Socioeconomic status over 12 years and subclinical cardiovascular disease: The Study of Women’s Health Across the Nation

Abstract

Background and purpose

The inverse relation between socioeconomic status (SES) and cardiovascular disease (CVD) is well-established. However, few studies have investigated SES assessed repeatedly over adulthood in relation to subclinical atherosclerosis. We aimed to test whether consistently low SES, as indexed by education, income, and financial strain over 12 years of midlife was related to later carotid intima media thickness (IMT) and plaque among women.

Methods

The Study of Women’s Health Across the Nation is a multi-site longitudinal study of midlife women. Education was assessed at the study baseline, income and financial strain were obtained yearly over 12 years, and a carotid ultrasound was obtained at study year 12 among 1402 women. Associations were tested in linear and multinomial logistic regression models adjusted for demographic, biological, and behavioral risk factors.

Results

A high school education or less (OR (95%CI): 1.72(1.15,2.59), p<0.01), some college education (OR (95%CI): 1.65(1.17,2.32), p<0.01), consistently low income (OR (95%CI): 1.83(1.15,2.89), p<0.05) and consistent financial strain (OR (95%CI): 1.78(1.21,2.61), p<0.01) over 12 years were associated with higher carotid plaque, and consistent financial strain was associated with elevated maximal IMT (b(SE)=0.02(0.01), p<0.05) controlling for standard CVD risk factors. When SES indices were considered together, financial strain (b(SE); 0.02(0.01), p<0.05) and low education (high school education or less: OR (95%CI): 1.55(1.01,2.37), p<0.05; some college: OR (95%CI): 1.56(1.09,2.21), p<0.05) were most consistently associated with IMT and plaque, respectively, controlling for risk factors.

Conclusions

Findings indicate the importance of targeting economically disadvantaged women in efforts to prevent CVD among women.

Introduction

The inverse relation between socioeconomic status (SES) and cardiovascular disease (CVD) is well-established.¹ Typically, studies investigate the relations between SES assessed at a single point in time in relation to CVD risk. However, SES is not static, and chronic exposure to low SES may be most detrimental to health.² Few studies have investigated SES over adulthood in relation to CVD risk.

CVD typically develops over decades, with subclinical levels present before a clinical CVD event occurs. Using subclinical CVD measures is useful to investigating relations of SES to CVD to mitigate potential SES biases in CVD event presentation and detection. Carotid intima media thickness (IMT), one of the most widely-used and well-validated subclinical CVD measures, utilizes ultrasound to quantify the thickness of the lumen-intima interface and the media-adventitia interface of the carotid artery. Carotid ultrasound can also be used to quantify plaque in the carotid artery. Both IMT and plaque ultrasound measures are highly reproducible and significantly associated with later CVD events, even among low risk populations.³

Several studies have utilized subclinical CVD indices to examine SES disparities in CVD risk. Most of these studies are cross-sectional in nature.^4–6 No studies have examined SES measured over midlife in relation to carotid IMT. Midlife is a particularly relevant time to consider in women, as atherosclerotic changes and adverse changes in key CVD risk factors may be accelerated.^{7, 8}

We tested whether consistently low SES, as indexed by education, income, and financial strain over 12 years of midlife was related to higher carotid IMT and plaque among women. We tested this hypothesis in the Study of Women’s Health Across the Nation (SWAN), a longitudinal study of White, Black, Hispanic and Chinese women transitioning through the menopause in which SES measures were obtained yearly over 12 years. We additionally considered whether traditional CVD risk factors explained these associations and whether these relations varied by race/ethnicity.

Methods

SWAN is a community-based, multicenter, longitudinal study designed to characterize the menopause transition.⁹ Briefly, SWAN is conducted at seven sites: Boston, MA; Chicago, IL; the Detroit, MI area; Los Angeles, CA; Newark, NJ; Pittsburgh, PA; Oakland, CA. From 1996–1997, 3,302 women aged 42–52 years were enrolled. Each site recruited Caucasian women plus one racial/ethnic group: African American (Pittsburgh, Chicago, Detroit, Boston), Chinese (Oakland), Japanese (Los Angeles) and Hispanic (Newark). Women were recruited from lists of names or household addresses. Seventy-three percent of the women selected were contacted and provided information to determine eligibility; 51% (N=3302) of eligible women enrolled. Baseline eligibility criteria included being aged 42–52 years, having a uterus and ≥one ovary, not being pregnant or lactating, not using oral contraceptives or hormone therapy (HT), and having ≥one menstrual cycle in the prior 3 months. 3,302 women were enrolled. Six sites collected carotid ultrasound data at SWAN visit 12 (Boston, Chicago, Newark, Pittsburgh, Oakland, Detroit); thereby Caucasian, African American, Chinese, and Hispanic women had ultrasound data. SWAN protocols were approved by the institutional review boards at each site. Each participant provided written informed consent. This investigation was an analysis of SES from baseline through SWAN visit 12 in relation to subclinical CVD at SWAN visit 12.

