Pilot Projects

Women exhibit greater age-matched Alzheimer’s disease and related dementias (ADRD) risk, and women and men have distinct biomechanisms of ADRD pathogenesis, justifying research into unique etiological processes in each sex. Despite women’s greater ADRD risk compared to men, women vary in ADRD susceptibility. Therefore, it is crucial to determine why some women have greater risk than others. We are galvanized by pilot results indicating that greater breastfeeding history is associated with diminished ADRD risk in post-menopausal women. Lactation is a sensitive period of plasticity for human mothers as well as other mammals. During this life phase, permanent, re-organizing effects transpire in various physiological systems. Therefore, it is plausible that women’s lactation patterns modify the risks and mechanisms involved in neuropathogenesis across the lifespan. We capitalize on our trans-disciplinary perspective. Our project is cost-efficient in utilizing available data and resources from an NIA-funded, large cohort study. We will analyze data collected from 7,479 post-menopausal women age 65+ who participated in the Women’s Health Initiative (WHI) Memory Study (WHIMS) and from the subset of 2,304 women in the WHI Study of Cognitive Aging (WHISCA). Our three aims reflect three top candidate biomechanisms that could plausibly account for ways in which lactation endows cognitive resilience against ADRD in mothers: improvements in glucose homeostasis, lipid metabolism, and reduced inflammation. We will utilize the existing WHI Specimen Test Results database, which includes ~700 blood-based biomarkers measured at baseline and annual follow-ups. The main activities during the grant period are to identify appropriate biomarkers and composite scores, then conduct mediation modelling in an experimental cohort subset to narrow down the most promising biomechanisms, and finally, fit mediation models in an analytic cohort subset to evaluate the hypotheses in our aims. This effort will involve examination of Apolipoprotein-E levels and APOE haplotypes, which may be implicated in female-specific etiological processes, among many other biomarkers. Ultimately, our research program is working towards developing biomimetic therapeutics to help women avoid ADRD by determining why the natural variability in lactation history endows variability in ADRD risk, and emulate the biological processes we determine to govern this resilience. Our goal is for this pilot study to generate preliminary results that we can leverage for an NIA R01 responsive to the Notice of Special Interest: Sex and Gender Differences in AD/ADRD which invites proposals on life course factors e.g., reproductive history, and sex-specific risk factors, e.g., pregnancy. Our planned R01 will address NIA’s ADRD Research Implementation Milestones 1.I to “identify critical periods of life and critical lifestyle and other parameters with respect to cognitive impairment and dementia prevention” and 1.B to “employ a life-course approach” examining “information on early life exposures/events.” PI’s chances of funding success will be bolstered by NIH Early30 Stage Investigator (ESI) status, as this will be her first R01 application.

Hearing loss is highly prevalent among older adults and is significant associated with adverse outcomes across mental, physical, and cognitive health domains. While the associations between epigenetic aging and agerelated diseases are documented, the link between epigenetic aging and hearing loss remains unexplored. This proposal utilizes data from the Health and Retirement Study to investigate the association between hearing loss and epigenetic aging among older adults in the U.S. Previous research suggests that epigenetic modifications could directly cause auditory hair cell damage and might also be linked to hearing loss through pathways involving cardiovascular diseases and cancer. Conversely, hearing loss could impact epigenetic aging through its effects on psychosocial health. This study aims to assess the association between hearing loss and epigenetic aging, considering various covariates, mediators, and moderators. The expected findings aim to enhance our understanding of the biological and social mechanisms connecting epigenetic aging and hearing loss, illuminating how hearing loss is interconnected with other health outcomes among older adults.

Through the advent of electronic patient portals and recent federal mandates such as the 21st Century Cures Act Final Rule, a majority of patients in the US – including older adults – have access to their biological laboratory results contemporaneously with their physicians.1 For most lab tests, the text accompanying the reported lab value is the same for all patients regardless of age or other sociodemographic and biological clinical factors that can impact one’s risk of poor outcomes associated with a particular lab result. Taking hemoglobin A1c (HbA1c) as an example, a 75-year-old with a HbA1c of 5.7 % and essentially zero risk of developing clinically meaningful diabetes is shown the same text accompanying the lab value as a 30-year-old with a 10% risk of developing diabetes (i.e., the accompanying text in the EHR states that both patients have “prediabetes”). This “one-size fits all” approach is outdated in the age of precision medicine and misses an opportunity to provide patients and providers with personalized information. Numerous models have been developed using a wide array of risk factors to identify non-elderly individuals with prediabetes most likely to develop diabetes.2-6 Tailoring care intensity based on risk level makes sense,7,8 and has encouragingly improved outcomes for young and middle-aged individuals with prediabetes in controlled clinical trial settings.9 In the real world, where patients and their physicians do not have ready access to individualized risk information, older adults with who have essentially zero risk of developing diabetes often go on to receive unnecessary and potentially harmful healthcare10. This unnecessary care includes billions of dollars spent every year on repeat testing that adds little or no value to the patient’s health, 10 and psychological distress arising from a clinically meaningless diagnosis that negatively impacts quality of life. Our goal is to change this status quo so that older adults can see their own personalized risk that integrates their individual epidemiological, medical and biological data. This will allow older adults to engage in truly informed shared decision making about their own healthcare and lifestyle behaviors. As a first step towards this goal, we propose to build and test models that examine the ability of a HbA1c lab result for adults greater than 65 years to predict incident diabetes across age, gender, race and ethnicity using UC Health Clinical Data Warehouse, a repository of universal claims and electronic health records from 6 University of California Health systems since 2012. Building and testing these models directly addresses one of the primary aims of the USC/UCLA CBPH: integrating epidemiological, medical, and biological information into investigation of population health outcomes in order to better understand health differences across socioeconomic, race/ethnic, and demographic subgroups. Completion of these aims will provide critical preliminary data to include in a proposal to NIA to measure the effect of giving patients and their providers the option to know their individual clinical biological risk

Neighborhood social stressors have been linked to an elevated risk of age-related diseases and mortality. As the global population of older adults continues to grow, with the majority preferring to age in their communities, the need to unravel the underlying mechanisms driving neighborhood-based health disparities grows urgent. Epigenetics, biochemical processes responsive to environmental stimuli and exerting systems-level impact on the aging epigenome, may function as a key molecular pathway to adverse health outcomes among residents of high-stress neighborhoods. This study aimed to examine the association of neighborhood stressors (i.e., neighborhood socioeconomic deprivation, low social cohesion, disorder, and crime) with epigenetic aging based on DNA methylation levels across diverse sub-populations. Furthermore, this research identified individual-level social stress (i.e., social isolation, strain and low support, and loneliness) that could explain these disparities. Data is from the Health and Retirement Study, a nationally representative survey of U.S. adults ages 50 and older. This dataset contains information on a wide range of neighborhood stressors, either self-reported or observed, and includes blood samples for assessing epigenetic aging. This project advanced our understanding of the biological underpinnings through which exogenous neighborhood stressors get under the skin to create and maintain health disparities in later life.

The objective was to estimate the effects of social integration on disability trajectories among older adults in two harmonized cohorts of long-term members of Kaiser Permanente Northern California: (1) the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) study and (2) the Study of Health Aging in African Americans (STAR). KHANDLE enrolled 1,712 Asian, Black, Latino, and white adults aged 65 years and older by January 1, 2017 and STAR enrolled 764 Black adults aged 50 years or older by January 1, 2018. Both studies measured social integration at baseline using the Berkman-Syme Social Network Index (score range: 0-6). At each study visit (KHANDLE: 3 waves complete, wave 4 underway; STAR: 3 waves complete) participants were asked questions about mobility limitations (walking a quarter of a mile; waking up 10 steps without resting); functional limitations (stooping, crouching or kneeling; lifting or carrying something as heavy as 10 pounds); activities of daily living (ADL) limitations; and instrumental activities of living (IADL) limitations. This study used generalized estimating equations to estimate the effect of social integration at baseline on the trajectories of the probability of disability in each domain (mobility limitations, functional limitations, ADL, IADL). It adjusted for potential confounders, including sociodemographic and health characteristics. Next, it used gestimation to consider different hypothetical interventions that shift the social integration index by a certain amount (e.g. increase everyone’s index score by 1 or 2 points) and compared the expected average probability of disability in each domain at baseline and year 5 under this hypothetical scenario to the average probability of disability at baseline and year 5 given the observed social integration index distribution. There are stored blood samples collected during KHANDLE wave 1. Biomarker data will become available from these samples in the future, which will provide potential opportunities to explore biological mechanisms linking social integration and disability among older adults after the completion of this pilot project.

This project aimed to develop and validate an A/T/N classification scheme using AD protein biomarkers from blood correlating with cognitive function among older adults in the Health and Retirement Study (HRS) and evaluate transcriptome profile associated with AD protein biomarkers such as ratio of Amyloid beta 42/40 (Aβ42/40) and Glial fibrillary acidic protein (GFAP) [A: amyloid pathology], phospho-tau181 (p-tau181) [T: tau pathology], Neurofilament light chain (NfL) and Glial fibrillary acidic protein (GFAP) [N: Neurodegeneration]. A recent genome-wide association study (GWAS) of brain biochemical phenotypes identified distinct genetic architecture of protein biomarkers of AD. Here, the project proposed to use the E-VAE classifier model to analyze RNAseq data from ~3700 participants in the HRS to classify participants based on A/T/N endotypes of AD using protein biomarkers measured from blood. Utilizing the differential gene expression analysis (conventional linear models) and biological interpretability of the latent embedding from E-VAE classifier model (deep learning model). It characterized the transcriptome signatures associated with amyloid-β plaque (A), tau neurofibrillary tangles (T) and severity of neurodegeneration (N) among the HRS participants. The project validated the association between expression profiles of the identified genes with A/T/N endotypes of AD in an external dataset (e.g.) Long Life Family Study. This novel approach integrating deep learning-based transcriptome signatures of blood biochemical endotype of AD/ADRD advanced our understanding of biochemical subgroups in AD/ADRD and advanced identification of novel drug targets to slow down the progression of AD/ADRD.

A growing body of research suggests historically marginalized students experience cultural mismatch (CM) during the transition to college. The stress from this mismatch negatively disrupts overall wellness, including self-reported mental and physical health, as well as academics. However, our understanding for the capacity of CM to “get under the skin” is quite limited. The purpose of this project was to extend this body of work by examining associations between CM and allostatic load – an index of cumulative burden or “wear and tear” of the body, as well as epigenetic age – biological age based on DNA methylation levels, among historically marginalized students during the transition to college. The project also examined whether perceptions of stress mediate these associations. The project utilized data from a Pilot Transition to College Study of historically marginalized students (N = 94; Mage = 18.00; 78.7% Latinx) conducted at the Health Equity Research and Education (HERE) Center. Results provided an in-depth understanding of the influence of CM on underlying biology during the transition to college. This is important because CM is a major barrier experienced by historically marginalized students during this period in development. If results were successful, they will provide critical, nuanced, scientific information that will fuel larger scale research and grant proposals surrounding the underlying biology linking CM to short and long-term health outcomes.

