Lipid-lowering therapy (LLT) decreases mortality in select patient populations. LLT has also been shown to have antiarrhythmic effects, thus favorably influencing the incidence and recurrence of atrial fibrillation (AF). However, data are lacking regarding the effect of LLT on mortality in patients with AF. The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study was the one of the largest multicenter trials comprising of 4,060 patients with AF at high risk for stroke and death. This is a post hoc analysis of the National Heart, Lung, and Blood Institute limited-access dataset of AFFIRM patients who were on LLT at the time of randomization (n = 913). The control group consisted of AFFIRM patients who were not on LLT (n = 3,147). Cox proportional hazards analysis was performed controlling for baseline differences. The end point was all-cause mortality, cardiovascular mortality, and ischemic stroke. A separate analysis was carried out for the combined end point of death, ventricular tachycardia, ventricular fibrillation, cardiac arrest, ischemic stroke, major bleeding, systemic embolism, pulmonary embolism, and myocardial infarction. Patients on LLT were younger and on more cardioactive medications but also had more cardiovascular morbidities. On multivariate analysis, LLT use was associated with lower all-cause mortality (hazard ratio [HR] 0.77, 95% confidence interval [CI] 0.62 to 0.95, p = 0.01), cardiovascular mortality (HR 0.71, 95% CI 0.53 to 0.95, p = 0.02), ischemic stroke (HR 0.56, 95% CI 0.36 to 0.89, p = 0.01), and combined end point (HR 0.81, 95% CI 0.69 to 0.96, p = 0.01). In conclusion, a decrease in mortality and adverse cardiovascular events was observed using LLT in AF.
Although lipid-lowering therapy (LLT) has been shown to decrease morbidity and mortality in primary and secondary prevention populations, no previous study has evaluated whether similar benefits of LLT exist in patients with atrial fibrillation (AF). We hypothesized that use of LLT would positively influence outcomes in patients with AF.
Methods
We performed a post hoc analysis of the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial. A public-use limited-access dataset was obtained from the National Heart, Lung, and Blood Institute (NHLBI). None of the authors are affiliated with the NHLBI or were part of the AFFIRM trial. The NHLBI limited-access dataset is devoid of all records with personal identifiers. Appropriate institutional review board approval was obtained from Wayne State University (Detroit, Michigan).
Details of the AFFIRM study have been described previously. Briefly, it was 1 of the largest studies of AF, involving 4,060 patients with risk factors for stroke or death. Patients were randomized to a rate control (n = 2,027) or a rhythm control (n = 2,033) strategy. This is a post hoc analysis of patients with AF who were on LLT (n = 913) versus those who were not on LLT (n = 3,147) at randomization.
End points of our analysis were all-cause mortality, cardiovascular mortality, and ischemic stroke. A separate analysis was carried out for the combined end point of death, ventricular tachycardia, ventricular fibrillation, cardiac arrest, ischemic stroke, major bleeding, systemic embolism, pulmonary embolism, and myocardial infarction. Only the combined end point was available in the public-access dataset and, hence, further breakdown into individual end points could not be performed.
Categorical and continuous variables were compared using chi-square and analysis of variance tests, respectively. Categorical variables are presented as numbers and percentages, and continuous variables are presented as means ± SDs. Variables included in the initial model included LLT, age, gender, coronary artery disease, heart failure, smoking, stroke or transient ischemic attack, diabetes mellitus, hypertension, myocardial infarction, randomization arm, rhythm at time of randomization, warfarin use, β-blocker use, angiotensin-converting enzyme inhibitor use, and aspirin therapy. Other important clinical variables shown to potentially interact in previous AFFIRM analyses were also included into the model irrespective of their univariate p value. Multivariate Cox regression was carried out using various end points, i.e., all-cause mortality, cardiovascular mortality, ischemic stroke, and combined end point as dependent variables. Variable selection in the model was conducted using stepwise selection. With the goal of having the most parsimonious model, only variables with a p value <0.05 were included in the final model. SAS 9.1 (SAS Institute, Cary, North Carolina) was used to perform statistical analysis and SPSS 17 (SPSS, Inc., Chicago, Illinois) was used to plot graphs.
Results
Patients on LLT were younger, were more likely to be men, have a history of cardiovascular co-morbidities (hypertension, diabetes, coronary artery disease), and be on more beneficial medications (angiotensin-converting enzyme inhibitor, β blockers, and aspirin) compared to patients who were not on LLT ( Table 1 ). Mean follow-up was 3.5 ± 1.2 years per patient. Total deaths were 666 (124 in LLT and 542 in non-LLT groups) including 331 cardiovascular deaths (64 in LLT and 267 in non-LLT groups) occurred during this period. On multivariate analysis, LLT was associated with lower all-cause mortality (hazard ratio [HR] 0.77, 95% confidence interval [CI] 0.62 to 0.95, p = 0.01), cardiovascular mortality (HR 0.71, 95% CI 0.53 to 0.95, p = 0.02), ischemic stroke (HR 0.56, 95% CI 0.36 to 0.89, p = 0.01), and combined end point (HR 0.81, 95% CI 0.69 to 0.96, p = 0.01; Figures 1 to 4 ). Our results are similar to other previous published analyses by the AFFIRM investigators except for inclusion of LLT in the model ( Tables 2 to 5 ).
