01: Age-related methylation signals of breast cancer risk in blood

BACKGROUND: Age is the biggest risk factor for developing breast cancer, which suggests that the biological aging process is a direct driver of cancer etiology. In all normal tissues, DNA methylation status changes systematically with age and is believed to mediate the biological consequences of aging. DNA methylation patterns, commonly referred to as 'Epigenetic Clocks', can be used as measures of aging. However, the chronologic and epigenetic ages can have subtle differences in different individuals. We hypothesize that accelerated epigenetic aging (i.e. DNA methylation indicating an older age than the chronologic age of the individual) is a risk factor for breast cancer development.

METHODS: We used DNA methylation data from blood samples and clinical data from n=2,107 participants in the Women's Health Initiative (WHI) and tested the association between breast cancer risk and two of the most commonly used epigenetic clocks by Horvath (based on 353 CpGs) and Hannum et al. (based on 71 CpGs). DNA methylation in whole blood was measured using the Illumina HumanMethylation450 BeadChip. We used Cox proportional hazard models to assess the association between two epigenetic age predictors (calculated using the algorithms by Horvath and Hannum et al.) and subsequent risk of breast cancer. The model was adjusted for several breast cancer risk factors including: chronologic age at the time of blood sampling, observational vs. clinical trial, clinical trial arm, race/ethnicity, education, BMI, waist-hip ratio, smoking, alcohol, age at menopause, age at menarche, number of pregnancies, age at first birth, previous mastectomy, months breastfed, and cell count estimates. Family history data and BRCA mutation status was incomplete in the WHI and therefore could not be included in our analysis.

RESULTS: Increased epigenetic age determined by the Horvath clock relative to chronological age was associated with increased future incidence of invasive breast cancer, even after adjusting for known risk factors (HR=1.04, P=0.03). Utilizing the Hannum clock, we found no significant association between epigenetic age and breast cancer risk (HR=1.01, P=0.568). When we included both age predictors as independent variables in a single model, the strength of the association between the Horvath epigenetic age and breast cancer risk increased (Horvath HR=1.09, P=6.3e-5; Hannum HR=0.95, p=0.077), such that every one year increase in epigenetic age relative to chronological age was associated with a 9% increased risk of future breast cancer. These results suggest that the aging signal in the Horvath clock that is unique from that captured by Hannum is what drives the specific association with breast cancer.

CONCLUSIONS: Our results support the hypothesis that “accelerated” epigenetic aging measured in the blood increases breast cancer risk. We also demonstrate that the two epigenetic clocks capture different aspects of aging, only some of which have implications for breast cancer risk. Epigenetic clocks may assist in targeting breast cancer screening to higher risk populations in the future, and understanding the biological mechanisms that are altered by the epigenetic changes may lead to new risk reduction strategies.

Citation Format: Hofstatter EW, Levine M, Hatzis C, Pusztai L. Age-related methylation signals of breast cancer risk in blood [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-05-01.

Correction: A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: A cohort study

[This corrects the article DOI: 10.1371/journal.pmed.1002718.].

A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: A cohort study

BACKGROUND

A person's rate of aging has important implications for his/her risk of death and disease; thus, quantifying aging using observable characteristics has important applications for clinical, basic, and observational research. Based on routine clinical chemistry biomarkers, we previously developed a novel aging measure, Phenotypic Age, representing the expected age within the population that corresponds to a person's estimated mortality risk. The aim of this study was to assess its applicability for differentiating risk for a variety of health outcomes within diverse subpopulations that include healthy and unhealthy groups, distinct age groups, and persons with various race/ethnic, socioeconomic, and health behavior characteristics.

METHODS AND FINDINGS

Phenotypic Age was calculated based on a linear combination of chronological age and 9 multi-system clinical chemistry biomarkers in accordance with our previously established method. We also estimated Phenotypic Age Acceleration (PhenoAgeAccel), which represents Phenotypic Age after accounting for chronological age (i.e., whether a person appears older [positive value] or younger [negative value] than expected, physiologically). All analyses were conducted using NHANES IV (1999-2010, an independent sample from that originally used to develop the measure). Our analytic sample consisted of 11,432 adults aged 20-84 years and 185 oldest-old adults top-coded at age 85 years. We observed a total of 1,012 deaths, ascertained over 12.6 years of follow-up (based on National Death Index data through December 31, 2011). Proportional hazard models and receiver operating characteristic curves were used to evaluate all-cause and cause-specific mortality predictions. Overall, participants with more diseases had older Phenotypic Age. For instance, among young adults, those with 1 disease were 0.2 years older phenotypically than disease-free persons, and those with 2 or 3 diseases were about 0.6 years older phenotypically. After adjusting for chronological age and sex, Phenotypic Age was significantly associated with all-cause mortality and cause-specific mortality (with the exception of cerebrovascular disease mortality). Results for all-cause mortality were robust to stratifications by age, race/ethnicity, education, disease count, and health behaviors. Further, Phenotypic Age was associated with mortality among seemingly healthy participants-defined as those who reported being disease-free and who had normal BMI-as well as among oldest-old adults, even after adjustment for disease prevalence. The main limitation of this study was the lack of longitudinal data on Phenotypic Age and disease incidence.

CONCLUSIONS

In a nationally representative US adult population, Phenotypic Age was associated with mortality even after adjusting for chronological age. Overall, this association was robust across different stratifications, particularly by age, disease count, health behaviors, and cause of death. We also observed a strong association between Phenotypic Age and the disease count an individual had. These findings suggest that this new aging measure may serve as a useful tool to facilitate identification of at-risk individuals and evaluation of the efficacy of interventions, and may also facilitate investigation into potential biological mechanisms of aging. Nevertheless, further evaluation in other cohorts is needed.