This pilot is to plan to conduct a pilot study for an eventual national sample in Colombia (COSA). This will be an addition to the HRS sister studies and be a representative sample of adults ages 50 and older to collect preliminary data needed for a planned R01 submission. The objectives of the pilot are to pilot a survey that can be used to clarify the role of early life education and experiences, life course stressors and adversities, social and economic conditions, and health care access and utilization on later life health and well being in a racially and ethnically diverse population with historically low educational attainment. Health measures include assessments of physical performance, cognition, disability, and disease. This pilot used survey instruments that can be harmonized with other studies in the Health and Retirement Study family of international studies. The pilot will be used to establish protocols for instrument development, field work, and data capture and analysis that will be necessary for a larger study. This pilot has been successful in allowing work with a survey firm and investigators in Colombia to design and begin to pilot a national study of aging in Colombia.

Although disease morbidity and mortality risk are elevated among those with sleep disturbances, the specific molecular pathways altered by sleep loss, which impact human disease, are poorly defined, and may be particularly magnified in late life. Potential pathways include inflammatory, oxidative stress, metabolic, and (neuro)-endocrinological pathways, each of which modifies biological aging. No research to date has tested whether treatment of insomnia in older adults modifies metabolic profiles. As insomnia is a modifiable behavioral target with established treatment efficacy, whether this treatment also improves metabolic patterns relevant to health would be a valuable addition to our understanding of how poor sleep drives risk. Cognitive Behavioral Therapy for Insomnia (CBT-I) is an effective intervention to treat insomnia. Our prior trial demonstrated the effectiveness of Tai Chi Chih as an alternative intervention successful at improving sleep among older adults with insomnia. Moreover, we observed improvements in inflammation among CBT-I participants pointing to biological shifts because of treatment. In this proposal we extended this work by investigating additional relevant metabolites for health using novel metabolomic approaches. We hypothesized that treatment of insomnia using either CBT-I or TCC would relate to improvements in concentrations of important metabolites relevant to cardiometabolic health and biological aging, namely improved lipid, amino acid, and oxidative stress profiles. We used existing stored specimens to complete this work, and will use the results to inform future grant applications to support junior investigator, Dr. Kusters.

It is thought that adverse working conditions contribute to premature aging, but it is unclear whether this occurs at the molecular level. The objective of this pilot project was to examine how occupational categorizations and specific occupational exposures are associated with accelerated epigenetic aging, and how they contribute to the earlier aging of disadvantaged social groups. This project utilized data from the Health and Retirement Study’s 2016 Venous Blood Study, complemented with data from the Occupational Information Network (O*NET), to assess how a set of several physical, psychosocial, and environmental occupational characteristics are associated with epigenetic aging as estimated through five epigenetic clocks. The results from this study contributed to our understanding of the biological processes through which working conditions influence health and the aging process.

How biological age may prospectively predict differential classification into trajectories of chronic health conditions (CHCs) associated with greater morbidity and mortality in older adults, and how this association may differ based on experiences of perceived discrimination are still unclear. The objective of the proposed research is to fill this gap using 2014-2020 Health and Retirement Study (HRS) data and biological clock data from the HRS Venous Blood Study. We constructed CHCs trajectories using latent class mixture models and regress these on PhenoAge, an epigenetic clock that measures biological age. We further examined the moderation effects of perceived discrimination on this association in Black, White and Hispanic participants. Exploring differences in vulnerability to health declines by biological age can enhance our understanding of drivers of accelerated health deterioration that have previously been insufficiently identified or quantified. It also provided insights into subgroups of adults who, due to accelerated biological aging and exposure to negative social determinants of health, may benefit from access to social and medical resources, particularly in cases where such resources (ex. Medicare) are reserved for those of a certain chronological age. Through this seed funding, we built a foundation for additional research on biological and social contributors to disparities in adverse health trajectories. Our findings informed the design of a future NIH R01 application focused on understanding factors contributing to differential health declines among diverse adults of similar chronological age and health profiles.

While commercial apps offer brain gaming to keep the brain healthy and improve cognitive functioning, no studies have examined the effectiveness of cognitive training over an extended period of time on a large scale. We proposed to implement an innovative way to engage individuals in regular cognitive training over a long period of time in a population-representative sample. Our specific aims were (1) to recruit members of the Understanding America Study (UAS) to participate in the Online Memory Training Study, an ongoing initiative developed by Prof. Aaron Seitz and Prof. Susanne Jaeggi and funded by the National Institute of Mental Health; (2) to assess adherence to regular cognitive training for gamified and non-gamified tasks and under different incentive schemes; and (3) to evaluate the short- and medium-term effects of online cognitive training on cognitive functions, mental health, and subjective well-being. The results of this proposal drew more generalizable and feasible implications of online brain training and also provided additional scientific evidence on the extent to which online cognitive training has a protective effect against cognitive decline, as well as positive, independent benefits for population health in the form of improved mental health and subjective well-being.

Past research has shown that childhood adversity (CA) affects adult health; however, the biological mechanisms underlying this association are unclear. Though past research has implicated DNA methylation (DNAm), studies utilizing representative data from older adults and reliable DNAm measures are needed to answer key questions about how stable DNAm changes associated with CA are in later life and how much DNAm mediates the association between CA and later life health. This project clarified the association between CA and later life health by 1) generating methylation risk scores (MRSs) for childhood adversity based on an epigenome wide association study conducted in an independent sample and validating that measure in a nationally representative sample of older adults living in the US from the Health and Retirement Study (HRS), including 2016 methylation data from the HRS Innovative Subsample of the Venous Blood Study and 2) assessing how much those MRSs mediate the association between childhood adversity and adult health. This research helped clarify the biological processes underlying the association between CA and adult health.

Health disparities are influenced by social determinants of health. However, there is still limited understanding of how social determinants influence health disparities. More work is needed to uncover the biological mechanisms through which social factors contribute to inequalities in health outcomes. One hypothesis is that social stressors cause epigenomic changes that have negative health consequences. Epigenomic changes refer to chemical modifications to DNA that lead to phenotypic alterations without changing the underlying genetic structure. One of the main epigenic changes is DNA methylation (hereafter, DNAm), a biological process in which methyl groups are added to a DNA molecule, changing gene expression. While it is not fully understood yet how DNAm may accelerate aging, DNAm profiles have been implicated in heart disease, mortality risk, chronic inflammation, and cortisol dysregulation. Recent studies have documented that low SES is associated with DNAm (e.g., McDade et al. 2019). These associations do not establish, however, whether this relationship is causal. There may be other differences between low- and high-SES individuals that confound the relationship. This pilot was to prepare for a larger project that studied whether a poverty alleviation program can affect change in DNAm and slow epigenetic aging. The program, known as a “Graduation program”, helps ultra-poor households in low-income countries start a self-employment activity. The program has consistently shown large, lasting effects in several randomized controlled trials conducted in different countries. A new randomized controlled trial conducted to evaluate the introduction of the program in the Philippines. This trial provided a unique opportunity to study the causal effects of poverty alleviation on DNAm. The larger project collected dried blood spots (DBS) from participants at baseline and one year after the end of the intervention. We used data from DNAm assays to construct epigenetic clocks. The hypothesis is that increases in epigenetic age will be slower in the treatment group than in the control. The pilot served for us to develop a protocol for the collection of the DBS samples, to test the protocol in the field, and to prepare for the blood collection in the Philippines that begun later in 2021 (depending on the pandemic). The pilot project was used to develop background material to support an administrative supplement application to fund the blood data collection at the baseline interview for the RCT which has been discussed in multiple calls with Jon King and John Phillips. The larger project has the potential to make several important contributions. It will test the hypothesis that epigenomic change can occur in response to a poverty alleviation program of two years’ duration. Results may inform the design of policies to reduce health disparities and the development of strategies to increase life expectancy and improve quality of life during old age.

Past research has used epigenetic clocks to connect social adversity to epigenetic pathways of aging across the life course. However, the majority of studies have not examined the long reach of socio-emotional or physical abuse or trauma on epigenetic profiles at older ages and/or have just used parental education or occupation to measure childhood disadvantage. Using the Multi-Ethnic Study of Atherosclerosis (MESA), this proposal examined whether greater exposure to early life adversity (ELA) in childhood was associated with adult epigenetic profiles that are indicative of accelerated biological aging. Specifically, this study examined (1) whether the exposure to greater overall ELA is associated with older epigenetic age as measured by AgeAccel (estimated by epigenetic clock age – chronological age) and whether this association was independent of both childhood and adult SES; (2) whether age of participants at the time of the epigenetic assessment moderated the observed association with older age associated with weaker ELA associations; and (3) whether moderation of the ELA-AgeAccel associations by sex and race/ethnicity revealed stronger associations among women but no difference by race-ethnicity. A better understanding of the relationship between early life adversity and epigenetic aging has lent insight into the specific social and behavioral mechanisms that influence differences in health and disease across older adult populations.

This research aimed to clarify the biopsychosocial pathways through which mental and physical health may be differentially linked across racial groups and stages of adulthood (i.e., young, middle, older adulthood). This has long been a central focus of the CBPH. The overall objective was to evaluate the role of allostatic load in shaping the relationships between mental and physical health among Black and White adults. Using data from the Nashville Stress and Health Study (N=1,252), a community epidemiological study of Black and White adults aged 21-69 in Nashville, Tennessee, this project addressed three specific aims: Aim 1: Examine racial differences in the associations between mental health and (a) allostatic load and (b) physical health, accounting for sociodemographic characteristics, social stress exposure, psychosocial resources, and health behaviors. Aim 2: Evaluate the extent to which allostatic load mediates the association between mental health and physical health among Black and White adults. Aim 3: Assess whether allostatic load moderates the association between mental health and physical health among Black and White adults. Aim 4: Determine whether racial differences in the associations between mental health, physical health, and allostatic load also vary across age groups (i.e., young, middle-aged, and older adults). Results from this project shed new light on the biopsychosocial pathways that may account for the race paradox in mental health. Specifically, this work helps to identify underlying protective factors, which if identified, may be harnessed to improve the physical health of this population. As such, this research informed public health interventions to reduce the burden of chronic physical health conditions and accelerated aging among this population. Findings from this project also supported my National Institute of Aging K01 application, in which I examine racial differences in combined physical-mental health risk across the life course.

In this pilot project, we used RNA transcriptome profiling data collected from three US population health studies conducted at different stages of the adult lifespan to map the role of the multiple distinct molecular signaling pathways that regulate inflammatory gene expression and identified the specific pathways that mediate SES-related health risk. The analyzed studies cover three distinct lifespan phases: 1) young adulthood (represented by mid-30’s leukocyte RNA profiles from the National Study of Adolescent to Adult Health; Add Health), 2) middle age (mid-50’s RNA profiles from the Study of Midlife in the US; MIDUS), and 3) older age (60s-80s RNA profiles from the US Health and Retirement Study; HRS). Integrative analyses of genome-wide RNA profiling data from these three studies addressed four specific aims: Aim 1: Identify which specific pro-inflammatory gene pathways are most strongly associated with low SES. Aim 2: Determine which specific pro-inflammatory gene pathways are most strongly associated with chronic disease risk. Aim 3: Quantify the extent to which multiple distinct pro-inflammatory gene pathways can potentially mediate the overall relationship between SES and disease risk. Aim 4: Determine whether the association between low SES and specific pro-inflammatory gene pathway activity changes systematically over the course of adult development (i.e., young, middle, and older adulthood) or differs by sex. The overarching goal of this project was to provide new insights into the biological pathways through which low SES contributes to health disparities, and identify specific molecular pathways that can be targeted to ameliorate SES-related disparities in health outcomes and enhance both lifespan and healthspan for all members society.

