2020 Request for Pilot Proposals

Due Date: Monday, October 12, 2020 @ 5:00pm

The Network on Life Course Health Dynamics and Disparities in 21st Century America (NLCHDD), funded by the National Institute on Aging, invites interested researchers to submit pilot proposals that have potential to better understand how US adult health and mortality outcomes are shaped by meso-level contexts.

The US has poorer population health relative to most other high-income countries in the world and wide socioeconomic, racial/ethnic, and gender disparities across a range of outcomes. Poor health and very wide disparities have been magnified even further by the Covid-19 pandemic. There is also evidence that population health in some US contexts is improving while in other contexts key outcomes are stagnating or even getting worse. The NLCHDD is looking for proposals that promise to advance science in this important area of study and to lead to fundable grant proposals.

NLCHDD 2020 Pilot Proposal Instructions can be found HERE

NLCHDD Pilot Proposals FAQs

FAQs about the Pilot Grants

  • Will proposals be accepted after October 12, 2020?

No, October 12, 2020 is the submission deadline.

  • What is the time deadline for submission of proposals?

Proposals must be submitted prior to 5:00 pm Pacific Time.

  • Will I receive a submission confirmation?

Yes, you will receive a confirmation email.

  • Is it possible to extend the project past 06/30/2021?

If the R24 parent award is extended, you may apply to extend the period performance.

  • When can I apply for an extension for my project?

You may apply for a no-cost-extension 30-40 days before the end of your period of performance.

  • How do I apply for an extension of my project time period?

You can send an email to agingnetwork@maxwell.syr.edu listing the date you would like to extend your project and a brief justification for the extension.

  • Which entity will be the awarding sponsor of the PILOT grants?

Syracuse University will be the awarding entity on the PILOT grants through a grant from NIH/NIA.

  • If I need more funding to finish my project can I ask for more?

No, the funding is limited to the amount awarded.

  • Can I submit more than one proposal?

You can submit only one proposal as a PI. You can submit a second proposal as a Co-I.

  • I applied for a pilot grant in a prior year. Can I apply again?

Yes, you may apply again.

  • I received a pilot grant in a prior year. Can I apply again?

Yes, you may apply again.

Year 6 – Health Declines and Growing Disparities Across States

The Roles of State and Local Immigration Policy in Shaping Spatial and Life Course Patterns of Health Inequality | PI: Courtney Boen, University of Pennsylvania

The Roles of State and Local Immigration Policy in Shaping Spatial and Life Course Patterns of Health Inequality

PI: Courtney Boen, MPH, Ph.D.
Co-Investigators: Robin Ortiz, Atheendar Venkataramani
University of Pennsylvania

Abstract

By 2060, projections estimate that nearly 30 percent of Americans will be Hispanic and 1 in 5 individuals in the U.S. will be foreign born. Given these broader population trends, understanding the sociopolitical drivers of health among U.S.- and foreign-born Hispanics is critical to determining future patterns of health. Existing research on the relative health and well-being of Hispanics offers mixed evidence. Studies typically find that Hispanics have lower infant mortality rates and longer life expectancies but higher rates of diabetes, disability, and biophysiological dysregulation than non-Hispanic Whites. In these ways, Hispanics live long lives compared to other racial-ethnic groups, but they also face tremendous health risk across the life course due to the social contexts in which they are born, live, and age. In particular, over the past two decades, a surge of state-level immigration policies has restricted access immigrant to public benefits, militarized border communities, increased immigrant detentions and deportations, and contributed to heightened levels of xenophobic and racial hostility across the U.S. Emerging scholarship suggests that these policies and accompanying institutional practices contribute to life course and geographic patterns of health and mortality risk. Still, despite a growing body of research in this area, two critical gaps in the literature warrant our attention. First, studies on the health impacts of restrictive immigration policies and enforcement activities typically focus on single state policies or specific immigration raids. As such, there is limited evidence of how temporal and spatial variation in policy and immigration enforcement climates and contexts patterns population health inequality more broadly. Second, little research assesses how immigration policy and enforcement shapes life course patterns of health risk as individuals age. The current project expands understanding of the links between state and local policy, institutional practice, and population health by examining how state-level restrictive immigration policies and state- and county-level immigration enforcement activities shape patterns of health inequality from birth through late life. Linking a variety of nationally-representative individual-level data with longitudinal state- and county-level policy and immigration enforcement data, this project examines how the passage of state-level restrictive immigration laws and changes in state and local immigration enforcement activities contribute to racial-ethnic and immigrant-native disparities in birth outcomes and markers of adult physiological functioning across time and space. Using a variety of analytic techniques—including difference-in-differences modeling and longitudinal fixed effects modeling—findings from this project will provide new evidence of the links between state and local sociopolitical environments and population health inequality. 

State Moderation of Race Disparities in Polygenic Associations | PI: Jason Boardman, University of Colorado Boulder

 State Moderation of Race Disparities in Polygenic Associations

PI: Jason Boardman
University of Colorado Boulder

Abstract

This project will use data from the National Longitudinal Survey of Adolescent to Adult Health (Add Health) to describe state-level differences in genetic associations for health outcomes and health behaviors among members of different racial and ethnic groups. A large body of work has shown that most genetic associations are stronger among non-Hispanic and white adults than individuals with different racial and ethnic identities. The propose study will evaluate the extent to which these differences are consistent across states. Departure from consistency will provide important information regarding this important, yet not evaluated to date, research question in the health-disparities literature. After describing state-level variation in disparities of genetic associations, we will explain the differences across states using state level information on social demographic features, specific health policies. We will specifically focus on indicators of state-level differences in the exposure to discrimination and different forms of institutional racism to add to the existing health-disparities literature by linking these explanatory models to work in geneenvironment interaction research. That is, using genetic information to highlight the social determinants of health with a specific emphasis on disparities by race and ethnicity.

