OUP user menu

Association between statin therapy and reductions in atrial fibrillation or flutter and inappropriate shock therapy

Sanjeev P. Bhavnani, Craig I. Coleman, Charles M. White, Christopher A. Clyne, Ravi Yarlagadda, Danette Guertin, Jeffrey Kluger
DOI: http://dx.doi.org/10.1093/europace/eun128 854-859 First published online: 21 May 2008


Aims In patients without implantable cardioverter defibrillators (ICDs), statins have been shown to reduce the incidence of atrial fibrillation and atrial flutter (AF/AFL). We sought to determine if statin therapy could reduce the occurrence of AF/AFL with rapid ventricular rates with and without inappropriate shock therapy among a large heterogeneous ICD cohort.

Methods and results We prospectively followed 1445 consecutive patients receiving an ICD for the primary (n = 833) or secondary (n = 612) prevention from December 1997 through January 2007. Outcome measures include incidence of AF/AFL that initiated ICD therapy or was detected during ICD interrogation. Cox hazard regression analyses were conducted to determine the predictors of AF/AFL with and without inappropriate shock delivery and did not include inappropriate shocks resulting from lead dysfunction or other exogenous factors. Patients in this study (n = 1445) were followed over a mean follow-up period of (mean ± SD) 874 ± 805 days. There were 563 episodes of AF/AFL detected, with 200 episodes resulting in inappropriate shock therapy. Overall, 745 patients received statin therapy and 700 did not. The use of statin therapy was associated with an adjusted hazard ratio of 0.472 [95% confidence interval (CI), 0.349–0.638, P < 0.001] for the development of AF/AFL with shock therapy and 0.613 (95% CI, 0.496–0.758, P < 0.001) without shock therapy when compared with the group without statin use.

Conclusion Among a cohort with ICDs at high risk for cardiac arrhythmias, statin therapy was associated with a reduction in AF/AFL tachyarrhythmia detection and inappropriate shock therapy.

  • Statins
  • Implantable cardioverter defibrillator
  • Atrial tachyarrhythmias
  • Inappropriate shock therapy


Implantable cardioverter defibrillators (ICDs) reduce sudden cardiac death (SCD) among individuals at high risk of ventricular arrhythmias.1,2 Unfortunately, the delivery of ICD shocks is painful, causes patient anxiety, and is correlated with poorer patient perceived quality of life.3,4 Although the delivery of a shock to convert a sustained ventricular arrhythmia (termed appropriate shock) causes the same amount of pain as a shock delivered unnecessarily (termed inappropriate shock), the risk benefit for inappropriate shocks cannot be justified. In ICD patients, paroxysmal atrial fibrillation (AF) is the leading cause of inappropriate shocks.5 Utilizing anti-arrhythmic drugs to prevent inappropriate shocks is limited by the risk of pro-arrhythmia and of adverse effects. Statins have been shown to reduce the occurrence of AF among patients with heart failure and can prevent post-cardiothoracic surgery AF.68 Statin therapy has been postulated to attenuate inflammatory responses by impeding cytokine production and reactive oxygen species generation. Unfortunately, the effect of statins on the occurrence of atrial arrhythmias in the ICD population is currently unknown.

The primary objective of this study was to determine the impact of statin therapy on the occurrence of atrial fibrillation and atrial flutter (AF/AFL) tachyarrhythmia detection with and without inappropriate shock therapy as an adjunct to ICDs among a large heterogeneous cohort of high-risk primary and secondary prevention patients.