This pilot project linked new contextual data on lead and mercury exposure into the Health & Retirement Study (HRS), examining relationship to HRS DNA methylation data and cognition. It addressed the three specific aims: Aim 1: Construct respondent-specific chemical air pollutant exposure measures. Aim 2: Identify chemical exposures associated with cognitive function. Aim 3: Examine the inflammatory and methylation pathways mediating the above association. Results from the proposed pilot project were used to support an application for R01 funding from NIA. With the data generated from this pilot we were able to show how chemical air pollutants could be used in population-level studies of aging to trace the pathways linking environmental exposures to ADRD outcomes. The pilot results provided key preliminary findings as well as demonstration of feasibility for an R01 proposal. We expect to expand on the work proposed in this pilot by incorporating additional years of chemical exposures in the residential environment, as well as occupational exposures, which will provide a life course perspective on exposure.

Despite a growing body of research linking social adversity to these biological mechanisms of aging, relatively little is known about protective factors that may ameliorate the negative biological effects of adversity. This research addressed these gaps in the literature by investigating associations between chronic stress (e.g., stressful life events, negative work-family spillover, perceived discrimination, spousal strain, family and friend strain) and accumulated daily stress (exposure, response) and biological aging, assessed via gene expression of the DNA damage response (DDR), cell stress, and cellular senescence in a national sample of adults aged 25-74 (the MIDUS Refresher cohort; N = 863). The study also investigated social support (spousal support and joint decision making, family and friend support) as a protective buffer of associations between chronic and accumulated daily stress and markers of biological aging. Importantly, potential differences in stress-related effects on biological aging as a function of age and race/ethnicity were also be explored. This research may have a direct benefit for population health as it aims to identify social and demographic factors that accelerate or slow the progression of aging and contribute to differences in vulnerability for age-related disease.

We proposed to measure family systems-level biological processes in a cohort of parents and their children participating in a field study of cultural influences on health in adolescents and their parents. We examined in particular the degree of family clustering (or co-determination) in a well-described immune cell gene expression profile known as the conserved transcriptional response to adversity (CTRA). The proposed research saught to leverage a unique opportunity to assay and analyze patterns of gene expression for evidence of CTRA in a sample of approximately 120 parent-adolescent dyads who provided dried blood spots (DBS) at three time points, at two-year intervals, when the parents were between 45 and 50 years of age and the youth were between 15 and 21 years of age. The unique opportunity afforded by these data allowed for the pursuit of three specific aims: Aim 1: Estimate the intergenerational association of the CTRA pattern of gene expression between parents and adolescents; Aim 2: Examine how the CTRA pattern of gene expression changes across time for parents and adolescents (including cross-lagged panel analyses to quantify the relative magnitude of parental effects on child and child effects on parent); and, Aim 3: Explore the association of family stressors with the CTRA pattern of gene expression through mechanisms of (a) mutual exposure of the same stressor by parents and adolescents, and (b) spillover of the effects of stressors experienced by one family member to the other family member. By addressing these aims, the overarching goal of the project was to open new avenues by which the impact of family stress on inflammatory-related biology can be assessed in population health studies using family-based sampling and research designs.

The aim of this project was to analyze collected neuroimaging data from the Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD) to investigate how associations between magnetic resonance markers of brain integrity and age and literacy may change in dementia. The specific aims are two. The first aim was to characterize the associations between cognitive function (memory and global cognition) and a variety of structural and functional markers of brain integrity, including brain morphometry (cortical thickness, subcortical volumes, gray matter density), brain microstructure metrics derived from diffusion MRI, and functional connectivity of resting state networks. The second aim was to investigate how those associations between cognitive function and brain reserve change at different stages of the disease (dementia, mild cognitive impairment, healthy), age, and literacy. This study offered a unique opportunity to investigate changes between neural reserve and compensation during the transition between clinical stages, and in particular how literacy and aging in India subjects affect brain resilience to the progression of Alzheimer’s disease. The results helped us further develop the Wave 2 LASI-DAD project to link brain reserve to the progression of cognitive impairment and the onset of dementia, which will increase our understandings about potential strategies to promote resilience against cognitive aging and disease.

A major concern about cognitive decline is that it occurs years before it can be diagnosed as cognitive impairment or dementia. There are no cures for clinically significant cognitive decline, but the existence of modifiable risk factors suggests that if those risk factors are identified and interventions occur, diagnosis may be delayed, which would greatly reduce the cost of dementia and related conditions. This proposal requested funding for pilot data processing and analyses leading to fund of an R01 to test the hypothesis that an index of cognitive decline based on subjective memory impairment (SCI), memory performance, volumetric data from structural imaging, blood biomarkers of metabolism and inflammation, and functional performance can successfully identify individuals at risk for cognitive impairment. It has become clear that multiple pathways lead to SCI including affect, personality, volumes of medial temporal lobe structures, white matter hyperintensities, ApoE status, and blood biomarkers. We computed volumetric measures from previously collected MRI data as well as assays from existing and now frozen blood samples. Data was analyzed for determination of the most effective markers of longitudinal decline in a wide variety of cognitive measures, and possible conversion to poor cognitive function.

Cardiovascular risk factors (CVFs) prior to age 60 are consistently associated with elevated risk for subsequent cognitive decline and Alzheimer’s disease and related dementias (ADRD). Associations after age 60 are less clear, but the literature supports that there is a pattern of accumulating insults over time such that midlife CVFs contribute to both developing cardiovascular disease and subtle cognitive changes; then in later life, CVD confers further damage, accelerating cognitive decline and progression to ADRD. This hypothesis has been referred to as a dynamic pathway of injury. It has yet to be evaluated in an epidemiologic sample, with objective and quantitative measures of CVFs. Of further importance, the hypothesis has not been evaluated for the differential effects of APOE ?4 status and sex. The overall goal of this project was to address these questions in two population-representative samples of adults aged 50 and older. The first aim was to (1a) characterize the dynamic relationship between CVFs and cognitive abilities over time, from middle to later adulthood, evaluate (1b) whether the relationship differs by APOE ?4 status and (1c) between men and women. The second aim was to examine how these relationships predict trajectories of progression to ADRD. The study proposed to use dynamic latent variable modeling that will inform causal interpretations on the changing relationships between CVFs and cognitive changes from middle to later life. To test the hypothesis, this study efficiently leverages the large-scale longitudinal U. S. Health and Retirement Study (HRS) and English Longitudinal Study of Ageing (ELSA), both of which are the only panel studies that comprise the necessary components for testing models for inferring causal relationships: at least three assessment waves on objective biological markers for CVF and on cognitive abilities for the same individuals, genetic data with which to determine APOE ?4 status, and reliable methods for determining cognitive status, or the presence of ADRD. Findings of this project provided insight on the temporality underlying the CVF-cognition pathway, and how differences between APOE ?4 carriers, men and women contribute to progressive damage and ADRD.

Educational attainment and occupational complexity are two predictors of cognitive functioning and dementia risk potentially modifiable by public policies and intervention. For this reason, it is important to understand influences on educational and occupational achievement, including the extent to which relationships reflect shared genetic influences, environmental influences through the rearing family, and environmental influences through school and community. Prior to adulthood, biological parents provide rearing environments correlated with their children’s genotype. In adulthood, educational and occupational achievement are achieved partly through increasingly active transactions between an individual and their environments. Key resources made available just before the transition from adolescence to adulthood could offer insights on mechanisms for supporting life-long cognitive health in an aging population. Additionally, as environmental influences on cognitive functioning may reemerge in late-life, average educational attainment of neighborhood residents may potentially offer protective aging contexts to community-dwelling older adults. Project Talent (PT) participants were originally interviewed in high school (1960) and have been surveyed again in 2014, providing a rich source of information relevant to the experience of growing up and growing old in the United States. The Project Talent Twin and Sibling (PTTS) subsample is particularly suited for conducting genetically-informed studies on aging Americans. Using PTTS, we considered the contributions of specific parental and school resources on complexity of individual’s occupation in adulthood. We did so by examining how these early-life resources mediate the genetic and environmental correlations between educational attainment and occupational complexity. As women tend to live longer than men, resulting in higher prevalence (and possibly incidence) for dementia, we tested for sex differences in mediating effects of resources provided in adolescence, which may be particularly relevant given persisting gender differences in occupational opportunities. This PTTS pilot worked on teasing apart the relationships of parental and school resources, educational attainment, and occupational complexity in the first half of the lifespan relevant to population education and health policies. The results laid the basis for a subsequent application to consider how these influences are related to risk for cognitive impairment and dementia.

This research addressed the gap in knowledge on social adversity and biological aging by extending a larger William T. Grant and NICHD-funded UCLA Families and Health Study conducted by Dr. Theodore Robles. The study is comprised of 78 adults (from 39 couples) with an average age of 43 years (SD = 7.2), and at least one child between 9 and 13 years of age that have not been diagnosed with a chronic medical condition. At study entry, participants completed comprehensive questionnaire and interview assessments of their relationship and stressful life events. During a sophisticated 8-week daily diary protocol, participants completed surveys of their daily stress and couples’ interactions (e.g., amount of conflict and affection with their partner). At the end of the diary period, participants provided a blood sample that we used to assess aging biology. This pilot project added an additional novel biomarker of aging, the epigenetic clock, and conducted new analyses of the existing data to investigate associations between chronic stress, couples’ relationship quality, and several hallmark indicators of aging: telomere length, gene expression of cellular aging pathways (cell senescence marker p16INK4a, DDR, SASP), and the epigenetic clock.

The excessive rates of cardiometabolic conditions among AAs compared to other race/ethnic subgroups is well-documented, and a growing body of evidence has highlighted the role of racism and discrimination in those health disparities. This study centered on the associations between unfair treatment experienced by AAs within select structural domains (employment, criminal justice, education, and housing) and their cardiometabolic risks. This study aimed to identify the structural domains that may be associated with the greatest cardiometabolic risks among AAs, laying the groundwork for future studies to address cardiometabolic disparities from a structural perspective.

Mental processing speed is a fundamental component of many other cognitive abilities. Thus, examination of how and what influences processing speed changes through aging offers critical information on how cognitive function changes through aging. It is becoming increasingly clear that there are sex differences in processing speed, and those differences are greater at specific age ranges, particularly adolescence and older ages. Although, on average, women show faster processing speed while young, they show steeper declines over age compared to men. The variation in processing speed observed at earlier ages may be due to different genetic factors than those that influence change in later ages. Similarly, variation observed across sex may due to different genetic liabilities. Evidence from studies on menopause and cancer suggest that sex hormones may play a role in the sex differences in processing speed in older age. Therefore, there may be a biological pathway involved linking genes to changes in processing speed; however to date, no genome-wide association scan (GWAS) studies identifying genetic variants involved have been published on processing speed. A disadvantage of performing traditional GWAS studies on this phenotype is that although GWAS can identify novel genetic susceptibility loci, if risk is elevated for only a subgroup (e.g., women) or there are opposite genetic effects in different subgroups (e.g., men and women), there will be a weak marginal effect that is unlikely to be detected as significant. The aim of this project was to identify whether there are underlying genetic associations with baseline level and longitudinal change in processing speed performance during aging, and whether variants differ between men and women. We addressed this aim by applying an innovative genomewide-by-environment interaction scan (GWIS) approach that enabled us to efficiently identify susceptibility loci for processing speed that are the same or differentially associated by sex. Further, we used bioinformatics approaches to identify plausible biological mechanisms for findings.

