Abstract
BACKGROUND: Self-reporting is often used in large epidemiologic research to identify hypertensive participants, but its validity in young adults has not been sufficiently assessed. We aimed to validate self-reported hypertension diagnosis from questionnaires by using medical records among a sample of young adults aged 22-39 years in the Growing Up Today Study (GUTS), during a time of transition in national hypertension definitions. METHODS: A sample of 1,000 GUTS participants were asked for permission to access their medical records, to confirm their self-reported hypertension status on questionnaires from 2010 to 2019. Paired self-reported and medical record information was available for 318 participants. Medical records were reviewed by a clinical hypertension specialist. We evaluated agreement, kappa statistic, sensitivity and specificity of self-reporting. We assessed the correlations between blood pressure measurements self-reported on the 2019 questionnaire and those from medical records and compared the measurements using paired t-tests. RESULTS: The selected sample was generally representative of the full GUTS cohort. Agreement, kappa coefficient, sensitivity and specificity of self-reported hypertension were 85.5%, 0.72, 100%, and 75.3%. Although the absolute differences in blood pressure measurements between self-report and medical records were small (e.g., the average difference in typical recent blood pressures was 3.5/1.2 mm Hg), these measures were only moderately correlated. CONCLUSIONS: Validity of self-reported hypertension was high in GUTS, ensuring use as an endpoint in future studies with confidence. We demonstrated that young adults likely without formal medical training are able to report hypertension status with reasonable accuracy.