Of the 1,566 women who underwent carotid measurements, 149 women were excluded due to reporting having had a stroke, angina, or myocardial infarction prior to their carotid scan and 15 woman for having an invalid scan. Compared to women included in the analysis, excluded women were more often African American, smokers, earlier in the menopause transition, more often used antihypertensive, glucose-lowering, lipid-lowering, and anticoagulant medications, had lower education, higher body mass index (BMI), higher systolic blood pressure (BP), a poorer lipid profile, higher homeostatic model assessment (HOMA) index, higher IMT, and higher plaque (p’s<0.05). These differences were expected given sample exclusions. 1,402 women were included in primary models (1,402, 1,369, and 1,373 in models for plaque, mean IMT, and maximal IMT, respectively).

SES

Education, income, and financial strain were assessed. Education was assessed at baseline (categorized as ≤ high school, some college or vocational school, ≥ college). Total annual household income was self-reported annually (categorized as low: ≤$34,999; medium: $35,000–$74,999; high: ≥$75,000). Income over the course of the study was categorized as: consistently low (≥50% of visits with low income), medium or mixed (≥50% of visits with medium income, 50% of visits with medium and 50% of visits with low or high income, or income distributed between the three categories without reaching 50% for any one category), and high (≥50% of visits high income). Only 57 women had mixed income, these women did not differ from medium-income women on any of the outcome measures, and thereby were grouped with medium-income women. Financial strain was assessed in response to: “How hard is it for you to pay for the very basics like food, housing, medical care, and heating?” (“very hard”, “somewhat hard”, “not hard at all”; categorized as “hard” and “not hard” due to cell sizes). Financial strain over the study was categorized as consistently low (no visits with difficulty paying for basics), moderate or mixed (>0–<50% of visits difficulty paying for basics), and consistently high (≥50% of visits difficulty paying for basics).

Carotid ultrasound

Ultrasound images were obtained using a Terason t3000 Ultrasound System (Teratech Corp, Burlington, MA) equipped with a variable frequency 5 to 12 Mhz linear array transducer. Digitized images were obtained from the near and far wall of the left and right distal common carotid artery, 1 cm proximal to the carotid bulb. IMT measures were obtained by electronically tracing the lumen-intima interface and the media-adventitia interface across a 1-cm segment for each of these 4 segments; one measurement was generated for each pixel over the area, for a total of ~140 measures for each segment. The means of the average and maximal readings at the 4 locations were used in analyses. Presence and extent of plaque were evaluated in each of 5 segments of the left and right carotid artery (distal and proximal common carotid artery, carotid bulb, proximal internal and external carotid arteries). Plaque was defined as an area protruding into the vessel lumen that was ≥50% thicker than the adjacent IMT. For each segment, plaque was graded between 0 (no plaque) to 3 (plaque covering ≥50% of the vessel); grades across segments summed to yield the plaque index.¹⁰ Technicians at the study sites were trained by the University of Pittsburgh Ultrasound Research laboratory and monitored for reliability. Images were read at the University of Pittsburgh Ultrasound Reading Center using the AMS software developed by Dr. Thomas Gustavsson.¹¹ Reproducibility of IMT measures was excellent with an intraclass correlation coefficient between sonographers of ≥0.77, and between readers of 0.90, and the plaque index similarly reliable with an intraclass correlation ranging from 0.86 to 0.93.¹²