Study design and population

Preparation of this report was in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement9 for the reporting of observational studies. All patients undergoing ICD implantation (n = 1445) for the primary (n = 833) or secondary (n = 612) prevention of SCD from 1 December 1997 through 31 January 2007 at a tertiary care, community-based teaching hospital constituted the study cohort. The data were extracted from a prospectively collected database over a (mean ± SD) follow-up of 807 ± 805 days. Type of ICD device implanted, pacemaker functionality (biventricular, single chamber or dual chamber) and programmed detection zones for atrial tachyarrhythmia (according to atrial cycle length and regularity) occurred at the discretion of an electrophysiologist (R.Y., C.A.C., and J.K.), and the device implantation occurred according to published guidelines. Data were gathered on concomitant co-morbidities, treatments, and procedural data and then entered in an electronic medical record designed in Microsoft Access© (2000) by certified nurse practitioners familiar with the database design, input, and management. All data utilized in this analysis were obtained at the time of ICD implantation and data input occurred before the generation of this hypothesis. Implantable cardioverter defibrillator interrogation and rhythm classification were entered only after electrogram evaluation by an electrophysiologist. Statin medication use was determined at the time of ICD implantation. Type and dose of statin therapy was prescribed at the discretion of the primary physician/cardiologist with no restrictions imposed in this analysis. All patients not receiving a statin at the time of ICD implantation were included into the study as controls. Each patient was evaluated 1 week after ICD implantation and approximately every 3 months thereafter until the end of follow-up or patient loss to follow-up. Each outpatient follow-up included a detailed evaluation consisting of a clinical evaluation and an electrocardiogram for baseline rhythm and ICD interrogation. The study received institutional review board approval with a waiver of informed consent.

Implantable cardioverter defibrillator interrogation and rhythm classification

Implantable cardioverter defibrillator interrogation with the analysis of stored electrograms resulting in AF/AFL with either controlled or rapid ventricular rates was reviewed and classified by an electrophysiologist (not blinded to statin allocation) on outpatient follow-up. In patients with dual-chamber devices, AF/AFL was detected by a combination of median atrial cycle length and an atrial electrogram demonstrating a rapid atrial rate greater than a ventricular rate and electrogram morphology with consecutive R-wave-R-wave irregularity for AF, and a fixed atrioventricular (AV) relationship in patients with AFL or atrial tachycardia. In patients with single-chamber devices, ventricular electrogram with QRS morphology similar to sinus rhythm and irregular ventricular rate was required for AF/AFL detection. Detection of AF/AFL occurred only by the analysis of stored electrogram and was not confirmed by electrocardiogram or 24-h Holter monitoring.

Endpoints of the study

The primary endpoints of the study included the first incidence of AF/AFL with rapid ventricular rates resulting in tachyarrhythmia device detection with inappropriate ICD shock therapy and device detection alone without shock therapy. Inappropriate shocks were defined as shocks (singe or cluster) associated with an electrogram failing to exhibit interval and morphological criteria of ventricular tachycardia or ventricular fibrillation and satisfying the above criteria for atrial tachyarrhythmias. It did not include therapy resulting from lead noise, external electromagnetic interferences, or other identifiable exogenous factors.

Statistical analysis

Continuous variables are presented as mean values with standard deviations and were compared using the Student’s t-test or the Mann–Whitney test where appropriate. Dichotomous variables are presented as percentages and compared using the χ2 or Fisher’s exact test where appropriate. Multivariable analysis was performed with the Cox proportional hazard modelling to control for potential confounders. All baseline variables demonstrating a significant association upon univariate analysis (P ≤ 0.20 for inclusion) between the occurrence of the endpoint (dependent variable = AF/AFL with and without inappropriate shock therapy) and treatment and co-morbidity characteristics (independent variables) were entered into the multivariable model. Previously identified independent predictors of inappropriate shock therapy resulting from atrial tachyarrhythmias10 were included in the model regardless of their strength of univariate correlation. The model was adjusted for statin use, age, ischaemic cardiomyopathy, previous myocardial infarction, pre-existing atrial tachyarrhythmias, ejection fraction, heart failure greater than New York Heart Association class I, primary indication for ICD implantation, single chamber, dual chamber, bi-ventricular pacemakers, chronic obstructive pulmonary disease, diabetes, obesity, end-stage renal disease, use of β-blockers, calcium channel blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers, aldosterone receptor antagonists, digoxin, dofetilide, amiodarone, sotalol, clopidigrel, and corticosteroids. In the multivariable model, variables were selected by stepwise, backward elimination, and a P-value of <0.05 was considered significant. Adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated for all independent predictors. All analyses were performed with SPSS 15.0 (SPSS Inc, Chicago, IL, USA).


