NEW YORK (Reuters Health) – While a hypertension risk prediction model developed in the Framingham Heart Study (FHS) shows good discrimination, its performance is not significantly better than a model with systolic blood pressure (SBP) alone, a new study suggests
This finding highlights the importance of blood pressure elevations among individuals without hypertension, said lead author Dr. Paul Muntner, of the University of Alabama, Birmingham, and colleagues in their paper, published online May 5th in Hypertension.
They add, Aggressive interventions for individuals with elevated SBP may be warranted.
The FHS model includes SBP, gender, parental history of hypertension, body mass index, cigarette smoking, and the interaction between diastolic blood pressure (DBP) and age.
Dr. Muntner and colleagues compared the FHS model, SBP alone, and a model of age-specific diastolic blood pressure (DBP) categories for the prediction of hypertension in 3013 non-hypertensive, non-diabetic subjects. The mean age of the participants was 58.5 years, and 53% were women.
As part of the Multi-Ethnic Study of Atherosclerosis, all participants had four examinations between 2000 and 2008, with a median of 1.6 years between exams. Overall, 849 patients developed hypertension between one examination and the next (i.e., SBP > 140 mm Hg or diastolic BP > 90 mm Hg, or initiation of antihypertensive drugs).
In multivariate analyses that included race-ethnicity and the components of the FHS model, incident hypertension was associated with Black race-ethnicity, SBP, age-specific DBP levels, and body mass index.
The c-statistic for the incidence of hypertension was 0.788 for the FHS model, 0.768 for SBP alone (compared to the FHS model), and 0.699 for the model of age-specific DBP categories.
The FHS model yielded a relative integrated discrimination improvement of 10% compared to SBP alone (and 146% compared to the DBP model).
The FHS prediction model underestimated the risk of hypertension in all of the deciles of predicted risk, but after recalibration and using a best-fit model the predicted risk for hypertension was