Preclinical indicators of disease such as inflammation, cortisol and glucose dysregulation, and multisystem dysregulation (allostatic load) are related to individual differences in the level of cognitive functioning across adulthood. This study examined whether individual biological systems and allostatic load are related to differential patterns of change in cognitive functioning over 9 years. We also explored whether the relationships between biomarkers and cognitive changes vary by age or educational attainment. We used data from the Midlife in the US (MIDUS) study, focusing on those participants who have biomarkers and cognitive data at the second wave of MIDUS (2005-06) and cognitive data 9 to 10 years later at the third wave (2014-15). The sample includes 863 men and women who ranged in age from 35 to 85 when the biomarker and cognitive data were first collected.   MIDUS biomarkers include a comprehensive range of biological and anthropometric measurements reflecting cardiovascular functioning, glucose metabolism, lipid metabolism, inflammation, HPA axis function, as well as sympathetic and parasympathetic nervous system function. Summary indices have been created to reflect dysregulation in each of these major systems as well as an overall index of multi-system dysregulation, or allostatic load. We used data from seven cognitive tests (immediate and delayed recall, number series, backward digit span, category fluency, speed of processing, and reaction time for attention switching) from the Brief Test of Adult Cognition by Telephone.  The tests are indicators of two factors, episodic memory and executive functioning. Regression analyses tested biomarkers as predictors of cognitive change, while controlling for preexisting diseases and medications. We also tested age, sex, and education as moderators. Identifying biomarkers as antecedents of cognitive changes in midlife and old age, can potentially aid in the early detection of cognitive impairments and increase the possibilities for preventive interventions.

Post-traumatic stress disorder (PTSD) is associated with a two to fourfold increase in the risk of dementia. This likely reflects both direct and indirect effects of PTSD on dementia pathologies.  There is dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis in PTSD, which may cause volume loss in amygdala, hippocampal, and prefrontal brain structures and promote Alzheimer’s Disease (AD) pathogenesis.  PTSD is associated with a number of medical and lifestyle factors that are in themselves dementia risk factors, including hypertension, hyperlipidemia, type II diabetes, coronary artery disease, stroke, metabolic syndrome, chronic kidney disease, depression, substance abuse, obesity, and smoking.  The two processes likely interact, such that chronic stress reduces the brain’s capacity to cope with new insults occurring during dementia processes. Finally there may be shared vulnerabilities for both illnesses, including psychosocial factors such as race/ethnicity, education, and gender. The work described here is a pilot study aimed at identifying those brain regions vulnerable to atrophy in older Veterans with PTSD and dementia, and determining neuropathological, medical, and psychosocial predictors of brain disease. To this end, we conducted a chart review study at the Greater Los Angeles VA Healthcare System (GLA VA). We identified 200 Veterans over the age of 60 who have completed clinical MRI brain scans. Patients were stratified into four groups based on dementia (DEM) and PTSD diagnoses (DEM+/PTSD+; DEM+/PTSD-; DEM-/PTSD+; DEM-/PTSD-). MRI scans were delineated into cortical and subcortical volumes.  White matter hyperintensities (WMH), a biomarker of cerebrovascular disease, was quantified.  FDG-PET scans (as available) were used to create an Alzheimer’s Disease (AD) biomarker. The primary analysis identified which brain regions are atrophic in DEM+/PTSD+ versus all other groups, to elucidate the neural signature of the combined diagnoses (aim 1). Subsequent analyses focus on the dementia groups (DEM+/PTSD+; DEM+/PTSD-), and the primary outcome will be whole brain volume, a marker of overall brain disease. For each group, the association between whole brain volume and dementia biomarkers (WMH and AD; aim 2), and whole brain volume and medical and psychosocial factors (aim 3, exploratory) will be assessed to determine the relative contribution these factors have on brain disease. We hypothesized that patients with PTSD and dementia would show smaller hippocampal and medial prefrontal cortical volume relative to those with dementia but no PTSD. We also hypothesized that DEM+/PTSD+ would show greater WMH burden than DEM+/PTSD-. In DEM+/PTSD+ whole brain volume will be predicted by both cerebrovascular and AD biomarkers, suggesting that while there may be a more prominent role for cerebrovascular disease, both pathologies are related to dementia risk in PTSD. Finally cardiovascular disease, metabolic disease, and lower education would predict smaller whole brain volume in DEM+/PTSD+, suggesting a role for both medical and psychosocial factors in disease expression. This preliminary work provided a first step in clarifying the mechanisms by which PTSD raises dementia risk. Importantly, the work described here supports optimal development of methods to refine the MRI data processing and chart review procedure, as well as preliminary analyses necessary for effect size estimates for a larger grant application. Ultimately this work saught to clarify the mechanisms by which risk factors translate to dementia, identifying priority areas for interventions in individuals with PTSD.

This project used the HRS genetic data and linked air pollution data to study cognitive change. The proposal examined the role of humanin related genes on cognitive outcomes.  Humanin is the first member of a new class of mitochondrial encoded signaling peptides. Humanin has already been implicated in neurodegenerative and related diseases and we have found that its administration in mouse models decreases inflammation and preserves cognitive function during aging. By sequencing the mitochondrial genome and measuring humanin levels in a small group of volunteers (n=150), we have identified a specific SNP (rs2854128) that is associated with lower circulating levels of humanin and importantly, predicts the degree of cognitive decline. Furthermore, this SNP is more prevalent in African-American populations and the associations with this SNP are stronger in this group. ApoE4 is the single strongest genetic risk factor for Alzheimer’s Disease. We discovered an effect of ApoE status on the circulating levels of humanin and believe that some of the effects of ApoE status on cognition could be related to this interaction. ApoE status has also been recently found to modify the role in susceptibility to air pollution and air pollution by itself has been shown to increase inflammation and decrease cognitive function, two processes that are ameliorated by humanin treatment in mice. It is unknown how the humanin SNP interacts with these factors, but it is our hypothesis that this SNP would be predictive of humanin-related phenotypes such as cognitive decline and metabolic dysfunction, and its interaction with both ApoE and air pollution will be associated with a worse outcome, particularly in African-Americans. This pilot study introduced a novel genetic biomarker for age-related phenotypes and also further supported our biological data. We anticipated that this data would be used to submit an R21 or R01 grant further examining the intersection of the mitochondrial-derived peptide SNPs and human health outcomes.

There is a large body of research on the effects of obesity on mortality. Although findings vary and sometimes significantly, there is consensus that obesity but not overweight produce excess adult mortality at adult ages. In some cases obesity appears to have beneficial effects, particularly at very old ages. The mainstay model used in the literature is one where the researcher retrieves estimates of effects of a variable for obesity (or obesity and overweight) with no controls for antecedent illness (a two-state model). This pilot project aimed at showing that the conventional two-state estimates of obesity effects on mortality do not, as a norm, have a clear interpretation. Under some conditions routinely found in limited panel studies, the two-state estimates typically used in the literature are biased and that in all cases comparability of estimates of total effects is jeopardized since the two-state estimates conflate different processes that may be very different in the populations being compared. This study showed that total effects advocated in this body of research conflate direct and indirect effects and tend to underplay the role that obesity may play as a factor that increases adult mortality. This project also aimed at proposing a strategy to minimize the problems of interpretability and comparability of estimates under most conditions generating limited panel data. This study proposed a two-stage estimation procedure to empirically assess the impact of obesity on older adult mortality patterns in U.S., Costa Rica, Mexico and Puerto Rico using available panel data.

The demographic transition in developing countries is occurring more rapidly than in developed countries. For example, while the old-age dependency ratio in the United States is expected to double in 2050, that in Mexico is expected to quadruple. By 2050, the elderly proportion of the population in the United States and Mexico will be similar. Higher rates of poverty among older persons can increase the urgency of policies to address these populations. In Mexico, 29 percent of persons at least 60 years of age live in poverty. To address high poverty in old age, more than 40 low-and-middle-income countries have introduced non-contributory pension programs or cash-transfer programs for the older population. Such programs can boost health as well. Drawing data from a unique randomized controlled trial (RCT) study of a non-contributory pension program for adults over 70 years in the State of Yucatan, Mexico (2008-2012), this project evaluated the impact of the program on glycosylated haemoglobin (HbA1c) and inflammation (CRP). This was a unique opportunity to analyze the effects of income on health using a RCT with biomarkers data. The results of the project were used to submit an R01 describing the impact of this type of program on health and extend the analysis by linking to administrative records (e.g., on other biomarkers, hospitalizations, and emergency visits) of the health-care system.

Using data from the MESA study, the aim of this pilot project was to examine whether an index of allostatic load is related to CVD risk independent of the new American Cardiology Associations’s CVD risk index. The goal was to determine whether allostatic load offers additional information on CVD risks over and above the ACA risk index. Preliminary analyses had been completed and had indicated that allostatic load provides additional prediction, over and above the ACA’s new risk index. Dr. Goldwater is drafting a manuscript based on these findings.

Recent advances in technology offer the potential to gain information from survey respondents on both behaviors and multiple physiological systems that are determinants of health. During the last two years, devices have both come down in cost and increased in capability. We proposed to pilot test a number of new measurement devices, some for the purpose of fine-tuning our assessment of the usefulness of devices with people of varying ages (e.g., testing the usability, acceptability, ability to retrieve information from a new wearable device within the target population along with cost) and others for ability to measure new physiological conditions for future data collection. This informed the research community about feasibility and reliability of these new approaches for use with population representative samples of different ages. Specific aim is to evaluate a set of integrated devices that provide data on 1) sleep, 2) physical activity and movement, 3) heart rate variability, and for some 4) stress. Devices were assessed for validity and reliability, ease of use, ability to capture data, as well as cost. The pilot tested devices among 1) younger volunteers for initial feasibility and validity and to select the best two or three devices, 2) older volunteers with face-to face contact to see how the devices are used by older persons, 3) Finally, with a set of older volunteers with whom we do not have face to face contact but who are sent the devices in the mail and provided instructions on the internet. This approach allowed us to both test the reliability of the validity of devices so that we could recommend the use of specific devices for use on large-scale studies.

Using data from the Midlife in the United States (MIDUS) Biomarker and Daily Diary Projects, we proposed to examine relationships between the dynamics of the hypothalamic pituitary adrenal (HPA) axis and the parasympathetic system, in order to identify common characteristics that are indicative of dysregulation and predictive of poor future health.  In particular, we examined the HPA and parasympathetic systems’ capacity to react (or dynamic range), their actual reactivity to the standardized challenge, and their rate of recovery after the challenge is completed. Our objective was to identify the particular characteristics that are most predictive of altered resting states in regulatory physiological systems. The ultimate objective of this line of work was to create and validate a comprehensive score for physiological dysregulation across multiple systems – allostatic load, which includes not only measures of altered resting levels (which have traditionally been the primary focus for scoring of allostatic load) but also measures of dysregulated system dynamics.

This was the first study we know of to examine inflammation and fatigue in Hispanic colorectal cancer surviving older adults, and also used a novel genome-wide transcriptional profiling methodology. The objective of this study was to understand the dynamics of fatigue and inflammation (genomics-based strategy to identify genes that are differentially expressed in leukocytes) in Hispanic colorectal cancer survivors after their completion of cancer treatment.