Implementation of the ACA: Releasing former Prisoners from a High Risk Death | PI: Carmen M. Gutierrez, University of North Carolina at Chapel Hill

Implementation of the ACA: Releasing Former Prisoners from a High Risk of Death

PI: Carmen M. Gutierrez
Co-Investigator: Evelyn Patterson

University of North Carolina at Chapel Hill

Abstract 

Former prisoners face an elevated risk of death upon their release from confinement, especially from causes amenable to health care, including drug overdose, untreated chronic health conditions, and suicide. Despite their significant burden of poor health, formerly incarcerated individuals have historically gone without needed medical care outside of prisons, in part due to their overwhelming lack of health insurance. Implementation of the Patient Protection and Affordable Care Act (ACA) significantly increased health insurance coverage among formerly incarcerated individuals. Whether and how the risk of death following release from prison has changed accordingly, however, remains unknown. Leveraging variation across states in the implementation of the ACA, the proposed project will explore whether and how changes in the provision of health insurance affect the risk of death among individuals recently released from prison. By investigating changes in the risk of death among formerly incarcerated individuals before and after ACA implementation across Medicaid expansion and non-expansion states, this project will provide evidence on how government policy interventions can address the relationship between incarceration and mortality, and will raise important questions about how state-level contexts shape the risk of deaths amenable to health care. In light of how incarceration is deeply concentrated among Black and Latino men, results from this project will also have profound implications for our understanding of how health insurance status, incarceration, and state policy environments combine to shape population disparities in health across race/ethnicity and gender.

Exploring the Role of Structural Sexism in Health and Mortality Across U.S. States | PI: Patricia Homan, Florida State University

Exploring the Role of Structural Sexism in Health & Mortality Across U.S. States

PI: Patricia Homan 
Florida State University

Abstract

Influential new research on U.S. health disparities has begun to document the ways that state-level contexts matter for health. Despite persistent gender inequality in the U.S, very few studies have examined the variation in gender equity across U.S. states and how it may be related to health. This project builds on an emerging line of structural sexism research to fill this gap by examining how systematic gender inequality in power and resources manifest in U.S. state-level institutions. Using U.S. state-level administrative data combined with individuals’ state of residence and health data from the Health and Retirement Study (HRS), I will measure structural sexism in state-level political, economic and cultural institutions and document its relationship to health and mortality in later life. First, I will generate multi-state life tables for men and women who live in U.S. states characterized by high vs. low levels of structural sexism in order to estimate healthy, unhealthy, active, disabled and total life expectancies at age 50 as a function of sexism exposure. Next, I will investigate a potential link between structural sexism and biomarkers, thereby building knowledge about how social inequalities get under the skin to affect biological function and long-term health. Finally, I will evaluate the potential of personal resources (particularly educational attainment) to offset the health risks associated with living in a sexist social environment. At a time when sociologists and epidemiologists are looking upstream to understand the larger social forces that influence population health, the proposed research will generate novel evidence that that systematic gender inequality in U.S. states is not only a human rights issue, but also a pressing public health problem.

Unmet Caregiving Need Across State Policy Contexts | PI: Adriana Reyes, Cornell University

Unmet Caregiving Need Across State Policy Contexts

PI: Adriana Reyes 
Cornell University

Abstract

Approximately 18 million older adults have limitations that require caregiving assistance, however about a third of them have needs that are not fully met. Being able to get the care one needs requires either having the resources to pay for formal care or having family to provide unpaid care. Families provide large amounts of unpaid care despite having little support to balance competing demands of family and work. As the population ages, policies and programs will need to better support family caregiving arrangements. Some states have begun to pass policies that support caregivers such as paid family leave. This project has three aims: to investigate state differences in unmet care need, create measures of state differences in support for caregiving, and test the association of state support for caregiving and unmet need for care. It is hypothesized that states with more supportive policies and spending on caregiving support programs will reduce the barrier to providing care and therefore reduce the levels of unmet care in those states. This project will be one of the first to systematically assess how states’ support of caregiving is associated with outcomes for caregiving and caregivers across the contiguous U.S. Data on unmet care will come from the Health and Retirement Study and be merged with newly created measures of states’ support for caregiving. Three aspects of the state policy context will be used to assess states’ support for caregiving: state care-work policies, spending on caregiver support, and the sociopolitical climate. The association between state support for caregiving and unmet care will be tested using multilevel regression. Understanding how different state policy contexts are related to unmet care for older adults will advance research on how best to care for the aging population. Better understanding how to enable family to provide the best care for their family will both improve the health and well-being of older adults and improve the well-being of family caregivers.