Baseline characteristics

A total of 1445 patients were eligible for the analysis over the follow-up period (range 1–3284 days). Table 1 compares the baseline characteristics between the statin (n = 745) and no-statin (n = 700) groups prior to ICD implantation. Statin medications (mean dose ± SD, mg/day) were identified as atorvastatin (n = 429, 29.1 ± 22.0), simvastatin (n = 265, 33.1 ± 19.2), lovastatin (n = 26, 28.9 ± 14.2), rosuvastatin (n = 18, 16.4 ± 12.0), and fluvastatin (n = 7, 53.3 ± 26.5). Left ventricular ejection fraction (LVEF) as assessed by echocardiography or radionuclide imaging, history of heart failure greater than NYHA class I, concurrent therapy with amiodarone or sotalol, and type of pacemaker device (single chamber, dual chamber, or bi-ventricular) were not significantly different between groups. In contrast, the group receiving statins was more likely to be of older age, to be male, to have a primary indication for ICD implantation, to have a prior myocardial infarction or a history of coronary intervention (angioplasty/bypass surgery), and to have hypertension. The statin group was more likely to receive β-blockers, aspirin, ACE inhibitors, and angiotensin receptor blockers. The no-statin group was more likely to be on digoxin, have a history of non-ischaemic dilated cardiomyopathy, and have a prior diagnosis of AF (32 vs. 27%, respectively, P = 0.019). Atrial fibrillation had been documented prior to implantation in 425 patients representing 29% of our total population.

View this table:
Table 1

Patient baseline demographic and clinical characteristics by the use of statin medications prior to implantable cardioverter defibrillator implantation

CharacteristicTotal group (n = 1445, %)Statin use (n = 745, %)No statin use (n = 700, %)P-value
Age, years (mean ± SD)65 ± 1367 ± 1163 ± 15<0.001
Male1123 (78%)604 (81%)519 (74%)0.002
Ischaemic cardiomyopathy961 (67%)599 (80%)362 (54%)0.004
Dilated cardiomyopathy215 (15%)66 (9%)149 (21%)<0.001
History of HF > NYHA class I375 ( 26)197 (26%)178 (25%)0.68
LVEF, % (mean ± SD)29 ± 1428.4 ± 1329.3 ± 160.26
 Ejection fraction < 30%984 (68%)520 (70%)464 (66%)0.18
Prior myocardial infarction769 (53%)481 (64%)288 (41%)<0.001
Prior percutaneous coronary intervention163 (11%)118 (16%)45 (6%)<0.001
Prior CABG Surgery554 (38%)350 (47%)204 (29%)<0.001
Prior atrial fibrillation425 (29%)199 (27%)226 (32%)0.02
History of diabetes mellitus416 (29%)230 (31%)186 (27%)0.08
History of hypertension568 (39%)329 (44%)239 (34%)<0.001
End-stage renal disease47 (3.3%)20 (2.7%)27 (3.8%)0.21
History of obesity43 (3.0%)22 (2.9%)21 (3.0%)0.95
History of COPD193 (13%)97 (13%)96 (14%)0.68
History of pulmonary hypertension28 (1.9%)10 (1.3%)18 (2.6%)0.09
History of cerebrovascular disease97 (6.7%)50 (6.7%)47 (6.7%)>0.99
History of hypothyroidism166 (11.5%)81 (10.9%)85 (12.2%)0.44
Primary indication for ICD implantation833 (58%)457 (61%)376 (54%)0.004
Single-chamber pacemaker418 (29%)213 (29%)205 (29%)0.75
Dual-chamber pacemaker672 (47%)349 (47%)323 (46%)0.83
Bi-ventricular pacemaker354 (24%)183 (25%)171 (24%)0.98
Medications therapy prior to ICD implantation
 β-blockers1147 (79%)620 (83%)527 (75%)<0.001
 CCB100 (7%)55 (7%)45 (6%)0.48
 Aspirin737 (51%)461 (62%)276 (39%)<0.001
 Warfarin546 (38%)268 (36%)278 (40%)0.13
 Clopidrigrel167 (11.6%)121 (16.2%)46 (6.6%)<0.001
 ACE-inhibitors705 (49%)386 (52%)319 (46%)0.02
 Aldosterone receptorantagonists389 (27%)212 (28%)177 (25%)0.32
 ARB246 (17%)146 (20%)100 (14%)0.007
 Digoxin461 (32%)219 (29%)242 (35%)0.03
 Amiodarone216 (15%)107 (14%)109 (16%)0.51
 Sotalol156 (11%)82 (11%)74 (11%)0.80
 Dofetilide27 (1.8%)16 (2.1%)11 (1.6%)0.42
 Diuretics831 (58%)414 (56%)417 (60%)0.10
 Hydralazine40 (2.8%)22 (2.9%)18 (2.6%)0.67
 Nitrates122 (8%)68 (9%)54 (8%)0.34
 Corticosteroid use64 (4.4%)30 (4.0%)34 (4.8%)0.44
Duration of follow-up (days)874 ± 805861 ± 816889 ± 7930.51
  • HF, heart failure; NYHA, New York Heart Association; CABG, coronary artery bypass graft; ICD, implantable cardioverter defibrillator; COPD, chronic obstructive pulmonary disease; LVEF, left ventricular ejection fraction; CCB, calcium channel blocker; ACE, angiotensin-converting enzyme; and ARB, angiotensin receptor blocker.

