Erik O Udo1, Norbert M van Hemel2, Nicolaas P A Zuithoff3, Pieter A Doevendans2, Karel G M Moons3. 1. Department of Cardiology, UMC Utrecht, The Netherlands, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands; Julius Center for Health Sciences and Primary Care, UMC Utrecht, The Netherlands, Universiteitsweg 100, 3584 CG Utrecht, The Netherlands. Electronic address: e.o.udo@umcutrecht.nl. 2. Department of Cardiology, UMC Utrecht, The Netherlands, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands. 3. Julius Center for Health Sciences and Primary Care, UMC Utrecht, The Netherlands, Universiteitsweg 100, 3584 CG Utrecht, The Netherlands.
Abstract
BACKGROUND: Right ventricular pacing (RVP) is associated with an increased risk of heart failure (HF) events. However, the extent and shape of this association is hardly assessed. OBJECTIVE: We quantified whether the undesired effects of RVP are confirmed in an unselected population of first bradycardia pacemaker recipients. Furthermore, we studied the shape of the association between RVP and HF death and cardiac death. METHODS: Cumulative percentage RVP (%RVP) was measured in 1395 patients. Using multivariable Cox regression analysis with %RVP as time-dependant co-variate we evaluated the association between %RVP and HF- and cardiac death, both unadjusted and adjusted for confounders, including age, gender, pacemaker-indication, cardiac disease, HF at baseline, diabetes, hypertension, atrio-ventricular synchrony, usage of beta-blocking drugs, anti-arrhythmic medication, HF medication, and prior atrial fibrillation/flutter. Non-linear associations were evaluated with restricted cubic splines. RESULTS: During a mean follow-up of 5.8 (SD 1.1) years 104 HF deaths and 144 cardiac deaths were observed. %RVP was significantly associated with HF- and cardiac death in both unadjusted (p<0.001 and p<0.001, respectively) and adjusted analyses (p=0.046 and p=0.009, respectively). Our results show a linear association between %RVP and HF- and cardiac death. We observed a constant increase of 8% risk of HF death per 10% increase in RVP. A model incorporating various non-linear transformations of %RVP using restrictive cubic splines showed no improved model fit over linear associations. CONCLUSION: This long-term, prospective study observed a significant, though linear association between %RVP and risk of HF death and/or cardiac death in unselected bradycardia pacing recipients.
BACKGROUND: Right ventricular pacing (RVP) is associated with an increased risk of heart failure (HF) events. However, the extent and shape of this association is hardly assessed. OBJECTIVE: We quantified whether the undesired effects of RVP are confirmed in an unselected population of first bradycardia pacemaker recipients. Furthermore, we studied the shape of the association between RVP and HF death and cardiac death. METHODS: Cumulative percentage RVP (%RVP) was measured in 1395 patients. Using multivariable Cox regression analysis with %RVP as time-dependant co-variate we evaluated the association between %RVP and HF- and cardiac death, both unadjusted and adjusted for confounders, including age, gender, pacemaker-indication, cardiac disease, HF at baseline, diabetes, hypertension, atrio-ventricular synchrony, usage of beta-blocking drugs, anti-arrhythmic medication, HF medication, and prior atrial fibrillation/flutter. Non-linear associations were evaluated with restricted cubic splines. RESULTS: During a mean follow-up of 5.8 (SD 1.1) years 104 HF deaths and 144 cardiac deaths were observed. %RVP was significantly associated with HF- and cardiac death in both unadjusted (p<0.001 and p<0.001, respectively) and adjusted analyses (p=0.046 and p=0.009, respectively). Our results show a linear association between %RVP and HF- and cardiac death. We observed a constant increase of 8% risk of HF death per 10% increase in RVP. A model incorporating various non-linear transformations of %RVP using restrictive cubic splines showed no improved model fit over linear associations. CONCLUSION: This long-term, prospective study observed a significant, though linear association between %RVP and risk of HF death and/or cardiac death in unselected bradycardia pacing recipients.