medRxiv · Apr 17, 2023

Reliability and Validity of self-reported Vascular Risk Factors in a Multi-Ethnic Community Based Study of Aging and Dementia

By Lee AJ, Sanchez D, Reyes-Dumeyer D et al.

INTRODUCTION: The reliability and validity of self-reported cardiovascular and cerebrovascular risk factors remains inconsistent in aging research. METHODS: We assessed the reliability, validity, sensitivity, specificity, and percent agreement of self-reported hypertension, diabetes, and heart disease, in comparison with direct measures of blood pressure, hemoglobin A1c (HbA1c), and medication use in 1870 participants in a multiethic study of aging and dementia. RESULTS: Reliability of self-reported for hypertension, diabetes, and heart disease was excellent. Agreement between self-reports and clinical measures was moderate for hypertension (kappa: 0.58), good for diabetes (kappa: 0.76-0.79), and moderate for heart disease (kappa: 0.45) differing slightly by age, sex, education, and race/ethnic group. Sensitivity and specificity for hypertension was 88.6%-78.1%, for diabetes was 87.7%-92.0% (HbA1c > 6.5%) or 92.7%-92.8% (HbA1c > 7%), and for heart disease was 85.8%-75.5%. DISCUSSION: Self-reported history of hypertension, diabetes, and heart disease are reliable and valid compared to direct measurements or medication use.

doi.org/10.1101/2023.04.12.23288492Open accessPubMed 37131736Biomarkers