Effect modification of aortic atheroma on the prognostic value of heart rate in hypertension

Background: Although some epidemiological studies have advocated a prognostic value of heart rate (HR) in hypertensive patients, the influence of vascular damages on this prognostic value has not been tested yet. Methods: HRs were collected by pulse palpation in 1204 primary hypertensive patients in sinus rhythm without cardiac-slowing drugs. Aortic damages were assessed by aortography, whereas cardiac disease was assessed by medical history, symptoms and electrocardiogram. Results: In a multivariable Cox model adjusted for major confounders, HR was of prognostic significance for all-cause [hazard ratio 1.12 (1.06-1.19) for 10bpm increment and 1.39 (1.18-1.64) for HR >= 82 vs. <82 bpm] and cardiovascular death [hazard ratio 1.10 (1.02-1.20) for 10bpm increment and 1.37 (1.09-1.72) for HR >= 82 vs. <82 bpm] after 35 years of follow-up. This association was particularly manifested at 15 years of follow-up. At that time, with the same multivariable survival model, the association between HR and cardiovascular death was stronger in patients with aortic atheroma [2.76 (1.47-5.18) for an HR >= 82 vs. <82 bpm] than in patients without [hazard ratio 1.36 (0.76-2.43) for an HR >= 82 vs. <82 bpm, P for interaction = 0.054]. Similarly, the association between HR and cardiovascular death was stronger in patients with an overt cardiac disease than those without (P for interaction = 0.044). Conclusion: In hypertensive patients, the prognostic significance of HR for cardiovascular outcome is modulated by the presence of aortic atherosclerosis or cardiac disease. This should prompt us to a thorough examination of cardiovascular damages in hypertensive patients when HR is elevated.