Year 5 – Developing and Identifying Geographic Data Resources

Local Contextual Data Resources for Aging Research | PI: Jennifer Ailshire, University of Southern California

Local Contextual Data Resources for Aging Research

Jennifer Ailshire
University of Southern California

Abstract

Since the early 1980s, U.S. adult health and longevity has become increasingly unequal across state and local areas. By the year 2000, the range in life expectancy across U.S. states exceeded the range across comparable high-income countries. The growing inequality reflects gains in the life expectancy in some parts of the country alongside stagnation or actual declines in other parts. Moreover, state and local contexts appear to have a particularly pronounced effect on the health and longevity of adults with low levels of education or income. The geographic divergence in older adult health and longevity has caught the attention of many researchers in recent years. While some progress has been made toward explaining the divergence, the lack of state and local contextual data that can be merged into individual-level survey data of older adults has been a major impediment. This lack of data also researchers interested in investigating the geographic divergence. Some contextual data resources do exist, but they are either local or state level and focus on either environment or policy. Also important, many researchers are unaware of existing contextual data available for data linkages with NIA funded surveys, or other surveys of adult health over the life course. This proposed project aims to address this lack of availability and awareness of state-level contextual data. It will (a) compile a catalogue of local-level contextual data resources that are currently linked with NIA funded studies or could be linked to data from population-based surveys, (b) enhance and existing contextual data resource by adding additional years of measures of several local area characteristics, and (c) publish a coauthored report (with Dr. Jennifer Karas Montez) that promotes the data catalogue and identifies promising directions for research. The proposal submitted here, which is focused on local-level data, is a companion to a separate proposal submitted by Dr. Jennifer Karas Montez focusing on state-level data.

The Geography of Health Inequalities: Contextual Data Merge for Wave V of the National Longitudinal Study of Adolescent to Adult Health | PI: Taylor W. Hargrove, University of North Carolina at Chapel Hill

The Geography of Health Inequalities: Contextual Data Merge for Wave V of the National Longitudinal Study of Adolescent to Adult Health

PI: Taylor W. Hargrove, LM Gaydosh & DW Belsky
University of North Carolina at Chapel Hill

Abstract 

The aim of this project is to develop a contextual database that summarizes the demographic, socioeconomic, health, and mobility characteristics of the environments in which participants from the National Longitudinal Study of Adolescent to Adult Health (Add Health) were living at the time of their Wave V interviews (2016-2018, ages 32-42). Specifically, we propose state-, county-, and tract-level data linkages to characterize levels of and trends in demographic characteristics of the population, chronic disease, health risk behaviors, health care access, economic opportunity, and inequality. Indeed, accumulating evidence indicates that the sociocontextual environments in which individuals are raised and live as adults are significant determinants of life course health and well-being. Yet, missing from the literature is the identification of pathways explicating how sociocontexutual mechanisms shape health and health disparities, particularly in earlier portions of life. Results of this project will thus make possible new research to test how place influences health, behavior, and social outcomes during the transition from adolescence to early adulthood and from early adulthood to midlife.

U.S. State Contextual Data Resources for Aging Research | PI: Jennifer Karas Montez, Syracuse University

U.S. State Contextual Data Resources for Aging Research

PI: Jennifer Karas Montez
Syracuse University

Abstract

Since the early 1980s, U.S. adult health and longevity has become increasingly unequal across state and local areas. By the year 2000, the range in life expectancy across U.S. states exceeded the range across comparable high-income countries. The growing inequality reflects gains in the life expectancy in some parts of the country alongside stagnation or actual declines in other parts. Moreover, state and local contexts appear to have a particularly pronounced effect on the health and longevity of adults with low levels of education or income. The geographic divergence in older adult health and longevity has caught the attention of many researchers in recent years. While some progress has been made toward explaining the divergence, the lack of state and local contextual data that can be merged into individual-level survey data of older adults has been a major impediment. This lack of data also creates significant “start-up” costs for researchers interested in investigating the geographic divergence. Some contextual data resources do exist, but they are either local or state level and focus on either environment or policy. Also important, many researchers are unaware of existing contextual data available for data linkages with NIA-funded surveys, or other surveys of adult health over the life course. This proposed project aims to address this lack of availability and awareness of state-level contextual data. It will (a) compile a catalogue of state-level contextual data resources that are currently linked with NIA funded studies or could be linked to data from other population-based surveys, (b) finish developing a database of time-varying state-level policies and demographics, and (c) publish a coauthored report (with Dr. Jennifer Ailshire) that promotes the data catalogue and identifies promising directions for research. The proposal submitted here, which is focused on state-level data, is a companion to a separate proposal submitted by Dr. Jennifer Ailshire focusing on local-level data.

Year 4 – America’s Worsening Population Health 

Wealth Ownership and Declining Health in the U.S.| PI: Lisa Keister, Duke University

Wealth Ownership and Declining Health in the U.S.

PI: Lisa A. Keister, James W. Woody
Duke University

Abstract

Recent declines in population health in the U.S. are likely to reflect dramatic changes in household wealth that have occurred over the same period. The association between broad measures of household wealth and overall health is well-documented, but the conditions under which the relationship holds are unclear. In particular, we know little about how specific financial states are associated with particular health outcomes and how this association operates for critical subpopulations. These details have the potential to clarify both the broad wealth-health connection and to provide information about the health of subpopulations such as middle- and lower-wealth households, racial/ethnic minorities, older adults, and women. 

 

We have three interrelated aims that will clarify how and when wealth affects health. First, we propose to provide updated estimates of the association between family wealth and health outcomes. We will explore how the association varies across the wealth distribution, by particular assets and debts, and across a large number of specific health outcomes. We will also study how the wealth-health relationship varies by race/ethnicity and gender. Second, we propose to study how the association between wealth and health varies over time including over the life course and as a result of the 2007-09 recession. We will focus on the potential protective effects of saved assets on health outcomes following retirement, and we will examine whether these patterns vary by wealth status, race/ethnicity, and gender. Third, we propose to explore whether family networks mediate the relationship between wealth and health. Our proposed research is particularly innovative because we will conceive of families as social networks, and we will use cutting-edge methods of social network analysis to study the dynamics of the wealth-health relationship and to disentangle the potential protective effects of family. We will study whether the structure, density, and dynamics of family networks ameliorate the relationship between wealth and health; and we will explore whether the role of family networks varies by wealth status, race/ethnicity, and gender. We will also examine these relationships over the life course and over time. 