Project assessed how adverse social environments affect gene expression in human leukocytes, helping to clarify the biological processes underlying social gradients in health risk. It utilized microarray-based transcriptome profiling to examine differences in leukocyte RNA profiles as a function of education and childhood circumstances in a small pilot sample of 121 HRS respondents. A number of large population studies are now collecting RNA samples; this study was a pilot for use of those data.

Building on an existing longitudinal, social-science research project in Cambodia, this pilot project laid the foundations for a new rider survey on the long-term consequences of nutritional deficiencies in the country in the late-1970s. The pilot project considered 4 tiers of additional data that were collected to this end: text-based questions on personal experiences in the late 1970s, anthropometric data, and health measurements without and with biological samples. Based on the pilot data, we analyze the cost-effectiveness of the different approaches, their potential impact on item-specific non-response rate, the individual loss-to-follow-up and overall refusal to participate, and data quality.

Evaluating cognitive performance by computer, including over the internet, has become increasingly common in research with younger respondents but has experienced much more limited penetration in studies of elders. Our goal was to create a tablet-based assessment that can be used in future studies with Project Talent (PT) participants – a U.S. population based study of individuals born 1942-1946. However, if successful, the potential applications were much broader, within the U.S. for studies of all ages within the U.S. and internationally.

This pilot examined the relationship between measured indicators of metabolic status and socioeconomic status (SES) among community-residing persons aged 45+ in two developing countries, China and India.

This pilot determined whether exposure to adverse local economic conditions, including economic decline, is related to biological risk in U.S. adults, and if differential exposures explain race/ethnic and socioeconomic disparities in cumulative biological risk. This project consisted of both data construction from secondary data sources and analysis using two national surveys: the Health and Retirement Study (HRS) and the Study of Midlife in the United States (MIDUS).

This pilot project assessed whether there are physiological benefits to older adults who participated in an inter-generational program designed to offer them opportunities for meaningful, socially-valued civic engagement while simultaneously embedding substantial doses of physical, cognitive, and social activity into program content as a vehicle for health promotion. The pilot built on prior work documenting cognitive and psychosocial benefits to older adults from program participation. Leveraging the introduction of a program into a demonstration school in Los Angeles, we proposed to evaluate hypothesized changes in biological mechanisms underlying expected physical, cognitive and psychosocial benefits.

This pilot project was longitudinally investigating the impact of cumulative social stressors (perceived stress, stressful life events, lack of social support, perceived discrimination) on level and change in allostatic load (AL) over time among midlife women. Eight years of data from the Study of Women’s Health Across the Nation (SWAN), an ethnically diverse community-based sample of women ages 40-55 at baseline, was analyzed using structural equation models (SEM) and latent growth curve (LGC) analysis (N=2,870). We were interested in identifying the predictive pathways of level and change of cumulative stressors on level and change in AL and in identifying and quantifying the mediating pathways that explain racial and SES differentials. We were also testing whether lower SES and higher cumulative stressors result in worsened AL profiles (‘moderated mediation’). Innovations of this application include: 1. The proposed analysis will be one of the first to test a key theoretical feature of AL, namely that accumulation is driven by exposure to greater social stressors or disadvantage. 2. Assessment of cumulative stressors may more accurately reflect the total burden (or benefits) women have available to them to cope with life stressors. Stressors may also interact in a non-linear fashion. 3. SWAN is one of the few data sets with up to 08 years of annual biomarker and anthropometric assessments which will allow for testing hypotheses pertaining to accumulation of biological risk over time. 4. SWAN biological data are collected over a critical period of women’s life course – as they transition through menopause and age from midlife to early old age – a time during which there are accelerated changes in physiological parameters as well as substantive alterations in social/role circumstances.

Telomeres cap the ends of chromosomes and promote chromosomal stability. Telomere shortening, which tends to occur with advancing chronological age, causes cellular senescence in vitro. Furthermore, a number of studies have found that leukocyte telomere length (LTL) is associated with morbidity and mortality independent of chronological age. Thus, telomere length has been proposed as a biomarker of aging. Although epidemiologic studies of LTL have become increasingly common, few studies include longitudinal measurement of telomere length. Thus, we know very little about (1) the average rate of change in LTL over time, (2) the extent to which the rate of change in LTL varies across individuals, (3) what factors increase or decrease the rate of change in LTL, and (4) whether the rate of change in LTL predicts health-related outcomes. This study addressed the following specific aims: Determine the average rate of change in telomere length over a ten-year period among adults aged 45-84 at baseline; identify sociodemographic, psychosocial, and behavioral predictors of the rate of change in telomere length; and determine whether the rate of change in telomere length is associated with risk factors for cardiovascular disease.

Measures of community-level social, economic, and demographic characteristics are commonly employed in scientific as well as program planning research to characterize the health-relevant qualities of a place, and to compare these features across locales. Measures such as median income, educational attainment, race/ethnic composition, poverty and crime rates, as examples, are often incorporated in multi-level ecological models as environmental factors that contribute to prediction of individuals’ health outcomes. The conceptual basis for using these measures is the social determinants of health model. In practice, however, the measures selected as markers are commonly selected by custom, consensus, or the untested policy of an institution or agency. Indeed, some projects engage community stakeholders in the variable selection process. It is not certain whether markers that are chosen this way permit the most valid judgment of population health significance; capacity for markers to represent a community’s vulnerability or resilience is assumed to be the same across diverse social groups. Little evidence has been offered of their validity for use with specific populations such as the aged. This study will use a combination of official statistical sources including the US decennial census and American Community Survey, in conjunction with recent epidemiological survey data from several studies covering Los Angeles County, to scientifically test equality of relationships between community markers and health outcomes, measured at the individual level, in distinct population groups. The anticipated results would provide an evidence base to recommend measures that best predict population health among old-age community residents, and promote the measures associated most closely with health outcomes in population subgroups defined by age group, race/ethnicity, nativity, primary language, and socioeconomic level. Such findings would more confidently inform epidemiological studies as well as future needs assessment and “healthy community” monitoring.

There is a dearth of information about how the timing of the advance directive (i.e., when it is completed in proximity to death) is associated with the frequency and type of health care transitions. Additionally, little is known about the relationship of an AD to the frequency and types of care transitions. Better understanding can inform research and practice in improving end-of-life care and reducing care setting transitions. The purpose of this pilot study is to better understand the relationship between timing of Advance Directive (AD) completion, stated care preferences, and the frequency and type of transitions made in the last six months of life. We also aim to examine the impact of sociodemographic factors and chronic illness on timing of Advance Directive completion and type and frequency of care setting transitions before and after completion of the Advance Directive.

A growing body of evidence indicates that engagement in generative activity, that is activity designed to contribute to the well-being of others, predicts more favorable trajectories of mental and physical health in later life. These health benefits may accrue, in part, from individuals’ perceptions of playing valuable and useful roles in the lives of others. Such perceptions have been found to predict decreased likelihood of physical disability development, institutionalization and mortality, and better cognitive functioning, in older adulthood. However, to date, efforts to identify the psychosocial and behavioral pathways through which perceptions of generativity and social usefulness are linked to beneficial health outcomes have indicated only minor mediating roles of hypothesized pathways. One potential explanation for the inability to better identify mediating mechanisms may be the temporal characteristics of previous measurements. Previous research in this area has utilized longitudinal survey data to explore associations between generativity perceptions, health outcomes, and potential psychosocial and behavioral mediators that asked participants to recall characteristics of these factors over considerable lengths of time (e.g., last month, last year) and often for variable recall lengths across measurements of different constructs. It may be that more fine-grained temporal measurements can better elucidate the associations between generative activity, generative perceptions, and the cognitive, affective, and behavioral processes that may lead to more favorable health states among the more generative. The proposed pilot study is designed to evaluate an electronic measurement system for capturing multiple daily assessments of generative activity, generative perceptions, and psychosocial and behavioral correlates. Specifically, the pilot study will examine the feasibility of implementing an experience sampling measurement protocol via the use of wireless tablet computers with cellular data connections (for real-time data transfer) in a sample of older adults age 60 years and over. Pilot study objectives are to examine implementation feasibility (e.g., recruitment success, measurement adherence), to use research team and user-participant data to fine-tune the assessment protocol, and to explore whether the length of the measurement period and the number of measurements is sufficient to capture intraindividual and interindividual variation in generative activity, generative perceptions, and cognitive, affective and behavioral factors.

Inflammation plays a pathogenic role in a number of serious illnesses, including many disorders of aging.  However, risk factors for chronic inflammation are still under investigation.  There is growing interest in the role of stress, particularly stress occurring in early life, as a predictor of inflammation and chronic disease later in adulthood, although longitudinal research in this area is extremely limited (1).  Moreover, despite intense interest in the impact of genetic factors on psychological responses to stress (e.g., depression; 2), there has been minimal investigation of how genetic and other factors shape inflammatory stress responses.  This longitudinal study will examine the impact of early stress exposure on genetic markers of inflammation, as well as genetic and psychosocial moderators of these effects, in a community sample of 500 young adults in Australia who have completed detailed stress assessments and provided blood for genetic and immune analysis.  Identification of early life predictors of lifespan health trajectories will enhance our understanding of risk factors for inflammation-related disorders in older age.

Researchers have proposed multiple biological pathways by which socioeconomic status (SES) might influence health. We propose to examine the associations of childhood SES, socioeconomic intergenerational mobility (comparing childhood SES and adult SES) and cumulative SES over the lifecourse, and cumulative biological risk. Moreover, we plan to examine cumulative biological risk over time, the first study to consider a longitudinal assessment of allostatic load. The data will include 6,814 men and women in the Multi-Ethnic Study of Atherosclerosis, a longitudinal population-based study on cardiovascular disease sampled from 6 sites in the United States.  The main outcome, a cumulative measure of biological risk, or “allostatic load,” includes a summary score of high risk indicators representing metabolic, cardiovascular and inflammatory systems (and where available neuroendocrine data on cortisol, norepinephrine and epinephrine). Socioeconomic measures will include childhood SES (parental education) and adult SES (income, education and wealth). Childhood and adult SES measures will be considered simultaneously to assess their independent associations with AL. Moreover, the combination of these levels (i.e., social mobility) will be assessed as well. These models will be repeated with change in AL as the outcome to examine how lifecourse SES may be association with change in AL over time.

The proposed pilot will collect biomarker data to better quantify health risks in an area of Los Angeles characterized by high rates and early morbidity and mortality.Park Mesa Heights (PMH) is a largely African American and Latino community in South Los Angeles, an area characterized by high rates and early morbidity and mortality from obesity, diabetes, hypertension, congestive heart failure, stroke and other chronic conditions.  To understand and address these health disparities, we formed a community-academic partnership, the Healthy Community Neighborhood Initiative (HCNI), which is composed of the Los Angeles Urban League (LAUL), Healthy African American Families (HAAF), Charles Drew University (CDU), and the UCLA Clinical and Translational Science Institute (UCLA, LA Biomedical Institute/Harbor-UCLA Medical Center, Cedars Sinai Medical Center, and CDU).  The study overall has the following specific aims: (1) to design and conduct a household survey and clinical assessment, neighborhood observations, and neighborhood asset map to better understand factors that can be intervened upon to influence health and health outcomes in the Park Mesa Heights community; (2) to develop a chronic condition registry in the community that will be monitored to identify trends in chronic disease incidence, prevalence, and control over time and to develop a registry of biomarkers of stress and inflammation in a community cohort and to link biomarker levels to baseline and longitudinal clinical outcomes; and (3) to engage the community in data analysis, understanding and interpreting and disseminating the results of data collection, and prioritizing issues identified as most important to address with interventions.  The USC-UCLA Biodemography Center pilot study leverages the HCNI infrastructure by building upon and extending the household data collection to support the development of protocols and materials that are specific to the health of middle-aged and older PMH residents.  We propose to include survey and clinical measurements of geriatric syndromes, functional status, and cumulative measures of biologic risk among residents 50 years and older.  In recognition of prior evidence of earlier health deterioration among African Americans and Latinos, measuring these conditions in middle age will allow us to detect earlier onset of biologic “wear and tear” through cumulative measures such as allostatic load, and the implications for functional decline and disability in this community.

