Many medications used to treat atrial fibrillation (AF) also reduce blood pressure (BP). The relation between BP and mortality is unclear in patients with AF. We performed a post hoc analysis of 3,947 participants from the Atrial Fibrillation Follow-Up Investigation of Rhythm Management trial. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) at baseline and follow-up were categorized by 10-mm Hg increments. The end points were all-cause mortality (ACM) and secondary outcome (combination of ACM, ventricular tachycardia and/or fibrillation, pulseless electrical activity, significant bradycardia, stroke, major bleeding, myocardial infarction, and pulmonary embolism). SBP and DBP followed a “U-shaped” curve with respect to primary and secondary outcomes after multivariate analysis. A nonlinear Cox proportional hazards model showed that the incidence of ACM was lowest at 140/78 mm Hg. Subgroup analyses revealed similar U-shaped curves. There was an increased ACM observed with BP <110/60 mm Hg (hazard ratio 2.4, p <0.01, respectively, for SBP and DBP). In conclusion, in patients with AF, U-shaped relation existed between BP and ACM. These data suggest that the optimal BP target in patients with AF may be greater than the general population and that pharmacologic therapy to treat AF may be associated with ACM or adverse events if BP is reduced to <110/60 mm Hg.
The seventh report of the Joint National Committee considers a blood pressure (BP) of <120/80 mm Hg (systolic or diastolic) to be “normal” or “optimal.” The cardiovascular risk is known to double with each 20/10 mm Hg increment in BP >115/75 mm Hg in the elderly. This report also outlines a linear relation of systolic blood pressure (SBP) and diastolic blood pressure (DBP) with cardiovascular mortality. However, in certain populations such as the elderly and in patients with acute coronary syndrome, a “J-shaped” relation between BP and outcomes has been observed. Low BP (<110/70 mm Hg) has been shown to be associated with an increase in adverse events, with the lowest mortality demonstrable in the BP range 130 to 140/80 to 90 mm Hg. Similar findings have also been demonstrated with chronic coronary artery disease (CAD), hypertension, and stroke. However, the “optimal” or “goal” BP in atrial fibrillation (AF) has never been studied in the past and had not been addressed in the Joint National Committee-8 guidelines. Given the loss of atrial contractility, the optimal BP in patients with AF may differ from the general population. Additionally, most medications used to treat AF, for either rhythm or rate control, result in decreased BPs. Therefore, it would be critical to define not only optimal BP but thresholds in BP below which adverse events may increase. Using patients enrolled in the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial, we hypothesized that patients with AF have a U-shaped relation between BP and outcomes. Furthermore, the goal of this study was to determine the optimal BP in patients with AF and also define lower boundaries of BP control at which mortality and/or adverse events increase.
Methods
We performed a post hoc analysis of patients enrolled in the AFFIRM trial. The details of AFFIRM trial have been described previously. In brief, AFFIRM was a prospective trial (n = 4,060) comparing survival in patients with AF and at least 1 risk factor for stroke randomized to a strategy of rate control (n = 2,027) versus a strategy of rhythm control (n = 2,033). Our inclusion criteria included subjects who had initial (baseline) BP recordings and at least 1 other BP recording during the first year of follow-up. There were 3,947 patients (with available BP data) who were included into our study sample. Patients with limited or no BP data were excluded (n = 113). BP measurements were consistent with the American Heart Association’s scientific statement on human BP determination by sphygmomanometer. Per study protocol, the BP was recorded in the preferred arm after the patient has been sitting quietly for at least 5 minutes.
The average BP was defined as the average of all available BP measurements taken during each postbaseline visit. SBP and DBP were categorized into 10-mm Hg increments to study the association between BP and primary and secondary clinical outcomes. The primary end point of our analysis was all-cause mortality (ACM). The secondary outcome considered was a composite of ACM and events including sustained ventricular tachycardia, ventricular fibrillation, pulseless electrical activity, concurrent AF, bradycardia, stroke, major bleeding, myocardial infarction, and pulmonary embolism. Our definitions remained consistent with the original trial. Follow-up data were available at 2, 4, 8, and 12 months and then 3 visits per year until a period of 6 years or study termination.