 

We propose using three survey datasets to study these processes. The Panel Study of Income Dynamics, the Survey of Income and Program Participation, and the Survey of Consumer Finances each include data about both wealth and health, and using them together will allow us to fully explore our specific aims. We will use various multivariate modeling techniques to study each relationship, including using quantile regression to study differences across the wealth distribution, growth models to study changes over time, and social network analysis to study the role of family networks.

Community-level Determinants of Suicidal Behavior: Preliminary Evidence from California | PI: Claire Margerison-Zilko, Michigan State University

Community-level Determinants of Suicidal Behavior: Preliminary Evidence from California

PI: Claire Margerison-Zilko
Michigan State University

Abstract 

Despite steady declines in mortality in the U.S. and other developed nations over the past century, recent evidence demonstrates a reversal of this trend for middle-aged white Americans. In these individuals, mortality has actually increased over the past 20 years. Of particular concern is evidence that this rising mortality is partially attributable to climbing rates of suicide. Reducing these rising rates will require more attention to modifiable, community-level factors. We propose to examine 3 community-level factors thought to influence suicide: rurality, economic insecurity, and community-level violence. Using the California Health Interview Survey (CHIS), we will examine relations between these 3 factors and self- reported nonfatal suicidal behavior, the most potent risk factor for subsequent suicide. We will also explore variation in the relationships between community-level factors and suicidal behavior by age, racefethnicity, and gender. This work will provide preliminary data for an ROl proposal examining relations between community-level factors and suicide fatality in the entire U.S.

Hard Times or a Long Time Coming? Examining Widening Inequalities in U.S. Adult Mortality | PI: Ryan Masters, University of Colorado Boulder

Hard Times for a Long Time Coming?  Examining Widening Inequalities in U.S. Adult Mortality, 1990-2015

PI: Ryan Masters, Ph.D.
University of Colorado Boulder

Abstract

Life expectancy for the U.S. population has declined for the first time in several decades, and rapid mortality increases among middle-aged Americans have alarmed researchers, health practitioners, and policy-makers. Moreover, U.S. health and mortality differences are widening between states and along rural-urban lines. Leading explanations for rising mortality among U.S. adults suggest that economic distress is driving increases in deaths from drug addiction, alcohol abuse, and self-inflicted harm. Increases in these “despair” deaths are thought to be especially high in rural white America (e.g., in pockets of the “Rust Belt,” the “Stroke Belt,” and Appalachia). However, the underlying causes of recent increases in U.S. mortality are unclear, and no study has fully investigated how cause-specific mortality trends are affecting Americans in different ways. The general aim of this project is to examine how changes in specific causes of death are driving spatial differences in U.S. adult mortality trends. This project aims to identify how U.S. mortality trends differ (a) by geographic area, (b) by cause of death, (c) by time periods and birth cohorts, and (d) by race/ethnicity and gender. The newly established Rocky Mountain Research Data Center at CU Boulder will be used to link official county death records from the National Vital Statistics System with official county population estimates from the U.S. Census Bureau. Annual population counts and deaths will be matched for every U.S. county by single-year age, by sex, and by race/ethnicity. Annual age-specific mortality rates will be estimated for male and female U.S. adult populations in every state and county between 1990 and 2015. Changes in between-state and within-state differences in all-cause and cause-specific mortality rates will be documented across this time period, thereby revealing where and how inequalities in U.S. mortality have grown in recent years. How U.S. mortality trends are associated with various county-level characteristics (e.g., median income, migration, educational attainment) will also be examined. 

The Contribution of Diabetes to Trends in Life Expectancy in the United States | PI: Andrew Stokes, Boston University School of Public Health

The Contribution of Diabetes to Trends in Life Expectancy in the United States

PI: Andrew Stokes, Ph.D.
Boston University School of Public Health

Abstract

This project focuses on the impact of diabetes on levels, trends, and differentials in life expectancy by sex and race/ethnicity in the United States. The prevalence of diabetes in the US has risen rapidly. The age-standardized prevalence of diabetes increased three-fold between cohorts born in 1920-29 and 1970-79. In 2011, the prevalence of diabetes at age 20 and above, when measured by Hemoglobin A1c, fasting plasma glucose or 2-hour plasma glucose level, was estimated to be 14.3%. Racial/ethnic disparities in diabetes were substantial; the age standardized prevalence was significantly higher among non-Hispanic Blacks (21.8%) and Hispanics (22.6%) than among non-Hispanic Whites (11.3%). The most commonly-cited estimator of the contribution of diabetes to American mortality is the frequency of its appearance on the death certificates as the underlying cause of death. However, the frequency with which diabetes is listed on the death certificate as the underlying cause of death is not a reliable indicator of its actual contribution to the national mortality profile and is likely to substantially underestimate its importance. In this project, we take a different approach to estimating the contribution of diabetes to US mortality levels, trends and differentials by using a nationally representative cohort—the National Health Interview Survey—to identify the excess mortality risk among people with diabetes. That excess risk will be used in combination with the prevalence of diabetes among deaths to estimate the percentage of deaths (population attributable fraction—PAF) that would not have occurred in the absence of diabetes. We will also estimate the contribution of this excess to US life expectancy and to disparities in life expectancy by sex and race/ethnicity. In addition, we will investigate the contribution of rising levels of diabetes and changes in diabetes-related fatality to changes in US life expectancy by sex and race/ethnicity between 1988 and 2011. Several recent articles on American mortality trends have noted a deterioration of death rates for white Americans since 1999. Using estimates of the risks associated with diabetes combined with changes in the prevalence of diabetes, we can evaluate the contribution of diabetes to age-specific mortality trends more precisely than has been the case in prior studies.

