This page provides study documentation for BA24. For description of the specimen results, see Specimen Results Description (open to public). Data sets of the specimen results are included in the existing WHI datasets located on the WHI Data on this site (sign in and a completed Data Distribution Agreement are required; see details on the Data site).
PI: Kathy Rexrode (Brigham and Women’s Hospital)
Investigative team: Christine Albert, JoAnn Manson, Nina Paynter, and Nancy Cook (Brigham and Women’s Hospital); Clary Clish (Broad Institute, MIT); Raji Balasubramanian (University of Massachusetts, Amherst); Leslie Tinker (Fred Hutchinson Cancer Research Center)
Coronary heart disease (CHD) is the leading cause of death in US women. Many of the strongest risk factors for cardiovascular disease (CVD) in women, such as obesity, diabetes, and insulin resistance, are associated with altered metabolism, but exactly how these metabolic perturbations directly impact the pathophysiology of CHD is not clear. Metabolite profiling techniques (metabolomics) provide a tool for measuring a full profile of small-molecule metabolites, providing a comprehensive picture of an individual’s metabolic status. Recently, members of our investigative group have successfully applied this technology to identify a panel of metabolites that predicted incident diabetes over 12 years. Although preliminary results for CHD are promising, only small studies with short duration of follow-up have examined these compounds prospectively for CHD. This proposal will measure a well-validated metabolomic profile, and test specific metabolic hypotheses for incident CHD. In addition, the proposal will examine how hormone therapy (HT) changes metabolomic profiles, and whether these changes explain the increased risk observed for combined HT.
Metabolomic profiles of >300 metabolites will be measured on a total of 2210 women in the Women’s Health Initiative (WHI), including the WHI Observational Study (WHI-OS) and the WHI-HT trials. Profiles will be measured at baseline on 400 incident CHD cases and 400 controls in the WHI-OS, and all CHD cases and controls from the WHI-HT trials both at WHI baseline (1994-1998) and year 1. A two-phase analysis will be conducted with a “discovery cohort” of 400 CHD cases in the WHI-OS, and a “validation cohort” using the 342 CHD cases and matched controls among the placebo arm of the WHI HT trials. Additionally, the effect of HT on metabolomic profiles from baseline to year 1 will be compared in estrogen plus progestin (E+P), estrogen alone (E-alone) and placebo arms, and then the degree to which this mediates the observed risk of CHD with E+P will be tested. Metabolomic data will be integrated with prior clinical, biomarker and genomic data from each of these cohorts using network and pathways approaches.
Specific Aims:
1. To test individual metabolite and metabolomic profiles associated with incident CHD in women. (Using 400 CHD cases and 400 controls in the WHI-OS, with replication among the 342 cases and 342 controls in the WHI-HT placebo arms)
a) To test specific metabolic hypotheses regarding CHD:
1) Branched-chain amino acids (AA) and aromatic AA are associated with increased risk of CHD.
2) Small and medium-chain acylcarnitines are associated with increased risk of CHD.
3) Metabolites of dietary phosphotidylcholine are associated with increased risk of CHD.
4) Urea cycle metabolites are associated with higher risk of CHD.
5) Desaturation of lipids is associated with CHD.
b) Agnostic examination of metabolites and metabolomic profiles associated with CHD.
2. To compare changes in metabolomic profiles from baseline to year 1 in E-alone, E+P and placebo controls in the WHI-HT, and to test whether these changes explain the increased risk of CHD with E+P, but not E alone, in the HT trials.
3. To explore the relationship of metabolomic profiles with previously measured clinical parameters, biomarkers and genomic data. (Including blood pressure, waist circumference, body mass index,markers of lipoproteins, inflammation, endothelial function, insulin resistance and thrombosis/fibrinolysis, as well as genomic data).
This unique prospective study of the metabolomics of CHD in the WHI presents an exceptional opportunity to advance our understanding regarding the metabolic and biologic determinants of CHD in a multiethnic cohort of women. Metabolomic changes with HT will be tested as mediators of the increased risk of CHD observed with combined HT. Moreover, the proposal will add to the rich biological resources of the WHI, complementing existing GWAS and biomarker data. Most all of the 2210 participants in this study will have had GWAS previously measured as part of GARNET, SHARe, and other projects. Thus, there will be rich future opportunities in to integrate metabolomic, proteomic and genomic findings. Moreover, given the rich clinical, lifestyle, dietary, behavioral, biochemical and genomic data in the WHI, the relationship of metabolomics with these factors can be examined. The proposed metabolomic analyses have the potential to identify precursors of CHD that may advance early detection, determine changes associated with hormone therapy, and examine precursors of CVD-free survival. The metabolomics data will complement existing WHI resources including genomic, proteonomic and biomarker data, enhancing the research potential of the WHI biologic data.
Summary of participants involved: A total of 2210 WHI participants will be involved in the proposed study. Baseline metabolomic profiles will be measured on a total of 800 participants from the WHI-OS (400 case/control pairs). Baseline metabolomic profiles will be measured on a total of 1410 participants in the hormone trials (HT), including 342 case/control pairs in the HT-placebo arms (Aim 1 validation), and 363 case/control pairs in the active treatment arms (Aim 2b). Year 1 metabolomic profiles will be measured on 227 case/control pairs in the HT placebo arms (a subset of the 342 pairs), and 363 case/control pairs in the HT-active treatment arms for Aim 2. Aim 3 will make use of all data collected for Aims 1-2; no additional specimens will be analyzed for Aim 3. An additional 10% percent quality control specimens will be analyzed throughout. A total number of 3729 samples, including QC samples, will be analyzed.
Samples and participants by aim:
Aim 1 will involve a total of 1632 samples (400 pairs of participants from baseline OS and 342 pairs of participants from HT-placebo arms, plus 10% QC).
Aim 2 will involve a total of 2849 samples (including 752 baseline samples already measured in aim1). A total of 363 pairs of participants in the HT-active arms for both baseline and year 1 (1597 samples including QC) and 227 pairs of participants in the HT-placebo arms for year 1 measures (500 samples including QC) will be measured for aim 2.
Aim 3 will utilize all 3729 samples that will be measured in the proposal.