Statin effect on atrial fibrillation and atrial flutter with and without implantable cardioverter defibrillator shock therapy

During the follow-up period, 563 patients (39.0% of the overall population) had an episode of AF/AFL tachyarrhythmia detection among which 200 patients had an episode of AF/AFL resulting in inappropriate ICD shock therapy, and 363 had an episode of AF/AFL detection without shock therapy representing 13.8 and 25.1% of the 1445 patients, respectively. The median device-detected ventricular rate of AF/AFL resulting in inappropriate shock therapy was 216 (range 175–222) and 150 b.p.m. (range 100–164) for device detection without shock delivery. The impact of statin therapy on the duration of AF/AFL episodes was not assessed.

The overall incidence of AF/AFL detection with inappropriate (72 vs. 128 patients) and without inappropriate (150 vs. 213 patients, P < 0.001 for both) shock therapy was lower among the statin-treated group when compared with the no-statin group, respectively. The group receiving statin therapy was associated with a reduced incidence of both AF/AFL with and without shock delivery [crude HR; 0.48 (95% CI, 0.35–0.65) and crude HR; 0.58 (95% CI, 0.45–0.73), P < 0.001 for both] when compared with the no-statin group upon univariate analysis.

Figures 1A and B depict the Kaplan–Meier curves for the cumulative hazard of AF/AFL detection with and without ICD shock delivery between the statin and no-statin groups. The use of statins was associated with an adjusted HR of 0.47 (95% CI, 0.35–0.64, P < 0.001) for the development of AF/AFL with shock therapy and adjusted HR of 0.61 (95% CI, 0.50–0.76, P < 0.001) without shock therapy when compared with the group without statin use. This effect appeared to be dose-related, with a 10 mg increase in atorvastatin dose equivalent (utilizing conventional dose conversion calculations between different statins) being associated with a 9.0% (95% CI, 2–17%, P = 0.048) reduction in the development of AF/AFL with shock therapy.

Figure 1

Kaplan–Meier curves depicting the effect of statin use on (A) cumulative hazard of atrial fibrillation and atrial flutter with inappropriate implantable cardioverter defibrillator shock therapy and (B) without shock therapy. AF, atrial fibrillation; AFL, atrial flutter; CI, confidence interval.

Multivariable analysis (where adjustment for the observational imbalances between groups are incorporated) of the characteristics demonstrating an independent association with reductions in AF/AFL with and without inappropriate shock therapy in addition to statin therapy is illustrated in Tables 2 and 3.