| Variable | LLT | Non-LLT | p Value |
|---|---|---|---|
| (n = 913) | (n = 3,147) | ||
| Age (years) | 68.7 ± 7.5 | 69.7 ± 8.2 | 0.001 |
| Men | 69.6% | 58.2% | <0.0001 |
| Coronary artery disease | 54.4% | 30.6% | <0.0001 |
| Angina pectoris | 44.9% | 20.2% | <0.0001 |
| Myocardial infarction | 32.3% | 13.0% | <0.0001 |
| Heart failure | 22.8% | 23.2% | 0.78 |
| Hypertension | 77.4% | 68.9% | <0.0001 |
| Peripheral vascular disease | 10.8% | 5.8% | <0.0001 |
| Diabetes mellitus | 26.1% | 18.3% | <0.0001 |
| Smoker | 12.8% | 12.0% | 0.53 |
| Previous stroke | 15.8% | 12.7% | 0.01 |
| Previous coronary artery bypass | 27.8% | 8.1% | <0.0001 |
| Previous percutaneous coronary intervention | 20.0% | 5.4% | <0.0001 |
| Pacemaker | 5.0% | 6.5% | 0.11 |
| Angiotensin-converting enzyme inhibitor use | 44.3% | 37.4% | 0.0002 |
| β blocker use | 51.6% | 40.0% | <0.0001 |
| Digoxin use | 47.5% | 54.7% | 0.0001 |
| Diuretic use | 43.5% | 42.4% | 0.56 |
| Heparin use | 16.4% | 18.0% | 0.27 |
| Warfarin use | 86.1% | 84.1% | 0.15 |
| Aspirin use | 33.6% | 24.6% | <0.0001 |
| Diltiazem use | 30.8% | 30.9% | 0.94 |
| Sinus rhythm at randomization | 55.5% | 53.7% | 0.35 |
| First episode of atrial fibrillation | 33.9% | 36.0% | 0.23 |
| Rhythm control arm | 48.6% | 50.3% | 0.37 |
| Duration of atrial fibrillation (years) | 2.8 ± 1.3 | 2.8 ± 1.3 | 0.14 |
| Variable | HR | 95% CI | p Value |
|---|---|---|---|
| Age | 1.06 | 1.05–1.07 | <0.0001 |
| Coronary artery disease | 1.66 | 1.40–1.98 | <0.0001 |
| Heart failure | 2.08 | 1.76–2.47 | <0.0001 |
| Diabetes mellitus | 1.44 | 1.19–1.72 | <0.0001 |
| Smoker | 1.74 | 1.39–2.17 | <0.0001 |
| Previous stroke | 1.58 | 1.30–1.94 | <0.0001 |
| First episode of atrial fibrillation | 1.24 | 1.06–1.46 | 0.009 |
| Sinus rhythm at randomization | 0.82 | 0.69–0.96 | 0.018 |
| Warfarin therapy | 0.65 | 0.53–0.80 | <0.0001 |
| Lipid-lowering therapy | 0.77 | 0.62–0.95 | 0.01 |
| Variable | HR | 95% CI | p Value |
|---|---|---|---|
| Age | 1.05 | 1.03–1.06 | <0.0001 |
| Coronary artery disease | 1.55 | 1.35–1.78 | <0.0001 |
| Heart failure | 1.72 | 1.50–1.98 | <0.0001 |
| Diabetes mellitus | 1.46 | 1.26–1.69 | <0.0001 |
| Smoker | 1.56 | 1.30–1.87 | <0.0001 |
| Previous stroke | 1.48 | 1.26–1.75 | <0.0001 |
| First episode of atrial fibrillation | 1.24 | 1.09–1.41 | 0.001 |
| Sinus rhythm at randomization | 0.81 | 0.71–0.92 | 0.001 |
| Warfarin therapy | 0.65 | 0.53–0.80 | <0.0001 |
| Lipid-lowering therapy | 0.81 | 0.69–0.96 | 0.01 |
| Variable | HR | 95% CI | p Value |
|---|---|---|---|
| Age | 1.02 | 1.00–1.04 | 0.028 |
| Female gender | 1.70 | 1.22–2.36 | 0.001 |
| Myocardial infarction | 2.04 | 1.39–2.99 | 0.0002 |
| Diabetes mellitus | 1.54 | 1.07–2.22 | 0.02 |
| Previous stroke | 1.65 | 1.09–2.46 | 0.01 |
| Sinus rhythm at randomization | 0.63 | 0.45–0.87 | 0.005 |
| Lipid-lowering therapy | 0.56 | 0.36–0.89 | 0.01 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
