Previous studies have found pulse pressure (PP), a marker of arterial stiffness, to be an independent predictor of atrial fibrillation (AF) in general and hypertensive populations. We examined whether PP predicted new-onset AF in comparison with other blood pressure components in the Losartan Intervention For Endpoint reduction in hypertension study, a double-blind, randomized (losartan versus atenolol), parallel-group study, including 9193 patients with hypertension and electrocardiographic left ventricular hypertrophy. In 8810 patients with neither a history of AF nor AF at baseline, Minnesota coding of electrocardiograms confirmed new-onset AF in 353 patients (4.0%) during mean 4.9 years of follow-up. In multivariate Cox regression analyses, baseline and in-treatment PP and baseline and in-treatment systolic blood pressure predicted new-onset AF, independent of baseline age, height, weight, and Framingham Risk Score; sex, race, and treatment allocation; and in-treatment heart rate and Cornell product. PP was the strongest single blood pressure predictor of new-onset AF determined by the decrease in the -2 Log likelihood statistic, in comparison with systolic blood pressure, diastolic blood pressure, and mean arterial pressure. When evaluated in the same model, the predictive effect of systolic and diastolic blood pressures together was similar to that of PP. In this population of patients with hypertension and left ventricular hypertrophy, PP was the strongest single blood pressure predictor of new-onset AF, independent of other risk factors.
Trial registration: ClinicalTrials.gov NCT00338260.