Baseline characteristics of the study population were compared across the SBP and DBP groups by the chi-square test for categorical variables and the 1-way analysis of variance or the Kruskal-Wallis test for continuous variables, depending on the distribution of the variable.
We hypothesized that if a J- or U-shaped relation is discovered between BP and outcome, it is more likely to be seen with average on-treatment follow-up BP (which is closer to the patient’s “actual” long-term BP) rather than a single baseline BP value. We also conducted formal tests of linearity for the relation between BP and outcomes. The R 2 of the quadratic model of SBP was significantly better than that of the linear model of SBP (p <0.001). Thus, the model with SBP as a quadratic function gives a better fit. Similar results were obtained using DBP and for secondary outcome as well. Hence, we decided to define the relation between average on-treatment BP and outcomes as a quadratic nonlinear relation. A nadir BP was calculated with the delta method, which is equal to coefficient of the linear term divided by 2 times the coefficient of the quadratic (square) term. The value of nadir BP was then used to determine the range of SBP and DBP at which the event rate would be the lowest, and this range of SBP (130–140 mmHg) and DBP (70–80 mmHg) was subsequently used as the referent group in the Cox proportional hazards models.
We created 2 separate adjusted models using baseline BP and average on-treatment follow-up BP. The predictive value for each model was calculated using Harrell’s concordance index (C index), and a comparison was drawn between them using bootstrapping. Because the predictive value of models with average on-treatment BP was significantly better than that of baseline BP models, all subsequent mention of BP relates to average follow-up BP and not to baseline BP.
Univariate Cox proportional hazards analysis was performed to investigate individual variables significantly associated with primary and secondary outcomes. Variables with p <0.2 in the initial univariate screen and those believed to be clinically relevant were included in the multivariate model. Initially, we constructed multivariate Cox proportional hazards models incorporating all potential confounding variables. Subsequently, we created reduced models including only variables with p <0.05 using backward regression techniques. These reduced models showed similar results and adjusted hazard ratio (HR) values compared with the full model. We decided to use the reduced model for our analysis to avoid “overfitting.” The variables included in the final multivariate models were BP group, age, history of hypertension, history of congestive heart failure (CHF), history of myocardial infarction, history of revascularization, history of stroke, history of diabetes, smoking status, use of warfarin, lipid-lowering therapy, diuretics, and randomized (rate vs rhythm) treatment group. This analysis was performed for the complete cohort and was also performed separately for the rate and the rhythm control arms. The adjusted HR for each category of BP was calculated in reference to the SBP and DBP range in which the event rate was the lowest, that is, for that group, the HR was considered as 1. We thoroughly checked for interactions and tested for collinearity using variance inflation factor, with variance inflation factor >10 signifying collinearity. No significant interactions or multicollinearity was observed. The variance inflation factor for all variables in the model was <10.
To evaluate the relation between SBP (in a continuous manner) and HR of primary and secondary outcomes, we performed restricted cubic spline analysis for the Cox proportional hazards model using the covariates listed previously. Separate subgroup analyses were also performed in important subgroups. A p value of <0.05 was considered statistically significant. All analyses were performed using the Stata software, version 11.0 (StataCorp LP, College Station, Texas).
Results
The baseline characteristics of the study population, stratified by average SBP and DBP ranges, are listed in Tables 1 and 2 . Participants with a lower mean SBP were likely to be younger, men, and have a lower body mass index, history of CAD, and history of myocardial infarction and/or CHF. Participants with a lower mean DBP were more likely to be older, leaner, and have a history of CAD, myocardial infarction, peripheral vascular disease, and diabetes. Lower SBP and DBP were significantly associated with a decreased left ventricular ejection fraction (LVEF; <30%). Higher SBP and DBP were significantly associated with a normal LVEF (>50%). Use of angiotensin-converting enzyme inhibitors, diuretics, calcium channel blockers, and other antihypertensive drugs was associated with a higher SBP. Use of diuretics, other antihypertensive drugs, digoxin, and/or amiodarone was associated with lower DBP. The mean follow-up period was 42.4 ± 14.7 months.