Covariates

Race/ethnicity was determined from the SWAN screening interview. Age, smoking status (current vs. past/never), psychosocial factors, insurance status, alcohol use, physical measures, and menopausal status were derived from questionnaires, interviews, and measures taken at visit 12 (the visit concurrent with the carotid ultrasound). Alcohol use was the reported average weekly number of servings of beer, wine, liquor, or mixed drinks. Menopausal status was obtained from self-reported bleeding patterns over the year. At the 12^th annual visit, most the women were postmenopausal (~85%); therefore women were categorized as postmenopausal (≥12 of amenorrhea), pre/peri-menopausal (a menstrual period within 12 months), or surgically menopausal (hysterectomy). Height and weight were measured and BMI calculated (kg/m²). BP was the average of two seated measurements. Given the high correlation between systolic and diastolic BP, the measure with the strongest association with the outcome was included in models. Women reported the use of cardiovascular medications (insulin, medication for blood thinning, medication for BP lowering, medication for lipid lowering) annually. Use of each these medications at any point over the 12 years were treated as four separate variables and covariates. Insurance status was also assessed (medical insurance with or without medication coverage). Several psychological factors were assessed via validated instruments, including depressive symptoms (Center for Epidemiologic Studies Depression Scale¹³), anxiety (Generalized Anxiety Disorder-7¹⁴) and positive and negative affect (Positive and Negative Affect Scale, PANAS¹⁵). Only positive affect was associated with any aspect of the carotid and therefore positive affect was included in final models.

Phlebotomy was performed in the morning following overnight fast within 90 days of the annual visit. Blood was separated, frozen (−80°C), and sent on dry ice to the University of Michigan Pathology Laboratory, CLIA-certified and accredited by the College of American Pathologists. Measurements were performed on a Siemens ADVIA 2400 automated chemistry analyzer utilizing Siemens ADVIA chemistry system reagents. Glucose was measured using a two-step enzymatic reaction utilizing hexokinase and glucose-6-phosphate dehydrogenase enzymes. Serum insulin was measured using radioimmunoassay. HOMA [(fasting insulin*fasting glucose)/22.5]¹⁶ was calculated. Lipid fractions were determined from EDTA-treated plasma. Total cholesterol and triglycerides concentrations were determined by coupled enzymatic methods.^17–18

Data Analyses

IMT (mean and max) and triglyceride values were natural log transformed for analysis. Given its skew, the plaque index was categorized as none (0), moderate (1) or high (>1). Baseline differences between included/excluded participants were tested using Wilcoxon rank-sum and chi-square tests. Associations between SES variables (considered separately and together) and each subclinical CVD outcome were estimated in linear and multinomal logistic regression models. Models were first adjusted for age, race and site, with additional adjustment for covariates selected based upon associations with outcomes at p<0.05. Interactions between SES and race/ethnicity were examined as cross product terms in multivariable models. Residual analysis and diagnostic plots were conducted to verify model assumptions. Analyses were performed with SAS v9.2 (SAS, Cary, NC). Models were 2-sided, α=0.05.

Results

Women were on average 59 years old, overweight, and postmenopausal (Table 1). Half of the sample was comprised of non-Hispanic Caucasian women, 30% of African American women, and the remainder Chinese or Hispanic women. Education was moderately correlated with income (ρ=0.39, p<0.0001) and financial strain (ρ=0.32, p<0.0001); income and financial strain were moderately-strongly correlated with each other (ρ=0.56, p<0.0001).

Table 1.

Participant characteristics (N=1402)

Age, years (Mean±SD)	59.5±2.7
Race/ethnicity n (%)
Black	417(29.7)
Caucasian	711(50.7)
Chinese	189(13.5)
Hispanic	85(6.1)
Education n (%)
≤ High school	307(22.1)
Some college	427(30.8)
≥ College	653(47.1)
Income n(%)
Consistently low	291(20.8)
Consistently medium/mixed	621(44.3)
Consistently high	488(34.9)
How hard to pay for basics
Consistently not hard	623(44.4)
Moderate/mixed	374(26.7)
Consistently somewhat or very hard	405(28.9)
Menopausal Status n (%)
Pre/peri-menopausal	44(3.1)
Natural postmenopusal	1199(85.5)
Surgical menopausal	159(11.4)
BMI, kg/m2 (Mean±SD)	29.9±7.3
SBP, mmHg (Mean±SD)	121.9±17.0
DBP, mmHg (Mean±SD)	74.3 ±10.1
HDL, mg/dL (Mean±SD)	62.2±16.4
LDL, mg/dL (Mean±SD)	120.5±31.7
Triglycerides, mg/dL (Median (Q1,Q3))	100.0(75.0, 139.0)
HOMA index (Median (Q1,Q3))	2.1(1.2, 3.8)
PANAS*: Negative Affect (10–50) (Median (Q1,Q3))	13.0(11.0, 17.0)
PANAS*: Positive Affect (10–50) (Mean±SD)	33.5±8.0
Smoker, current n (%) yes	123 (8.8)
Alcohol consumption n (%)
<1/month	711(51.7)
1/week	355(25.8)
>=2/week	310(22.5)
Antihypertensive use n (%)	603(43.6)
Glucose-lowering medication use n (%)	162(11.6)
Lipid lowering-medication use n (%)	464(33.4)
Anticoagulant use n (%)	179(13.0)
HT use n (%)	524(37.4)
Average IMT (Median (Q1,Q3))	0.78(0.71,0.86)
Maximum IMT (Median (Q1,Q3))	0.90(0.83,1.00)
Plaque n (%)
None (0)	808(57.6)
Moderate (1)	257(18.3)
High (>1)	337(24.1)

