W34 - Extension of WHI Stroke Genome-wide Association Study (W7) for Stroke (W-7)

This page provides study documentation for Core Study W34. 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).

Investigator Names and Contact Information
Core study approved by WHI Steering Committee


Genome-wide associations studies are proving to be a valuable tool for gaining novel insights into disease processes and pathways and continue to have potential to elucidate effects for each CT intervention. 

The WHI GWAS performed in W7- Genome-Wide Scan on Breast Cancer, CHD, and Stroke involved genotyping about 360,000 SNPs at Perlegen Sciences for 1000 cases and 1000 pair-matched controls, in pools of size 125, for each of CHD, stroke, and breast cancer.  About 9,000 SNPs meeting statistical criteria (p < 0.025) were selected for further testing, along with about 1000 additional SNPs from candidate genes for each of the three diseases.  These SNPs were then genotyped in an additional 773 CHD, 613 stroke, and 800 breast cancer cases and controls from the OS.  For each disease, there was a modest excess of SNPs beyond expectation meeting second stage (p < 0.02) criteria (excess of 14.0 for CHD, 14.9 for stroke, and 8.6 for breast cancer) suggesting that the SNPs tested were enriched for those relevant to these diseases.  When the relatively small number of SNPs (about 200 for each disease) meeting Stage I and II criteria were tested in cases and controls in the E+P trial (335 CHD cases, 258 stroke cases, 349 breast cancer cases), there was no longer an excess above expectation, except possibly for stroke, and we were not in a position to make a compelling case for any SNP-disease association with data available thus far.  A second important purpose of the W7 WHI GWAS was to examine interactions of SNPs with E+P effects in Stage III.  There is evidence for one or more such interactions for each disease, but the argument for a meaningful interaction would be enhanced if the corresponding SNP ‘main effects’ were more clearly established.
The WHI GWAS was designed to identify SNPs having odds ratios in the 1.4 - 1.5 range for the minor SNP allele, as it seemed natural to target first the ‘low-hanging fruit’ at a limited cost.  However, other GWAS based on similar numbers of tagging SNPs that are now starting to be reported tend to be identifying odds ratios in the 1.1 - 1.2 range, which require a much larger sample size for adequate statistical power.  For example, the National Cancer Institute C-GEMS breast cancer and prostate cancer studies include about 1000 cases and controls at the first stage, but about 5000 cases and controls in subsequent stages.  The recently reported UK breast cancer GWAS (Easton et al, Nature, 2007) which identified 5 novel independent breast cancer loci starting with Perlegen’s tag SNP set, involved 390 cases having a strong family history of breast cancer and 364 controls at the first stage, followed by nearly 4000 cases and controls at the second stage.  The small number (30) of SNPs that emerged were then tested in over 20,000 cases and 20,000 controls from a cohort consortium to precisely pin down the SNPs having these modest odds ratios.
It is proposed to test about 5,000 SNPs on the additional stroke cases in the CT and OS not yet included in genotyping related to the WHI GWAS.  This includes 1,328 CT stroke cases and 268 OS stroke cases, for a total of 1,596 stroke cases.  A refined SNP chip that includes about 5400 SNPs will be developed by Perlegen Sciences for this project. These SNPs will be selected primarily from the 10,000 that scored highest in case versus control contrasts in stage 2 of the WHI GWAS, and will also include candidate SNPs from other external sources. Analysis will include these cases and controls as well as the 613 cases and controls from stage 2 of the WHI GWAS.  Separate analyses will be carried out for major stroke subcategories, to acknowledge the heterogeneous nature of this disease category, and to take advantage of the carefully adjudicated stroke outcome data in both the CT and OS. Data available from the stroke umbrella study, and other available biomarker data on WHI stroke cases  will be considered to strengthen and extend project analyses. MRI data, available on some WHI women (WHIMS study), provide another pertinent source of information that could contribute, for example, to the further study of associations arising from this project.
See publications: 1104, 1653.  WHI publications by study lists published WHI papers that have been generated by ancillary studies. A complete list of WHI papers is available in the Bibliography section of this website.