Year 3 – Socio-economic Differences in Health 

Building a Data Resource to Test for Increasing Selectivity of Persons at the Low End of the Educational Continuum: Harmonizing Measures in the National Longitudinal Study of Adolescent to Adult Health and the Health and Retirement Study | PI: Daniel Belsky, Duke University School of Medicine

Building a Data Resource to Test for Increasing Selectivity of Persons at the Low End of the Educational Continuum: Harmonizing Measures in the National Longitudinal Study of Adolescent to Adult Health and the Health and Retirement Study

Daniel Belsky 
Duke University School of Medicine

Abstract

The broad aim of this pilot study is to develop a data resource to test for evidence of changing environmental and genetic influences on selection into the lowest-achieving population segment in terms of education, i.e. those who do not complete a high school degree. Specifically, we will test two versions of the hypothesis that individuals with less than a high school education in more recent birth cohorts come from increasingly disadvantaged backgrounds as compared to educational peers from earlier-born cohorts. This work aims to answer the question outlined in the request for pilot proposals “How is the changing composition of educational attainment influencing the selectivity of persons at the low end of the educational continuum?” The work will make use of sociodemographic, behavioral, ecological, and genetic data from the National Longitudinal Study of Adolescent to Adult Health and the US Health and Retirement Study. Research activities will include the development and analysis of a harmonized data resource to profile environmental and genetic risks for low educational attainment, with the goal of comparing burdens of these risks in birth cohorts born from early to late in the 20th Century. We will test if the least-educated members of more recently-born cohorts show higher concentrations of genetic and environmental risks as compared to educational peers from earlier-born cohorts. The ultimate objective of this pilot research is to build a database and make preliminary observations of it that can serve as the foundation for a larger program of research into pathways connecting educational attainment with healthy aging across the lifespan.

Assessing the Importance of Changing Educational Selection for Education-Mortality Trends in the U.S. | PI: Jennifer Dowd, City University of New York

Assessing the Importance of Changing Educational Selection for Education-Mortality Trends in the U.S.

PI: Jennifer Beam Dowd, Ph.D., Amar Hamoudi, Ph.D.
City University of New York

Abstract 

There is an urgent need for research to improve understanding of the social dynamics underlying observed increases in mortality among Americans with low educational attainment. Observed trends combine two important dynamics—a) contemporary increases in health inequality and b) the legacy of declines in educational inequality from the mid-20th century. The lagged relationship between educational expansion and mortality gradients makes it impossible to empirically disentangle these two processes using only contemporary data. Using historical education and mortality data from the U.S., we have previously shown that observed recent changes in the educational gradients in mortality among white women in the U.S. could reflect large but stable social inequalities in health, declining social inequalities in education, and increasing longevity for everyone(5). While this analysis raised important issues about changing selection into education, much more evidence is needed to fully understand the implications for education-mortality dynamics in the U.S. Furthermore, while much of the existing literature has focused on trends in those in the lowest education category (less than a high school education), the implications for selection dynamics in the higher education categories are less clear and remain unexplored. We propose to extend our prior analysis across the full spectrum of race/sex and educational subgroups, to bring in information about the social history of educational expansions in the U.S., and to investigate the implications of a range of underlying assumptions where information is incomplete. The project will lay the groundwork for a larger project that brings in richer detail on historical patterns in educational access, and variation in that access by race, geography, and occupation and other socioeconomic characteristics in order to refine the picture on educational and social gradients in mortality.

What is the Relationship between Chronic Pain and Death? | PI: Hanna Grol-Prokopczyk, University of Buffalo, SUNY

What is the Relationship Between Chronic Pain and Death?

PI: Hanna Grol-Prokopczyk
University at Buffalo, SUNY

Abstract

This research explores whether the strong, positive association between chronic pain and death is causal, and therefore whether chronic pain contributes to socioeconomic disparities not only in quality but in quantity of life (as suggested by Case and Deaton 2015). That is, does chronic pain actually increase mortality (by reducing physical activity, raising risk of depression, leading to deleterious use of opioid analgesics, etc.), or does it simply reflect the fact that many health conditions both cause pain and increase risk of death? If chronic pain does reduce life expectancy, by what specific mechanisms does it do so (and in particular, are opioid analgesics implicated)? These questions are addressed via secondary data analysis of 16 years of biennial Health and Retirement Study data (1998-2016), including the Prescription Drug Study (2005 and 2007) and Health and Well-Being Study (2009). 

This research is timely, as it examines the intersection of three recent trends in U.S. health: rising rates of chronic pain, rising rates of opioid use and misuse, and troubling slow-downs or even reversals in mortality gains among some population subgroups. Given the high prevalence of chronic pain in the U.S.—and the striking socioeconomic gradient in its distribution—understanding precisely why chronic pain predicts death will have implications for strategies to improve population health and reduce inequalities therein. This project also may clarify why American morbidity and mortality patterns differ from those of most other high- income countries, as the dominant treatment regime for chronic pain during the period of study relied far more heavily on opioid analgesics in the U.S. than in any other country.