Note: Medication use reflects use at any point during baseline through study year 12

^*PANAS: Positive and Negative Affect Scale

When each SES index was considered separately in relation to subclinical CVD, low education (some college/vocational education or less) and any financial strain were related to higher mean and maximum IMT (Table 2). All three SES indices were related to increased plaque. Controlling for covariates, financial strain, particularly when chronic, was related to higher maximal IMT and plaque. Low education and consistently low income were also related to increased plaque. Thus, in multivariable models, low SES on all three exposures were related to plaque, with financial strain additionally related to elevated maximal IMT.

Table 2.

Relations between SES indices (considered separately) and IMT and plaque

	Mean IMT	Max IMT	Plaque Index
			Moderate (vs. None)	High (vs. None)
	β(SE)	β(SE)	OR (CI)	OR (CI)
Model 1
Education
≤ High school	0.02(0.01)*	0.02(0.01)*	1.31(0.87,1.97)	2.01(1.40,2.88)†
Some college	0.02(0.01)*	0.02(0.01)*	1.31(0.93,1.84)	1.82(1.33,2.51)†
≥ College graduate	---	---	---	---
Income
Consistently low	0.002(0.01)	0.004(0.01)	1.23(0.78,1.93)	1.84(1.24,2.74)‡
Consistently medium/Mixed	0.01(0.01)	0.01(0.01)	1.18(0.85,1.63)	1.22(0.90,1.66)
Consistently high	---	---	---	---
How hard to pay for basics
Consistently hard	0.02(0.01)*	0.03(0.01)‡	1.51(1.03,2.21)*	1.92(1.37,2.68)†
Moderate/Mixed	0.02(0.01)§	0.02(0.01)*	1.48(1.05,2.08)*	1.12(0.81,1.56)
Consistently not hard	---	---	---	---
Model 2
Education
≤ High school	0.01(0.01)	0.01(0.01)	1.32(0.84,2.06)	1.72(1.15,2.59)‡
Some college	0.01(0.01)	0.01(0.01)	1.25(0.87,1.80)	1.65(1.17,2.32)‡
≥ College graduate	---	---	---	---
Income
Consistently low	0.001(0.01)	0.002(0.01)	1.36(0.81,2.27)	1.83(1.15,2.89)*
Consistently medium/Mixed	0.001(0.01)	0.001(0.01)	1.24(0.87,1.76)	1.19(0.85,1.67)
Consistently high	---	---	---	---
How hard to pay for basics
Consistently hard	0.02(0.01)§	0.02(0.01)*	1.66(1.09,2.53)*	1.78(1.21,2.61)‡
Moderate/Mixed	0.01(0.01)	0.01(0.01)	1.65(1.15,2.37)‡	1.05(0.73,1.50)
Consistently not hard	---	---	---	---

Model 1 adjusted for age, site and race

Model 2 adjusted for age, site, race, SBP, BMI, HOMA, HDL, LDL, triglycerides, smoking, alcohol use, use of antihypertensives, use of lipid lowering medications, use of anticoagulants, use of insulin, positive affect

^§p<0.10,

^*p<0.05,

^‡p<0.01,

^†p<0.001

We next considered SES variables simultaneously (Table 3). In models adjusted for covariates and with all three SES indices considered together, low education was related to increased plaque and consistent financial strain related to both increased likelihood of moderate plaque and higher maximum IMT. Thus, both low education and financial strain were related to plaque and/or maximal IMT when considered together and adjusting for covariates.

Table 3.