View this table:
Table 2

Independent predictors of shock therapy resulting from atrial fibrillation/atrial flutter

VariableP-valueHR95% CI
Prior statin use<0.0010.470.350.64
Age < 600.030.710.520.96
Single-chamber pacemaker0.031.401.041.88
Prior digoxin use0.011.451.081.96
History of myocardial infarction0.011.501.112.02
History of atrial fibrillation0.0011.671.242.26
Prior dofetilide use0.012.561.255.22
View this table:
Table 3

Independent predictors of atrial fibrillation/atrial flutter detection without shock therapy

VariableP-valueHR95% CI
Prior statin use<0.0010.610.500.76
Primary indication for ICD implantation0.011.341.081.67
History of atrial fibrillation0.0011.441.151.80
Single-chamber pacemaker0.0041.471.131.90
Bi-ventricular pacemaker0.011.551.122.15
History of chronic obstructive pulmonary disease0.0021.541.162.04


A recent meta-analysis demonstrated a 45% reduction in odds of developing AF with statin therapy in a non-ICD population.11 This is similar to a study by Hanna et al.6 which reported the results of a registry of 25 268 patients with reduced LVEF (≤40%) and demonstrated a 31% reduced odds of AF prevalence with statins.

In a nested cohort evaluation of the SCD in heart failure trial, Dickinson and colleagues11 reported a 28% reduction in incidence of AF/AFL; however, not all patients received an ICD and the results may not be generalized to a heterogeneous ICD population. In our study, statin therapy was associated with a significant 52.8% hazard reduction in rapid AF/AFL with inappropriate ICD shock therapy and 38.7% reduction in rapid AF/AFL detection among a large heterogeneous cohort of ICD recipients at risk for tachyarrhythmias.

Several clinical investigations have demonstrated reductions in different types of AF with statin therapy including post-operative,12 recurrent,13 recurrent after cardioversion,14 and new-onset AF.15 Our observations of an increased efficacy rate of statins on AF/AFL tachyarrhythmias with shock therapy when compared with detection of rapid rates alone may imply that statins exert an increased anti-arrhythmic effect at higher atrial rates. In a preliminary study,16 statin therapy reduced the incidence of AF and atrial high-rate episodes (defined as atrial rates >180 b.p.m.) among individuals with pacemaker implantation. Statins may exhibit a rate-dependent property which may suppress the occurrence of atrial tachyarrhythmias but not necessarily influence the ventricular response which is dependent on autonomic tone and intrinsic AV nodal properties. Our associated findings of a dose–response relationship may favour the likelihood of a drug-related effect rather than an increased risk for the development of AF/AFL among individuals with structural heart disease treated with higher doses of statins. Furthermore, reducing the device-detected cycle length during AF/AFL tachycardia episodes may also decrease subsequent inappropriate shock therapies.

In addition to statin therapy, we found several other positive- or negative-independent predictors or predictors that trended towards significance for AF/AFL tachyarrhythmias. A primary prevention indication for ICD implantation demonstrated an increased risk of AF/AFL tachyarrhythmia and trended towards increased risk for inappropriate ICD shock which may be due to the high prevalence of decreased ejection fraction and coronary artery disease, known risk factors for AF.17 The presence of bi-ventricular pacing was a positive predictor for inappropriate shocks. This is largely represented by significantly greater rates of left ventricular dysfunction, a history of AF, myocardial infarction, and a primary indication for ICD implantation as risk factors for the development of AF/AFL among these individuals when compared with those without bi-ventricular pacing. Single-chamber pacing was also a positive predictor for inappropriate shocks. Several investigations have associated asynchronous ventricular pacing with AF.1820 Prolonged right ventricular (RV) apical pacing may lead to atrial electrical remodelling, atrial dilatation, and disruption of AV synchrony,21 particularly when imposed on left ventricular dysfunction.22 Asynchronous ventricular pacing may induce retrograde atrial activation and changes in regional myocardial perfusion that may potentiate AF. Recently within the inhibition of unnecessary RV pacing with AV search hysteresis in the ICD study,23 atrial tachyarrhythmias accounted for 27% of all inappropriate shocks with similar rates of 48 and 46% in the dual chamber universal, rate responsive and ventricular pacing groups, respectively. In our cohort, increased β-blocker use may have led to bradycardia-induced pacing and thus indirectly increasing the risk of AF. Mechanically, inappropriate detection of high-atrial rates due to AF/AFL is mainly due to far-field R-wave oversensing, erroneous pacing, and inappropriate mode switching.24,25 Dual-chamber ICDs capable of automatic atrial shock therapies and atrial antitachycardia pacing are possible options for the management of AF as either adjunct to pharmacological or electrical-based therapy.26,27 Recent advances in dynamic AV algorithms to avoid unnecessary pacing and promote intrinsic conduction to mitigate the effects of atrial tachyarrhythmias are promising.