State Variation in Socioeconomic Disparities in Health | PI: Melissa Martinson, University of Washington

State Variation in Socioeconomic Disparities in Health

PI: Melissa L. Martinson, Ph.D.
University of Southern California

Abstract

Socioeconomic inequality has been called the ‘defining challenge of our time’ and has risen steadily since the 1970s. During this same period, health inequalities in the United States (U.S.) have also been on the rise and the disadvantaged population health of Americans relative to our peers has gained increasing attention. Within the U.S., income inequality varies by state as much as much as it varies between the U.S. and OECD countries such as France with significantly lower inequality, and the variation in poverty levels across states is similarly dramatic. States within the United States also vary in the quality of resources provided to low- income populations through social policies. The theory of fundamental causes, wherein income influences health through access to resources and the mechanisms shaping health inequalities adapt over time, suggests that health inequalities will be present as long as societal inequality is present. Therefore, states with higher levels of societal inequality and poverty should have greater health disparities, and the level of social protections and redistribution through social welfare policy has the potential to dampen this association. This study uses restricted geographic data and objective exam- and lab-based health measures of morbidity for adults age 25 to 65 from the National Health and Nutrition Examination Survey (NHANES) to address the following questions (1) assess the extent to which income gradients in objective measures of health vary by state and (2) examine the association between state-level socioeconomic inequality and individual-level socioeconomic disparities in objective health. This research project will establish important facts that need to be known about socioeconomic disparities by state to move the field forward in preparation for the next steps in the analysis. The aims associated with this pilot grant will set the foundation for a R01 grant application within the next 2 years, where I can examine a broader range of state-level variables, trends over time, age patterns, and gender differences in stratification.

Educational Pathways and Smoking among US Young Adults: A Cohort Comparison | PI: Katrina Walsemann, University of South Carolina

Educational Pathways and Smoking Among US Young ADults; A Cohort Comparison

PI: Katrina M. Walseman, Ph.D.
University of South Carolina

Abstract

Educational attainment is a very strong predictor of smoking behavior in the United States, where tobacco use remains the leading behavioral cause of premature mortality. Over the past four decades, tobacco use has become heavily concentrated among less educated persons. At the same time, there has been an increase in the prevalence of non-normative educational pathways – that is, young adults today are taking longer to attain their degree, and many of them never attain a degree after enrolling in college. The proposed study will take advantage of these two historical trends by examining cohort differences in the relationship between educational pathways and daily smoking in early adulthood. The specific aims of this project are to: (1) describe the educational pathways of two cohorts of young adults and the compositional changes in educational pathways between these two cohorts; (2) compare the association between educational pathways and daily smoking in early adulthood by cohort; and (3) examine the extent to which economic resources, occupational status, and marriage explain the relationship between educational pathways and daily smoking by cohort. The proposed study is significant because it will elaborate the ways in which education affects one of the most important behavioral causes of morbidity and mortality – cigarette smoking – across two successive birth cohorts in the US.

Year 2 – Racial-Ethnic Disparities in Health 

Racial Health Disparities from Aging in Changing Places| PI: Michael Bader, American University

Racial Health Disparities from Aging in Changing Places

PI: Michael Bader, Ph.D., Jennifer Ailshire, Ph.D.
American University

Abstract

The proposed research identifies how neighborhood change affects racial health disparities among older adults in the United States. Although older Americans prefer aging in place and policy makers increasingly support programs to make that possible, little research investigates how aging in place might exacerbate racial health disparities. Evidence from children and their caregivers demonstrates that Blacks and Latinos live in neighborhoods experiencing worsening economic conditions and increasing racial isolation, but little research examines how these changes affect the health of older adults. We examine the general hypothesis that racial health disparities can be explained, in part, by Blacks’ and Latinos’ exposure to increasing racial isolation and worsening economic conditions in neighborhoods. We aim to identify differences in the neighborhood change that Blacks, Latinos, and Whites experience in older age and to estimate the influence of neighborhood change on the hazard of all-cause mortality and cardiovascular disease. The proposed research will provide evidence to researchers and policy makers and support a future R01 to understand the mechanisms that link neighborhood change to racial health disparities.

The Contribution of Weight Status to Black-White Differences in Mortality and Diabetes | PI: Irma T. Elo, University of Pennsylvania

The Contribution of Weight Status to Black-White Differences in Mortality and Diabetes

PI: Irma T. Elo
Co-Investigators: Samuel H. Preston, Neil K. Mehta
University of Pennsylvania

Abstract 

The black-white health disparities in the United States are large and persistent. Cardiovascular diseases and diabetes account for a large fraction of the B-W difference in life expectancy at birth. Obesity is a key risk factor for developing heart disease and diabetes and it is more common among blacks than whites, especially among women. No study has systematically assessed the contribution of obesity to black-white differences in mortality or diabetes. We draw insights from our recent research that has shown the importance of accounting for weight histories in assessing the contribution of obesity to mortality and morbidity. The results of the proposed research will provide new estimates of the contribution of weight status to B-W differences in all-cause and cardiovascular disease mortality, life expectancy at age 40, and diabetes prevalence. 