Measurement of multiple dimensions of cognitive functioning is now done in the Health and Retirement Study (HRS) as well as other surveys in developed countries.  A number of studies in developing countries are attempting to harmonize with the HRS and related studies but it is not clear that it is possible to export cognitive scales to low education countries without appropriate adaptation. This project will develop scales of cognitive functioning for use in developing countries that are comparable to those in the U.S.  In this pilot, the cognitive scales will be added to the Indonesia Family Life Survey (IFLS). This will be similar to one done in Mexico by Rebecca Wong.  Approaches to measuring cognition will be developed in conjunction with Bob Willis and Jack McArdle.

This project aims to analyze a set of over 2 million genotypes of approximately 4,350 spouse pairs from the Health and Retirement Study (HRS), a representative sample of aging adults in the U.S. Study aims include: (1) Estimating genotypic resemblance in a large sample of spouses to investigate sources of spousal selection; and (2) Estimating the degree to which genotypic resemblance mediates phenotypic resemblance between spouses for alcohol-related measures. The HRS couple data have not been widely used, but provide a unique opportunity to estimate overall genotypic resemblance, study mechanisms of assortment, and estimate the degree to which genotypic resemblance mediates phenotypic resemblance for alcohol use in the U.S. The results will inform other genetic studies by investigating the degree of spousal assortment in the U.S. They can also be applied to a wide variety of traits measured in the HRS.

The human stress response is transduced through two endocrinological pathways: The hypothalamic pituitary adrenal (HPA) axis and the sympathetic adrenal medullary (SAM) axis. The systemic effector of the HPA axis is the steroid hormone cortisol while the systemic effector of the SAM is epinephrine. The peripheral sympathetic nervous system (SNS) predominantly uses norepinephrine as its fast neurotransmitter. The stable end-metabolites of both epinephrine and norepinephrine are 3‑methoxy‑4‑hydroxyphenylglycol (MHPG) and vanillyl mandelic acid (VMA). These are used as indices of NE and E release.

We have established the measurement of cortisol concentrations in hair-shaft samples as retrospective measures of integral secretion of this hormone over the period of hair growth. Thus cortisol secretion can be assessed retrospectively over a period of up to six months. This measure has profound implications for biological psychology research and is providing an empirical measure of allostatic load. The latter is implicated in a number of pathologies including affective disorders, dementia and the metabolic syndrome.

A feature of a number of stress-related diseases is an imbalance between measures of sympathetic activation and measures of HPA function. To assess this we propose retrospectively measuring SNS-activity by assaying the levels of markers of catecholamine release by liquid chromatography coupled to mass spectroscopy in hair samples analogously to the method developed for steroid hormones. These values will be related to measures to long-term cortisol release in a healthy population. The development of such a measure will provide a method of wide applicability to research into long-term stress-responses and their dysregulation in disease.

Thomas McDade is developing new methods for assessing the regulation of inflammation in dried blood spot (DBS) samples. DBS sampling represents a minimally-invasive alternative to venipuncture that has been recently incorporated into a large number of population- and community-based surveys. The validation of new laboratory methods for DBS samples will expand the analytic options for the large number of NIA investigators who are collecting DBS samples, and who are interested in inflammation as a mediator of diseases of aging. The two main aims are to develop and validate immunoassay protocols for quantifying IL-6, TNFa, and IL-10 at low concentrations in DBS samples and to investigate the feasibility of quantifying glucocorticoid receptor in DBS samples.

This project will clarify the biological mechanisms linking late life social isolation and poor sleep to increased risk for morbidity and mortality. One potential mechanistic pathway through which both social isolation and sleep disturbances in late life contribute to disease vulnerability and mortality is through accelerated aging and immunosenescence. The present project aims to address this hypothesis by adding to an ongoing project examining the effects of late life social isolation and sleep deprivation on telomere length and inflammation, an assessment of telomerase (a novel biomarker of telomere maintenance) within peripheral blood mononuclear cells. More specifically, as CD8 T cells and CD14+ monocytes may play a pivotal role in immunosenescence and inflammation, our aim includes assessment of telomerase in these subsets. Aims include: (1) to determine whether telomerase activity will be decreased in the socially isolated older adults compared to the age and gender matched socially integrated adults; (2) to determine whether a brief controlled sleep deprivation experiment alters telomerase activity, providing new insight into how sleep loss may contribute to the functioning of the telomere maintenance system and (3) to undertake a more critical examination of immune cell subsets to establish that differences in telomerase activity are not driven by sleep deprivation induced cell redistribution and also shed light on the degree to which sleep loss influences telomerase activity within cells highly relevant to immunosenescence. This pilot project is designed to support a proposed K01 application in the coming year.

This pilot project will evaluate evidence of sex differences in biomarkers in three Asian countries: China, Taiwan and Indonesia. The objectives are: (1) to examine sex differences in a set of cardiovascular and metabolic, and inflammatory biomarkers and markers of organ damage; (2) to examine the age pattern of sex differences in biomarkers in these countries; and (3) to develop preliminary evidence to support an application for NIH support to compare how behavioral, socioeconomic, and psychological factors are related to sex differences in biomarkers in Asian countries and the U.S. We expect this pilot to result in a submission for an R03.

The overall goals of this pilot project are to use two complementary approaches to explore the genetic determinants of inflammatory pathways in the Tsimane, an indigenous group of Bolivian Amazonian forager-horticulturalists. The first aim is to determine whether single nucleotide polymorphisms (SNPs) that have been previously associated with blood leukocyte counts in Caucasian populations also demonstrate such effects in the Tsimane. The second goal is to carry out large-scale sequencing of inflammatory candidate genes to identify rare genetic variants that may either occur at higher frequency in older Tsimane subjects compared to younger individuals or be unique to this indigenous Bolivian population compared to other populations. In the first set of analyses, nine SNPs previously identified in Caucasian populations were selected and analyzed in ~900 Tsimane individuals for association with eosinophil and neutrophil number. Of these variants, two were not polymorphic in the Tsimane and we obtained high quality genotypes for six others. None of these six variants exhibited significant association with either eosinophil or neutrophil counts.  There are several possibilities for these observations. First, the study may have been underpowered since phenotypic data were only available in ~700 subjects and all of these variants were initially identified in genome-wide association studies with large sample sizes (n>7000). Moreover, the minor allele frequencies of some of these SNPs are low in the Tsimane compared to other populations, which would further reduce power to detect association in our dataset. Alternatively, the Tsimane are chronically exposed to highly infectious/inflammatory conditions and have markedly elevated eosinophil and neutrophil counts compared to individuals living in industrialized countries. As a result, the genetic effects these variants have on leukocyte number in the Tsimane could be masked by such environmental effects. With respect to Aim 2 of the project, the deep sequencing of the interleukin 10 gene will be analyses completed over the remaining months of this pilot project. During this period, data generated from these experiments are being incorporated into manuscripts that will be submitted for publication.

Cross-sectional epidemiologic studies and neurotoxicological data have shown that exposure to air pollution may have deleterious effects on the brain and its functions, but convincing longitudinal studies linking air pollution exposure to cognitive decline in humans are still lacking. The purpose of this research is to link air pollution exposure estimates to two pre-existing datasets (CogNGCS and CogUSA) containing measurements of cognitive performance and demographics, and subsequently explore the adverse neurocognitive effects of air pollution. In CogNGCS (N=1,193), cognitive performance was assessed at two occasions by Woodcock-Johnson (WJ) tests measuring eight of the  broad cognitive abilities from the extended theory of fluid and crystallized intelligence. In CogUSA (N=1,434), a wide variety of cognitive performances were assessed at one time point by WJ tests measuring the same broad cognitive abilities as in CogNGCS, the Wechsler Abbreviated Scale of Intelligence (WASI), and the Health and Retirement Study (HRS) cognitive measures. Air pollution exposure models will be developed for ozone, NO2, PM2.5, PM10, and proximity to traffic. Adding air pollution exposure estimates to these national datasets will allow us to examine the longitudinal intra-individual changes and cross-sectional inter-individual differences in cognitive performances associated with exposure to ambient air pollutants, providing the preliminary data needed to seek R01 funding for future cohort studies.

This pilot involved preparing several datasets for an investigation of the role of life-cycle environmental conditions (such as infectious disease, social network availability, and occupational factors) in long-term trends in cardiovascular disease (CVD) across different cohorts.  While prior work has documented the rapidly increasing epidemic of CVD, particularly from the 1950’s through the 1970’s, trends over the entire century have not been the focus of prior research. Data to be examined include the first cohort to reach age 65 in the twentieth century. Such data offer the first opportunity to examine these longer-run trends in what has been (and continues to be) the major cause of disease and mortality. The project is important for shedding light on changes in heart disease across cohorts and on the etiology of heart disease in different cohorts. The use of this historical data has entailed the classification of 19th and early 20th century descriptions of heart conditions into modern equivalents and this has been completed by an MD experienced in 19th century medical terminology. Hand coding of the data is still underway. The initial data work has been an input into a working paper which investigates the effects of severe prisoner of war stress on cardiac sequelae. This work has found increased mortality from valvular heart disease and stroke but little effects on cardiac morbidity. The work on POW status provides evidence that debilitating events could leave either frailer or more robust survivors, depending on the extent of insults and mortality selection. It is expected that the disease classification scheme will be usable later on.

Seeman took advantage of an opportunity to introduce cognitive assessment into a major national study which has collected significant information on cognitive risk factors but has ignored this outcomes.  The pilot was successful in getting needed materials to implement cognitive assessments which are now on-going in full MESA data collection — data collection is scheduled to go through next November (2011). At that point proposals will be developed and papers written.

This integrative article provides a unique combination of information from basic cell-based experimental studies and linked population-based data from the MacArthur Study of Successful Aging to illustrate the pathways through which individual psychosocial experience (e.g., depression) may trigger sympathetic activity which in turn can interact at the cellular-level with differences in genotype to produce differences in inflammatory responses which are in turn related to differences in mortality risks.