Relations between SES indices (considered together) and IMT and plaque

	Mean IMT	Max IMT	Plaque Index
			Moderate (vs. None)	High (vs. None)
	β(SE)	β(SE)	OR (CI)	OR (CI)
Model 1
Education
≤ High school	0.02(0.01)*	0.02(0.01)*	1.18(0.77,1.81)	1.77(1.21,2.59)†
Some college	0.02(0.01)*	0.02(0.01)‡	1.24(0.88,1.76)	1.71(1.23,2.36)†
≥ College graduate	---	---	---	---
Income
Consistently low	−0.02(0.01)*	−0.02(0.01)*	0.94(0.56,1.59)	1.18(0.74,1.89)
Consistently medium/Mixed	−0.002(0.01)	−0.002(0.01)	1.01(0.71,1.43)	0.99(0.71,1.39)
Consistently high	---	---	---	---
How hard to pay for basics
Consistently hard	0.02(0.01)‡	0.03(0.01)†	1.50(0.97,2.32)*	1.57(1.06,2.33)‡
Mixed	0.02(0.01)*	0.02(0.01)‡	1.45(1.01,2.07)‡	1.01(0.72,1.43)
Consistently not hard	---	---	---	---
Model 2
Education
≤ High school	0.01(0.01)	0.01(0.01)	1.18(0.74,1.87)	1.55(1.01,2.37)‡
Some college	0.01(0.01)	0.01(0.01)	1.18(0.81,1.70)	1.56(1.09,2.21)‡
≥ College graduate	---	---	---	---
Income
Consistently low	−0.001(0.01)	−0.02(0.01)	1.04(0.58,1.86)	1.26(0.74,2.15)
Consistently medium/Mixed	−0.001(0.01)	−0.006(0.01)	1.05(0.72,1.53)	1.01(0.70,1.46)
Consistently high	---	---	---	---
How hard to pay for basics
Consistently hard	0.02(0.01)*	0.02(0.01)‡	1.61(1.01,2.58)‡	1.49(0.96,2.30)*
Mixed	0.01(0.01)	0.01(0.01)	1.62(1.11,2.36)‡	0.98(0.67,1.43)
Consistently not hard	---	---	---	---

Model 1 adjusted for age, site and race

Model 2 adjusted for age, site, race, SBP, BMI, HOMA, HDL, LDL, triglycerides, smoking, alcohol use, use of antihypertensives, use of lipid lowering medications, use of anticoagulants, use of insulin, positive affect

^*p<0.10,

^‡p<0.05,

^†p<0.01

We conducted several additional analyses. We tested whether associations between SES and subclinical CVD varied by race/ethnicity. These analyses included White, Black and Chinese women given the few Hispanic women in this sample. Some evidence of modification by race (p<0.05) suggested that consistent financial strain was most clearly related to plaque for White women (OR=2.02, 95%CI: 1.26–3.24, p=0.004) and mixed financial strain with plaque for Black women (OR=2.04, 95%CI: 1.12–3.72, p=0.02). However, interpretation of these differences may be limited by the small sample of certain racial groups the multiple interactions tested. We also considered insurance status as a covariate, with findings largely unchanged (data not shown).

Discussion

This study is the first to examine low SES across midlife and across several SES indices in relation to IMT and plaque in women. Low education, low income, and financial strain, particularly when experienced chronically over midlife, were associated with a greater plaque accumulation in the carotid artery among women free of clinical CVD. In addition, financial strain was associated with elevated IMT. When SES indices were considered together, financial strain and low education remained independently associated with subclinical CVD. These associations persisted after controlling for standard CVD risk factors and were broadly similar across racial/ethnic groups.

Prior research on SES and carotid IMT has largely considered cross-sectional relations between SES and subclinical CVD,^4–6 comparisons of SES measured once during childhood and adulthood,^19–23 or the relation between SES and IMT progression.^24–26 In general, these studies have yielded mixed findings, as the associations between SES and subclinical CVD have often,^{4, 19, 23} but not always,^{5, 6, 20} been explained by CVD risk factors. Two studies of a single cohort have examined relations between SES considered repeatedly over adulthood in relation to subclinical CVD (coronary artery calcification, IMT) among Black and White young adults. These studies found associations between lower SES over adulthood and higher coronary artery calcification, yet largely among the Black participants²⁷ and between occupational mobility and IMT, although primarily among the male participants.²⁸ No studies have tested relations between SES changes over midlife in women in relation to carotid IMT and plaque.

Our study considered three indices of SES, including education, and financial strain, and income, measured approximately annually over 12 years. Considered together, low education and persistent financial strain were most consistently related to subclinical CVD. Various indices may capture slightly different aspects of SES that may be differentially related to CVD risk.²⁹ Education can pattern later opportunities, and is consistently associated with cardiovascular health, particularly for women.³⁰ Considering resource measures, financial strain was more robustly related to health than low income, particularly when sustained. Financial strain may better account for saved resources that can protect against income shocks and better index the adequacy of resources to meet basic expenditures.