Work Histories and Racial/Ethnic Disparities in Multimorbidity Trajectories | PI: Michal Engelman, University of Wisconsin-Madison

Work Histories and Racial/Ethnic Disparities in Multimorbidity Trajectories

PI: Michal Engelman, Ph.D.
University of Wisconson-Madison

Abstract

As longevity in the U.S rises, the persistence of racial/ethnic disparities may increasingly manifest in poor health. There are substantial differences by race/ethnicity in the prevalence of individual chronic conditions, as well as in multimorbidity (the co-occurrence of multiple chronic conditions). While chronic diseases grow common with age, little is known about the relationship between working (a nearly universal life course experience) and the pattern of disease accumulation in later life. The proposed study will explore the link between occupational histories and subsequent multimorbidity trajectories from a disparities perspective. It will test the hypothesis that cumulative exposures associated with specific occupational histories influence the pattern of subsequent health declines. Because labor markets and health experiences varied markedly by race/ethnicity and gender for older Americans, all models will be estimated separately for each demographic group. I will develop novel work exposure measures by combining longitudinal data on respondents’ longest-held occupations from the Health and Retirement Study with data from the Occupational information network, which captures the required abilities and work characteristics that distinguish specific occupations. Using latent-class methods, I will analyze the role of occupations and specific occupational characteristics (including decision-making capacities, creative thinking, interpersonal relationships, and physical demands) in differentiating multimorbidity trajectories both across and within demographic groups defined by race/ethnicity and gender. Findings from this research will inform debates surrounding the legal retirement age, and have implications for understanding the impact of occupational factors on current health disparities as well future trends for America’s aging and increasingly diverse population.

Race and Ethnic Differences in Measured Indicators of Health and Biological Risk | PI: Uchechi Mitchell, University of Southern California

Race and Ethnic Differences in Measured Indicators of Health and Biological Risk

PI: Uchechi A. Mitchell, Ph.D.
University of Southern California

Abstract

Eliminating disparities in health has been and continues to be a primary objective of public health researchers and practitioners. Much like disparities in chronic disease, cumulative biological disease risk-a summary measure of dysregulation across multiple physiological systems—is also systematically patterned by race/ethnicity. Blacks are more likely to be high-risk on measured biological indicators of health and have greater cumulative biological risk scores. To contribute to this body of work, the proposed study will examine race differentials in mortality, incidence of elevated biomarker levels (i.e., high-risk levels) and behavior responses to information on cardiometabolic risk. The specific aims of this project are to (1) determine whether race differences in mortality are associated with cardiometabolic risk and explain change in risk or subsequent differentials in risk; (2) examine race differences in change in the cardiometabolic score and in transitions between high- and low-risk levels of individual biomarkers; and (3) determine if there are race differences in behaviors and subsequent biological risk, after becoming aware of being high risk on indicators of high blood pressure, high cholesterol, or glucose dysregulation (i.e., diabetes). This project will advance research on racial/ethnic disparities in health among older adults by using longitudinal data from a large, racially diverse and national sample of older adults; using measured indicators of health; accounting for differential mortality; and testing multiple explanations for race differences in biological risk. In the end, the knowledge gained from this project will inform efforts to improve the health of older minorities and reduce racial and ethnic health disparities in the U.S. Population. The proposed research, therefore, meets objectives set forth by both the NIA and Healthy People 2020.

Sources of Favorable Neighborhood effects on Mexican-American Health | PI: Fernando Riosmena, University of Colorado Boulder

Sources of Favorable Neighborhood ‘Effects’ on Mexican-American Health

PI: Fernando Riosmena
University of Colorado Boulder

Abstract

This project will contribute to debates on the Hispanic Health Paradox (HHP) and other race/ethnic health disparities as well as on the effects of residential segregation by providing a broader, more in-depth examination of neighborhood effects on Mexican American health. People of Mexican origin living in residential environments with higher concentrations of coethnics are in better health than Mexican Americans living outside of these enclaves. Because “barrios” tend to have fewer economic and infrastructural resources conducive to better physical health, these associations could signal the existence of countervailing, socio-cultural protective mechanisms of intra-ethnic social support/control. These patterns may question the pervasiveness of the health-deleterious consequences of high socioeconomic and race/ethnic stratification and segregation in the United States. More specifically, the barrio effect could suggest that the Hispanic Health Paradox, i.e., the more favorable chronic health that Latinos often exhibit relative to their socioeconomic position in U.S. society, may be the result of locally-embedded protection mechanisms. However, the barrio effect could be an artifact of the composition of neighborhoods due to selective in-/out-migration. Furthermore, prior research has almost exclusively focused on estimating barrio effects for the average individual and community; examined barrio effects only in traditional Mexican American gateways; and only used census geography boundaries (e.g., individual or clusters of census tracts) as proxies for residential environments. 

This project aims to present a more complete picture of barrio effects by: studying and dealing with self- selection in and out of neighborhoods; examining the extent of variation in barrio effects over time and across measures, gateway types, neighborhood settings, and individual life courses (under a synthetic cohort perspective); and by using more flexible spatial units as proxies of residential environments than prior work. We use nationally-representative biomarker and anthropometric data from restricted-access 1999-2010 National Health and Nutrition Examination Surveys merged to contextual data from the 1990 and 2000 Censuses and 2005-2009 American Community Surveys to assess the existence of barrio effects in inflammation, cardiometabolic, and behavioral risk factors. We depart from and extend prior work on the topic by 1) addressing the role of self-selection in and out of neighborhoods; 2) examining (net-of-selection) variation in barrio effects across traditional, re-emerging, and new Mexican gateways as well as across neighborhoods over time and according to their levels of demographic stability, economic disadvantage, and turnout, as well as across the life course and other individual characteristics; and 3) taking advantage of exact latitude/longitude data in the survey to construct indicators of the residential environment using a buffer-based approach to define different spatial scales and zonings by using standard census geographies as well as by drawing buffers of 0.5, 1.0, 1.5, and 2.0 km. around each sampled dwelling.

