Patient oriented and epidemiologic research projects, particularly multi-center projects, typically generate data with potential utility beyond the specific hypotheses and questions for which they were designed. The Women's Health Initiative (WHI) Clinical Trials enrolled 68,133 female participants aged 50-79 in dietary and hormone therapy trials. Self-reports of the primary and secondary outcomes (e.g. breast cancer, coronary heart disease) were confirmed by reviews of medical records. In contrast, self-reported incident type 2 diabetes mellitus was not independently confirmed by review of medical records, nor was routine diagnostic testing performed. Confirming that participant self-report of diabetes in WHI is accurate would greatly enhance the value of the WHI data as an unparalleled resource for further investigation of the effects of dietary, hormonal and other influences on diabetes in older women. Approximately 7% of WHI Clinical Trial participants reported diabetes at baseline, and the self-reported incidence of new-onset medication-treated diabetes in WHI has been close to 1% per year. In previous epidemiologic studies, self-reports of diabetes were confirmed at widely ranging rates from 64% to 98%. Preliminary data from medical record reviews at the Minneapolis WHI Field Center showed that the positive predictive value of self-reported incident diabetes during follow-up was 85% (74%- 94%) and the negative predictive value of never self-reporting diabetes was 99% (92%-100%). The principal aim of the proposed study is to obtain more precise estimates of the positive and negative predictive values of WHI self-reports of incident diabetes in the WHI clinical trials by expanding the study to 3 other WHI Field Centers. These additional Field centers collectively represent WHI closely with regard to diabetes risk factors, demographics, prevalence and incidence of self-reported diabetes. We will use existing data in the WHI database, supplemented by newly collected data from a questionnaire and medical record reviews in a subset of participants to carry out the study's aims. The new data collection employs similar methods to those already in place for adjudication of other self-reported WHI primary outcomes, and will be used to establish a gold standard to which the self-reports of diabetes in WHI can be compared. Secondary aims are to compare the confirmation rates of self-reported diabetes between intervention and control arms within the trials, and develop a prediction model for incident diabetes that incorporates clinical characteristics (e.g., age, race/ethnicity, education, BMI) along with self-reported diabetes status. This is the first study to examine the negative predictive value of self-reports of diabetes. The availability of serial fasting serum glucose data from a large cohort of postmenopausal women provides a unique opportunity to estimate of the rate of under-reported incident diabetes in this population. This study will examine whether post-menopausal women who self-reported diabetes in the Women's Health Initiative study really had diabetes according to standard diagnostic criteria. The results of the study will help determine whether self-reported diabetes can be used for other analyses of the WHI data, and for post- menopausal women in other studies. ? ? ?
| Jackson, Jody M; DeFor, Terese A; Crain, A Lauren et al. (2014) Validity of diabetes self-reports in the Women's Health Initiative. Menopause 21:861-8 |
| Luo, Juhua; Rossouw, Jacques; Margolis, Karen L (2013) Smoking cessation, weight change, and coronary heart disease among postmenopausal women with and without diabetes. JAMA 310:94-6 |