This study aims to evaluate the comparability of blood pressure assessed using an Omron automatic device (used in HRS and NSHAP) with that obtained by standard sphygmomanometer, and to evaluate the comparability of lipid and C-reactive protein measurements assessed using a Cholestech (proposed for use in Costa Rica) device with that obtained from venous sample. Recruitment was stratified by gender and disease status (hypertensive versus not), with 93 participants recruited. For blood pressure, participants were randomized to complete either Omron or traditional sphygmomanometer first. Three measurements from the first device were taken, with a one-minute rest in between measurements. Three measurements with the other device were then taken. Analyses were conducted on the average of the second and third readings from both devices. Distributions from the two devices were similar, but the Omron was more similar to the sphygmomanometer for systolic blood pressure (r=0.89) than with diastolic blood pressure (r=0.67). Of the 93 participants, 49 also participated in the evaluation of the point-of-service equipment manufactured by Cholestech for cholesterol assays. First, blood was collected via capillary puncture for the Cholestech measurement and then a venous sample was drawn for comparison. Comparability of the Cholestech with venous results was quite good (total cholesterol: r=0.82, HDL: r=0.72, LDL: r=0.88, triglycerides: r=0.87, CRP: r=0.92 (analyzed only for samples with detectable venous values). While the results are quite comparable, several features of the Cholestech equipment limit its feasibility for community-based population studies. Manufacturer operating instructions require running control samples each day the unit is used. Both costs for the sample cartridges and the failure of initial control samples on a portion of the days result in high costs. Also, we did not assess the ability of the equipment to withstand high temperatures and humidity. As a result the CRELES study did not pursue their original idea of using the Cholestech and is going with more well-established venous protocols. Findings from this pilot will be incorporated into current work developing web-based materials for investigators interested in making informed decisions regarding protocols for biomarker data collection (e.g., pros and cons of Cholestech equipment will be outlined based on pilot findings).

This pilot is collecting saliva and blood samples for later assays of DNA, RNA, cardiovascular, metabolic and inflammatory risk factors in this long running study of cognitive aging. The resulting data set will be available to researchers. The blood samples are now being collected by the CTSI at the University of California, Irvine. This pilot laid the ground work for a project award as part of the USC Alzheimer Disease Research Center.

What effects acculturation processes have on cardiovascular risk factors in the foreign-born over time spent living in the United States and over generations is unclear and an area of ongoing research. This study investigates the relationship of acculturation to two novel cardiovascular disease risk factors, inflammation and hemostasis,  among Latino and Asian immigrants. Baseline data from the Multiethnic Study of Atherosclerosis (MESA), a rich new data source on older Latinos and Asians, were analyzed using multivariate methods. MESA is unique in having detailed biometric, social and behavioral data on a large number of older Latino and Asian immigrants making this research possible. Analyses for this pilot project are complete and a manuscript is in preparation reporting on major findings – that native-born Hispanics and those who have lived in the US longer exhibit significantly greater burdens of inflammation.

The primary aims of this project are to examine migrant-nonmigrant differentials in cardiovascular risk factors and explore the extent to which these differentials might be associated with differences in psychosocial stress, health behaviors, and access to health care. In this study, data from a 2007 national probability sample of migrants and nonmigrants in China was analyzed.  This project is the first study to integrate psychosocial and biological information in the explorations of potential differentials in cardiovascular risk factors across migrant and non-migrant populations in China. Findings from the study is expected to lead to new RO1 funding to measure biomarkers on stored specimens and to further examine longitudinal relations between migration and health in China. This project is developing the capability of assaying dried blood spots in a lab in Beijing, China, which should be useful infrastructure for later students.

This study investigates the relationship between adult height and health using data from twins in order to control for genetic and nutritional factors. In this pilot study questions on early childhood exposure to infections are linked to adult height. Further it tests the hypothesis that height is related to childhood experience, in a case-control study of healthy identical twins discordant for height by at least 1 inch. The data are based on both members of 213 healthy identical twin pairs born from 1968 to 1982 from the population-based California Twin Program (CTS) who reported a difference of at least 1 inch in height. The answers are validated by interviewing their mother, asking similar questions about their early experience. Questions include information on childhood infections, contact with pets, fecal-oral exposures, day care and school attendance, growth, diet, exercise, and onset of puberty. This project provides another approach to the issue of the effect of inflammation on lifetime health. The analysis of early childhood risk factors for adult height is in progress. The preliminary analysis found that mothers of twins are more consistently reliable historians regarding illness, but their twins know more about each other’s physical development, especially as they approach puberty. Questions comparing members of the twin pair with regard to illness (e.g. Who missed more elementary school, you or your twin?), diet, growth, development and exercise often produce more information than absolute questions (How much elementary school did you miss on average due to illness? How much elementary school did your twin miss on average due to illness?). These data are being used to generate algorithms for using the questionnaire response data from twins and their mothers for the analysis of the main research question.

This pilot examined simultaneously the effects of serum beta-carotene level and inflammation burden on subsequent mortality. It also examined the role of beta-carotene and APOE in affecting cognitive functioning. The hypothesis was that there would be a link between adverse levels of serum beta-carotene level and inflammation factors and deterioration in physical and cognitive functioning among older persons. We also hypothesized that low beta-carotene level and high inflammation burden would independently predict functioning change and there will be a synergistic effect between the two in their association with higher subsequent overall mortality. An additional hypothesis is that the presence of APOE4 alleles would interact with the level of antioxidant status to affect cognitive functioning. The primary finding was that among high functioning older persons, the effect of low serum beta-carotene level on cognitive decline is confined to those who have APOE 4 alleles, suggesting a genetic influence on the effect of low antioxidant status.

The long run purpose of this study was to evaluate the protective effect of education and other indicators of mental ability or “mental exercise” against cognitive decline. The pilot research examined whether education relates to the timing of shifts from nondemented to mild cognitive impairment and from mild cognitive impairment to dementia, and from diagnosed dementia to death. The hypothesis is that people with higher education will drop faster once they start to decline, because their education protected them from crossing a diagnostic threshold for longer than would be the case for someone less endowed. Thus we would expect that having greater cognitive reserve will protect against onset of clinically diagnosed dementia, while decline after diagnosis will be more abrupt. The work of the pilot project was to develop the data from the ADRC files for such examination and test our ability to use the latent growth curve method with this data. This project will help us evaluate the role of changing educational levels in producing trends in population cognitive ability. This project was the basis of a section of a P01 project.

Data for this project came from the MacArthur Study of Successful Aging, a longitudinal cohort of relatively high-functioning women and men aged 70-79 years at baseline in 1988. All standard cardiovascular risk factors including CRP were measured at baseline. Repeat measurements were obtained on all risk factors except for HDL cholesterol and CRP in 1991. Cardiovascular events and cardiovascular mortality information till 1998 has been collected. The baseline measurements of risk factors among the participants without history of previous diabetes, stroke, or heart attacks, were used to construct a composite risk score. This score was internally validated using bootstrapping to compare its risk discrimination ability with the Framingham score. The pilot project used stored plasma samples from the 1991 blood draw to measure 1991 values of CRP and HDL cholesterol. These were used to create change scores reflecting change in the new risk scores between 1988 and 1991, and study the association of these change scores with risk of cardiovascular events and/or cardiovascular mortality between 1991 and 1998.

This project explores the role of early life conditions in increasing risk of dementia and Alzheimer’s disease (AD). It builds on a study showing tooth loss before age 35 was a significant risk factor for AD, even controlling for socioeconomic status of the rearing household. We use a different data set, the National Academy of Sciences-National Research Council (NAS-NRC) Twin Registry of World War II veterans/Duke Twins Study database, to further test the hypothesis. We propose to augment the Duke Twins Study database with medical information from 1940s military records, including dental records. That information is archived on microfiche film. The pilot project allows us to contract with the Institute of Medicine (IOM) to extract a portion of these data and transfer them to a usable format that permits merging with the Duke Twins database. The first specific aim is to determine whether tooth loss, as a marker of early life infection, is associated with risk of cognitive decline and dementia. The second specific aim is to determine whether the association between tooth loss, cognitive decline and dementia is primarily due to its role as an indicator of low socioeconomic status.

A greater proportion of adults reach old age without disability than ever before, yet clinicians continue to struggle with how to approach the substantial proportion (over 10%) of the older population who develop what is increasingly referred to as the geriatric syndrome of “frailty”: a constellation of problems such as weight loss, muscle wasting, and cessation of physical activity, that cannot be explained by identifiable medical or psychiatric diseases. This investigative team proposes to use data from the MacArthur Study of Successful Aging (MSSA), a longitudinal cohort of 1189 high-functioning Americans, to examine whether and to what extent “allostatic load” predicts geriatric frailty across demographic subgroups. Allostatic load explicitly considers the simultaneous and potentially cumulative impacts of physiological effects from multiple regulatory systems before they become clinically manifest as frailty, fitting very well with the theoretical framework of frailty as a cycle of declining reserves across physiologic systems. Proposed analyses offer the potential to identify dysregulatory pathways across one or multiple physiological systems preceding onset of geriatric frailty, thus suggesting possible opportunities to intervene to prevent frailty across the population. This exploratory investigation would provide preliminary data laying the groundwork for an RO1 application to NIA to extend these findings in larger more diverse samples of older adults including the Health ABC study and Hispanic EPESE.

This project combines the assessment of three methods for determining HbA1c levels as used in clinical trials currently conducted by the PI of this proposal, ongoing epidemiologic studies using a portable device that gives rapid feedback, and dried blood spots appropriate for large-scale studies. The results are expected to provide a ‘Rosetta Stone’ for comparing the three methods and for interpretation of results across different studies. Finally, the proposed pilot will provide additional pilot data for the PI to submit a K award to the NIA and other federal and foundation grants that should facilitate the entry of the PI to the field of Aging research and the study of the social determinants of health.

Investigate contextual effects of neighborhood socioeconomic status (SES) on mortality among older adults using the Asset and Health Dynamics Among Older Adults (AHEAD) study of 1993 and 1995. This is done by estimating the magnitude of between-neighborhood variation in mortality, and the extent of which this variation persists net of differences in the social status of the people who live in these different neighborhoods.

This study will follow subjects longitudinally to determine whether rapid decay of iconic memory predicts later development of AD by comparing iconic (visual) sensory memory and attention in older adults with normal cognition to those with mild cognitive impairment (MCI). The hypothesis is that the difference in iconic memory can accurately predict cognitive status. These psychophysical measures will be compared to standard neuropsychological cognitive assessments scores and to changes in those scores over the previous year, to determine whether iconic memory correlates with lower current scores, and declines in scores, on standard cognitive tests.

This project is investigating whether feelings of usefulness to others is associated with the likelihood of disability development and mortality using the MacArthur Study of Successful Aging (MSSA). This study examines depression, physical and social activity, and indicators of low-grade inflammation (IL6 and CRP) as mediators associated with feeling useful.

The purpose of this research is to recover and organize the database from the Hawaii Family Study of Cognition (HFSC) so that these classical data can be used in new longitudinal research to study a variety of contemporary dynamic propositions about aging, health and cognitive abilities.

This project will characterize the distribution of allostatic load (AL) and its consequences for important health markers (vascular stiffness and bone density) among a longitudinal sample of midlife women as well as the impact of demographic factors, personal coping strategies, spirituality, and social support of the accumulation of AL and its consequences on vascular stiffness and bone density using the data from the Study of Women’s Health Across the Nation (SWAN).

The first goal of this pilot project was to provide new evidence on the feasibility of using DBS to measure C-Reactive Protein (CRP), an indicator of acute inflammation and possibly a risk factor for coronary artery disease. The second goal of the project was to provide a careful and comprehensive characterization of the socio-demographic and health factors that are correlated with CRP. The analyses will combine detailed SES and health information in the WISE study, a random assignment treatment-control intervention study in Indonesia, and identify those characteristics of individuals and their families that are predictive of elevated CRP. The results of this pilot study have been the foundation for a research application that included measurement of CRP using the DBS collected from all adult respondents in WISE as well as follow-up interviews three years after the DBS were collected. These measures of CRP will be used to predict subsequent heart disease, other health problems and mortality and will thereby provide unparalleled opportunities to more fully interpret the information contained in CRP in low-income settings.