Several limitations merit consideration when the results of the present study are interpreted. First, these results are inherently limited by the retrospective and observational design of this investigation. Although we attempted to control for potential confounders using multivariate statistical analysis, we could not control for unknown confounders. However, having AF/AFL results in the same direction as the previous statin and AF/AFL studies support our methodology and efficacy results. Next, statin use was not defined as a time-dependent covariate. The individuals stratified within the no-statin group may have received statin therapy during the follow-up period. This may inadvertently underestimate the impact of statin therapy on AF/AFL. It should also be noted that both therapy for and detection of AF/AFL were dependent on the programming of VT detection rates, supraventricular tachycardia discriminators, and pacing modes that were not uniform and was at the discretion of the implanting or following electrophysiologist. Our data were adjusted on the basis of commonly accepted independent predictors of AF as identified in the American College of Cardiology/American Heart Association/European Society of Cardiology guidelines;28 however, the possibility exists that other extraneous factors might influence our results. Left atrial area, presence of mitral regurgitation, and atrial and ventricular dimension as risk factors for the development of AF/AFL were not measured and hence not included in this analysis. Atorvastatin was identified as the most commonly used statin, representing 62.1% of the overall statin usage, and also represented the majority of statins with hydrophilic (98.4%) and synthetic (94.5%) properties, and therefore analyses of individual statins and statin classes were not determined. Accurate determination of AF/AFL as detected by single-chamber pacing devices is limited by the lack of an atrial lead and analysis of the atrial electrogram. Morphology discrimination programmes are heterogeneous between pacemaker models, which may lead to variable detection rates of rapid AF/AFL by these devices.29 Furthermore, we did not assess an average percentage of ventricular pacing per patient. No patient was receiving prevastatin, and this likely represents a selection bias by prescribing physicians. Data regarding cholesterol levels or inflammatory markers were not uniformly available. Finally, this investigation does not provide insight into a precise mechanism of AF/AFL-related risk among ICD recipients or the mechanism conferring a benefit with a statin-based therapy.

Angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and aldosterone receptor antagonists are emerging as potential non-arrhythmic therapies for atrial tachyarrhythmias. Statin therapy has been postulated to augment the inflammatory response by reductions in C-reactive protein,30,31 reactive oxygen species generation,32 and modulation of nitrosative stress33 and has been hypothesized among the mechanistic roles in the prevention of AF. Statins have demonstrated benefit for the primary and secondary prevention of myocardial infarction and have been shown to reduce heart failure hospitalization among a cohort with systolic dysfunction.34 Whether the reduction in AF/AFL with statin therapy is primarily due to an electrophysiological mechanism or rather represents a surrogate marker of reverse remodelling is currently unknown. Future prospective investigations evaluating the anti-arrhythmic effects of statins, ideally in patients with normal cholesterol levels such that the lipid-lowering effects of these drugs will be minimized are needed. Such investigations should also assess the impact of statin dose on the development and recurrence of AF/AFL among individuals at risk for SCD.


Our data suggest that statin use is associated with significant reductions of AF/AFL tachyarrhythmias with and without inappropriate shock therapy among a cohort of ICD patients at high risk for cardiac arrhythmias.

Conflict of interest: none declared.


This evaluation was not funded.


  • Preliminary data in this paper were presented as an E-poster at the American Heart Association Scientific Sessions, Orlando, FL, USA, 4–7 November 2007.


View Abstract