Year 1 – Trends in Women’s Health

Changes in Cardiovascular Risk among Women and Men in the United States | PI: Jung Ki Kim, University of Southern California

Changes in Cardiovascular Risk Among Women and Men in the United States

PI: Jung Ki Kim, Ph.D.
University of Southern California

Abstract

The aim of this project is to examine recent trend in cardiovascular risk, the links between cardiovascular risk factors and mortality at multiple points in time, and onset of cardiovascular risk factors among men and women.  Data used for the analyses include the pooled data from National Health and Nutrition Examination Survey (NHANES) 1988-1994, 1999-2000 and after up to 2011-2012, and the Health and Retirement Survey (HRS) 2006/2008and 2010/2012 biomarker data.  Eight biomarkers that represent 9 cardiovascular risk factors such as systolic blood pressure, diastolic blood pressure, total cholesterol, HDL cholesterol, overweight, underweight, glycated hemoglobin, C-reactive protein (CRP)and cystatin C, and their summary scores are analyzed to show gender differences in presence and onset of cardiovascular risk among men and women and its link to mortality.  The results on longitudinal change of cardiovascular risk and its link to mortality would provide better understanding on worsening women’s health in the United States in the past decade. Also, connecting changes in biological risk to medical and behavioral practice such as menopause/hormone therapy, obesity and medication would suggest a projection of longer term trend in women’s health.

Educational and Cross-National Differences in Women's Health Expectancies | PI: Neil Mehta, Emory University

Educational and Cross-National Differences in Women’s Health Expectancies

PI: Neil K. Mehta, Max Planck 
Emory University

Abstract 

Recent research has demonstrated a number of troubling patterns with respect to U.S. women’s mortality both over time and relative to their European counterparts. In this project, we extend existing research on mortality to a focus on healthy and unhealthy life expectancy. We specifically examine the contribution of three key behavioral factors (obesity, smoking, alcohol) to cross-national and educational differences in women’s healthy and unhealthy life expectancy. In doing so, we provide methodological advancements in the estimation of health expectancies and their decomposition using high quality longitudinal data and innovative matrix population models. We focus on women ages 50 and above in the United States and Europe.

Explaining Inequalities in Women's Mortality Between U.S. States | PI: Jennifer Montez, Case Western Reserve University

Explaining Inequalities in Women’s Mortality Between U.S. States

PI: Jennifer Kara Montez
Consultants: Anna Zajacova, Mark D. Hayward 

Case Western Reserve University

Abstract

Inequalities in women’s mortality between U.S. states are large and growing. Accompanying the divergence in mortality across states was a decline in federal aid to states and an increase in states’ discretion over how to legislate and fund policies and programs. Thus, states may have contributed in important ways to the inequalities in women’s mortality. While states clearly differ in their contextual features (e.g., tax policies, education expenditures), they also differ in the characteristics of their populations. It is unknown how much of the inequality in women’s mortality between states reflects differences between states in their contextual features, net of women’s characteristics Distinguishing this contribution is imperative for identifying specific strategies and policies to reduce women’s mortality rates and inequalities. The specific aims of this project are to: (1) assess the extent to which the variation in women’s mortality across U.S. states reflects differences between states in their contextual features, net of women’s characteristics, and (2) identify the contextual features that contribute the most to the variation in women’s mortality. We hypothesize that inequalities in women’s mortality between states strongly reflect the states’ contextual features, net of women’s characteristics; and that the influence of states’ contexts on mortality is stronger for women than men. The project will use the 2013 National Longitudinal Mortality Study in which we link respondents to an extensive dataset of states’ contextual features that the project PI has assembled. These features span the states’ economic environment, social cohesion, sociopolitical orientation, physical infrastructure, and tobacco environment. The project will estimate multilevel, discrete-time event history models. It will also replicate the analyses for men and two age groups to examine how states’ contexts differentially impact the mortality of women and certain portions of the adult life course. Preliminary analyses indicate that states’ contexts have stronger and more pernicious consequences for women and older adults. The project is methodologically innovative and should make a significant contribution to public health by identifying policies and strategies from low mortality states that other states can implement to reduce mortality.

Trends in US Women's Health 1997-2014 by Education Level | PI: Anna Zajacova, University of Wyoming

Trends in US Women’s Health 1997-2014 by Education Level

PI: Anna Zajacova
Consultants: Jennifer Karas Montez, Linda G. Martin, Robert F. Schoeni
University of Wyoming

Abstract

Recent trends in morbidity and disability of non-elderly American women have been worrisome, with some indications of stagnation or even deterioration of health. Understanding the causes of these trends constitutes a necessary critical first step for reversing them. In particular, we urgently need to determine trends in different dimensions of health for women across educational-attainment levels in order to understand what groups are driving the aggregate trends, and why. This proposal will determine morbidity and disability trends for women at all levels of educational attainment and uncover the role of key health behaviors and socioeconomic risk factors in the observed health trends. At the completion of this project, we will have documented and (at least partly) explained the heterogeneity in recent morbidity and disability trends among US women. We analyze a large (N~150,000) sample of women age 45-84 from the National Health Interview Surveys collected in 1997-2014, focusing on multiple important dimensions of health including chronic conditions, functional limitations, and disability. Using nonparametric, semiparametric, and parametric models we will obtain detailed visualization of trends, as well as parsimonious results characterizing the trends and their determinants. This knowledge is important for the ultimate goal to identify the causes of women’s health trends, which in turn is critical for public policy and interventions to improve population health and decrease health disparities.