Midlife is a critical phase of life for women. It encompasses the menopause transition, a time of decline in endogenous estradiol and cessation of menstrual cycles. The menopause is associated with adverse changes in lipids⁸ and possibly accelerated IMT.⁷ Midlife is also a time of changing social roles, including caretaking aging parents, children leaving home, and retirement, all impacting financial status.³¹ Thus, capturing CVD risk status during midlife in women is particularly relevant, as social and economic factors may act synergistically with ovarian aging to accelerate risk.

The mechanisms underlying these relations are likely numerous. Standard CVD risk factors such as BP, lipids, or glucose, or medications to treat these risk factors did not explain these associations. The psychosocial stress associated with financial strain may be a pathway linking SES to CVD. Notably, the psychological factors assessed here were largely unrelated to subclinical CVD. However, the full spectrum of psychological factors was not assessed, and additional factors such as anger³² should be considered in future work. Key health behaviors that are potent CVD risk factors (e.g., smoking) as well as BMI did not explain the observed associations. Access to healthcare can vary by SES, although this access typically does not explain SES gradients in health.³³ Insurance status did not change findings here. Other novel risk factors, such as inflammation or variations in autonomic nervous system function, may play a role in the observed associations and can be considered in future work. Given the time period over which CVD develops, the pathways linking low SES to CVD are likely numerous.

This study had several limitations. Carotid scans were administered once at year 12. Thus, we are not able to test subclinical CVD progression. Education was measured once, at baseline. Therefore, any additional education women may have received in midlife, while rare and generally not associated with health,³⁴ is not captured here. Income had notable limitations in its assessment that may have increased error. It could not be indexed to household size (household size was not available), and it did not account for variations in cost of living across regions or over time. Change in income across categories was rare, limiting the ability to test income mobility and subclinical CVD. Men were not included here.

This study had several strengths. Women were assessed approximately yearly over 12 years on key exposures, allowing a strong test of consistency of exposure to low SES in relation to later subclinical CVD. Several racial/ethnic groups were included, allowing for testing of the consistency of effects across groups. Multiple SES indices were considered here. The subclinical CVD measures utilized here are useful in indexing disease development before the emergence of clinical disease.

In summary, consistent financial strain, low education, and to a lesser extent, low income over 12 years, were associated with elevated subclinical CVD among midlife women. Associations persisted controlling for standard CVD risk factors. There have been calls for a better understanding of sociocultural factors in relation to the development of CVD, including its intermediate phenotypes.³⁵ These results underscore the importance of considering socioeconomic adversity in relation to the development of atherosclerosis. They suggest the potential value in targeting economically disadvantaged women in efforts to prevent CVD among women.

APPENDIX

SWAN Clinical Centers

University of Michigan, Ann Arbor-Siobán Harlow, PI 2011-present, MaryFran Sowers, PI 1994–2011; Massachusetts General Hospital, Boston, MA-Joel Finkelstein, PI 1999-present; Robert Neer, PI 1994–1999; Rush University, Rush University Medical Center, Chicago, IL-Howard Kravitz, PI 2009-present; Lynda Powell, PI 1994–2009; University of California, Davis/Kaiser-Ellen Gold, PI; University of California, Los Angeles-Gail Greendale, PI; Albert Einstein College of Medicine, Bronx, NY-Carol Derby, PI 2011- present, Rachel Wildman, PI 2010–2011; Nanette Santoro, PI 2004–2010; University of Medicine and Dentistry-New Jersey Medical School, Newark-Gerson Weiss, PI 1994–2004; and the University of Pittsburgh, Pittsburgh, PA-Karen Matthews, PI.

NIH Program Office

National Institute on Aging, Bethesda, MD-Winifred Rossi 2012-present; Sherry Sherman 1994–2012; Marcia Ory 1994–2001; National Institute of Nursing Research, Bethesda, MD-Program Officers.

Central Laboratory

University of Michigan, Ann Arbor-Daniel McConnell (Central Ligand Assay Satellite Services).

Coordinating Center

University of Pittsburgh, Pittsburgh, PA – Maria Mori Brooks, PI 2012 - present; Kim Sutton-Tyrrell, PI 2001 – 2012; New England Research Institutes, Watertown, MA - Sonja McKinlay, PI 1995 – 2001.

Steering Committee: Susan Johnson, Current Chair

Chris Gallagher, Former Chair