This project has obtained and prepared a number of existing data sets that have information on markers of inflammation and cardiovascular and metabolic risk factors as well as indicators of physiological capacity such as height and weight, grip strength, stunting, etc. We are investigating these factors across the age range for populations living under very different conditions. We include populations from the United States, Mexico, and the Tsimane of Bolivia. Such populations have very different disease structures, economic status, expenditures of energy, nutritional status, and medical treatment. They have very different levels of both childhood and adult mortality and causes of morbidity. The aim of this project is to develop pilot data that will be used to prepare a program project. The use of physiological indicators from this variety of environments will provide the basis for groundbreaking work in biodemography. This project tests the hypothesis that disease exposure, diet, and intervention experienced in childhood have a lasting effect on the aging process. The effect of the increasing societal development including better public health, diet, and health care has been to delay the aging process. Reductions in exposure to infections and noxious environments in childhood have left cohorts of older persons progressively healthier because they have been exposed to lower lifetime levels of inflammation. Improvements in sanitation and public health have resulted in more resistant hosts through decreases in exposure to bacteria and parasites.

The aim of this project is to determine whether there is a link between telomere length and morbidity among a sample of older persons and to link psychosocial and demographic variables to telomere length. This pilot will provide unique longitudinal data on telomere length (and changes therein over time) for an older cohort, allowing for evaluation of hypothesized relationships between changes in TL and subsequent health outcomes. This study is measuring telomere length of white blood cells for the 900 subjects in the MacArthur Study who have blood samples archived from baseline, around 7.5 years earlier. It examines whether telomere length predicts cardiovascular related diseases and morbidity (from death certificates) and possible relations with allostatic load, which is already linked to cognitive decline, morbidity, or mortality, in the MacArthur Aging Study. As a secondary aim, we will explore how antioxidants and psychosocial factors linked to neuroendocrine reactivity or morbidity, including socio-economic status, depression, and social support, are linked to telomere length. Currently, final DNA samples have been extracted. Samples will be shipped to Dr. Cawthon’s laboratory once funding for laboratory assays and planned analyses of resulting data is available.

This study analyzed data from the National Health Interview Survey (NHIS) with linked vital statistics data to build a descriptive simulation model of mortality among Black, Latino, Asian and White adults. We first calculated all-cause and cause-specific mortality rates. We then examined mortality among Asians for only the more common causes of death given the more limited sample size of this ethnic group in the NHIS dataset. These rates were then used in a simulation model to estimate racial/ethnic differences in years of potential life lost (YPLL). The specific aims of this study were to, 1) Estimate all-cause and cause-specific mortality rates for Whites, Blacks, Hispanics and Asians, standardized to the age and sex distribution of the U.S. population, 2) Conduct a simulation model using the mortality rate estimates in order to project the racial/ethnic differences in YPLL, 3) Calculate mortality rates adjusted for education, years lived in the U.S., obesity and smoking and incorporate these estimates into the simulation model to predict the impact of these variables on racial/ethnic differences in YPLL.

In this study, we found that African-Americans have higher mortality rates from a wide range of causes, however, the disparity is concentrated in a few areas. HIV contributed most to the difference in YPLL (20.0%), followed by homicide (6.9%), hypertension (5.8%) and ischemic heart disease (4.8%). In fact, HIV accounted for a greater proportion of the racial disparity in YPLL than ischemic heart disease, cerebrovascular stroke, hypertension and congestive heart failure combined. Dr. Wong has submitted a proposal for a K award to extend this work in several ways. First, the previous study examined years of life lost over an average follow-up period of 7.4 years. The proposed work will use a simulation model to estimate the life years lost over a lifetime. The simulation model will also allow the estimation the racial disparity in years of life lost due to differences in incidence vs. fatality from specific diseases. Thirdly, the simulation model would serve as a policy tool to estimate the impact of future interventions targeting the elimination of racial disparities in health. Finally, the proposed project will estimate the difference in cause-specific mortality for Latinos and Asians compared to Whites.

Analysis of age-specific mortality rates can provide a measure of the rate of aging and can enhance our understanding of the mechanisms underlying aging and death. The goal of this research is to study the age-specific mortality of populations of billions of long-lived and control yeast. These studies will be performed in parallel with measurements to assess age-dependent deteriorative physiological changes that lead to loss of function and death.

This study found that long-term, multiphasic mortality rates in yeast populations are highly reproducible, both within and across genotypes. Replicate cohorts show striking consistency throughout the lifespan, which reached over 120 days in some instances. The complicated dynamics are not a result of random variation (P<0.0001, P<0.005 for the null hypotheses that the log-mortality patterns are flat or linearly increasing, respectively). The multiphasic mortality patterns may be truly reflective of the underlying biology of the organism, such that physiological changes over time cause the risk of death in individual yeast to fluctuate. Observed mortality patterns might be due to a changing frailty of the individual organisms caused by the byproducts of respiration (e.g., oxidative damage) or by other physiological effects of aging. This hypothesis predicts that other physiological characters, such as stress resistance, would also exhibit multiphasic behavior. The mutations that extend life in a variety of species appear to cause the organisms to be resistant to damage from heat and the corrosive effects of antioxidants. Extended longevity of these simple organisms appears to be associated with increased investment in maintenance and repair, sometimes at the expense of the growth rate. These findings could have important implications for the treatment of diseases. If drugs can be developed that make human cells younger and more resistant to multiple stresses, as in the long-lived organisms, then longevity would be extended and human cells would be protected against a variety of diseases.

 This project funded initial analyses of cognitive data from the multi-site, multi-ethnic Study of Women’s Health Across the Nation (SWAN). This is a study of menopause, focusing on ethnic variations through inclusion of cohorts of Hispanic, African American, Japanese American and Chinese American women, in addition to Caucasians. The pilot analyses explored ethnic variation in cognitive performance and its relationship to differences in educational attainment. These analyses are designed to provide some preliminary, cross-sectional data on which an R01 application was based for additional funding to obtain longitudinal cognitive data for the SWAN cohort (something that was not included in the original planned SWAN data collections). This project provides the basis for requesting longitudinal data on cognition in future waves of SWAN.

While perhaps the most general genetic risk factor currently described for the development of late-onset Alzheimer Disease, the effects of e4 allele of the Apolipoprotein E (APOE) gene on cognitive functioning more generally remain unclear. Cognitive decline in elderly persons is an early predictor of dementia and risks may be modified by a genetic predisposition. Multiple measures of cognitive function were assessed longitudinally in the MacArthur Successful Aging Study, a population-based cohort free of frank impairment at baseline. Subjects were 965 Caucasian and African-American men and women from Durham NC, East Boston MA, and New Haven CT, aged 70-79, and recruited in 1988-1989. Two follow-up evaluations were completed, one at 3 years and another at 7 years.

At the first follow-up, modest but significant declines in naming and spatial ability were associated with APOE-e4 genotype. By the second follow-up, more pronounced and significant associations were noted between APOE-e4 genotype and cognitive decline from six of the eight cognitive outcomes. After 7 years, APOE-e4 allele carriers were twice as likely to have declined on a global cognitive score (OR=2.0; 95%CI: 1.1, 3.6) as compared to non-carriers. This study shows that the APOE-e4 is associated with cognitive decline among a high-functioning elderly cohort with effects most pronounced after 7 years of follow-up. Hence, the e4 allele either may function as a risk factor for cognitive impairment in normal aging across a broad spectrum of domains or, alternatively, may exert detectable effects early in a long prodromal AD trajectory.

To determine whether demographic and health variables differentially interact to predict cognitive scores in Asset and Health Dynamics of the Oldest-old (AHEAD), a nationally representative survey of older Americans to test the developmental discontinuity hypothesis. This research examined cognition in older adults across a variety of measures and used modern scaling methods to identify patterns of cross-sectional decline across measures, to evaluate the role of educational attainment in cross-sectional age differences and in longitudinal decline across cognitive abilities, and to determine whether change in some abilities is more differentially affected by educational attainment than in other abilities. This work used both the Long Beach longitudinal Study and the AHEAD data sets.
Methods: Rasch modeling procedures were used to rescale cognitive measures in AHEAD to calibrate interval scores from raw scores, equating scales across measures, making it possible to compare directly effects of predictors. It also reduces the likelihood of obtaining spurious interactions. Tasks included combined immediate and delayed recall, the Telephone Interview for Cognitive Status (TICS), Series 7, and an overall cognitive score.

Direct comparisons of predictors across Rasch scales showed that demographic variables most strongly predicted performance on all scores, with health variables having smaller effects. Age interacted with both demographic and health variables, but patterns varied across variables. Demographic variables have strong effects on cognition. The developmental discontinuity hypothesis that health variables have stronger effects than demographic ones in cognition in the oldest-old was not supported.

The purpose of this study was to examine whether education and intellectual engagement earlier in life are associated with cognitive impairment later. A co-twin control design was used in order to control for genetic and other familial influences. 167 twin pairs were included in the study in which one member was clinically demented while the twin partner was cognitively intact. These discordant pairs were identified from the Study of Dementia in Swedish Twins and the OCTO-Twin Study, subsamples from the population-based Swedish Twin Registry.

Results from this research have been reported in two articles. The first is an analysis of education. The association between dementia and education was studied in 143 twin pairs discordant for dementia, using a matched pair design, and in 221 dementia cases and 442 unrelated controls from the same twin registry, using a case control design. Low education was defined as 6 years or less of schooling. Case control analyses with prevalent cases showed low education to be a risk for Alzheimer’s disease but not dementia in general. Low education did not significantly predict incident cases. In the matched pairs analysis, which controls for genetic and other familial influences, differences in education between demented twins and twin partners were not statistically significant. However, for Alzheimer’s disease, odds ratios resulting from matched pairs and case control analyses were similar. Twins’ comparative reports about intellectual involvement earlier in their lives suggest a longstanding difference on this dimension, with less involvement by the twin who will become demented. The second study examined whether an active lifestyle affects risk of Alzheimer’s disease: specifically, whether participation in leisure activities during early and middle adulthood was associated with reduced subsequent risk of Alzheimer’s disease and dementia in general. The study sample consisted of 107 same-sex twin pairs discordant for dementia for whom information on leisure activities was self-reported more than 20 years prior to clinical evaluation. A factor analysis of these activities yielded three activity factors: intellectual-cultural, self-improvement and domestic activity. Matched pair analyses compared activities within the discordant twin pairs while controlling for level of education. Odds ratios adjusted for level of education revealed that, within twin pairs, participation in a greater overall number of leisure activities was associated with lower risk of both Alzheimer’s disease (OR=.5, 95% CI=0.3-1.0) and dementia in general (OR=0.6, 95% CI=0.4-1.0). The association between overall activity level and Alzheimer’s disease was significant for women (OR=0.4, 95% CI=0.2-0.8) but not for men (OR=1/1, 95% CI=0.4-3.4). In addition, greater engagement in intellectual –cultural activities was associated with lower risk of Alzheimer’s disease for women (OR=0.4, 95% CI=0.2-1.0), although not for men (OR=0.9, 95% CI=0.3-2.6). Thus, within twin pairs, those who reported participating in more leisure activities than their siblings were less likely to develop Alzheimer’s disease or any type of dementia. Social and intellectual activities may be more important for dementia.