| Variable | Systolic Blood Pressure (mm Hg) | p Value | ||||||
|---|---|---|---|---|---|---|---|---|
| ≤110 (n = 143) | >110 to ≤120 (n = 426) | >120 to ≤130 (n = 849) | >130 to ≤140 (n = 1107) | >140 to ≤150 (n = 854) | >150 to ≤160 (n = 383) | >160 (n = 185) | ||
| Age (years) | 68.6 ± 9.5 | 69.0 ± 8.4 | 69.3 ± 8.2 | 69.1 ± 8.1 | 69.9 ± 7.8 | 70.4 ± 7.9 | 70.7 ± 7.5 | 0.02 |
| Men | 108 (75.5%) | 289 (67.8%) | 555 (65.3%) | 673 (60.7%) | 478 (55.9%) | 206 (53.7%) | 93 (50.2%) | <0.01 |
| Body mass index (kg/m 2 ) | 26.3 ± 5.4 | 28.3 ± 5.7 | 28.7 ± 5.6 | 29.1 ± 6.2 | 29.5 ± 6.1 | 29.3 ± 6.5 | 29.0 ± 5.9 | 0.001 |
| Hypertension | 63 (44.0%) | 221 (51.8%) | 510 (60.0%) | 806 (72.8%) | 695 (81.3%) | 343 (89.5%) | 170 (91.8%) | <0.01 |
| Coronary artery diseases | 78 (54.5%) | 190 (44.6%) | 336 (39.5%) | 396 (35.7%) | 295 (34.5%) | 138 (36.0%) | 72 (38.9%) | <0.01 |
| Myocardial infarction | 51 (35.6%) | 104 (24.4%) | 161 (18.9%) | 169 (15.2%) | 117 (13.7%) | 62 (16.1%) | 25 (13.5%) | <0.01 |
| Revascularization | 39 (27.2%) | 100 (23.4%) | 175 (20.6%) | 186 (16.8%) | 138 (16.1%) | 68 (17.7%) | 32 (17.3%) | 0.001 |
| Heart failure | 68 (47.5%) | 121 (28.4%) | 201 (23.6%) | 220 (19.8%) | 172 (20.1%) | 77 (20.1%) | 42 (22.7%) | <0.01 |
| NYHA HF class | <0.01 | |||||||
| 0 | 83 (58.0%) | 323 (75.8%) | 679 (79.9%) | 922 (83.2%) | 699 (81.8%) | 316 (82.5%) | 148 (80.4%) | |
| 1 | 24 (16.7%) | 48 (11.2%) | 88 (10.3%) | 108 (9.7%) | 98 (11.4%) | 38 (9.9%) | 23 (12.5%) | |
| 2 | 22 (15.3%) | 46 (10.8%) | 63 (7.4%) | 64 (5.7%) | 46 (5.3%) | 25 (6.5%) | 13 (7.0%) | |
| 3 | 14 (9.7%) | 9 (2.1%) | 19 (2.2%) | 13 (1.1%) | 11 (1.2%) | 4 (1.0%) | 0 (0%) | |
| Peripheral vascular disease | 10 (7%) | 33 (7.8%) | 50 (5.9%) | 72 (6.5%) | 52 (6.1%) | 39 (10.2%) | 11 (6%) | 0.14 |
| Prior stroke | 22 (15.4%) | 60 (14.1%) | 108 (12.7%) | 134 (12.1%) | 121 (14.2%) | 49 (12.8%) | 30 (16.2%) | 0.62 |
| Smoker | 26 (18.2%) | 66 (15.5%) | 101 (11.9%) | 134 (12.1%) | 90 (10.5%) | 43 (11.2%) | 21 (11.4%) | 0.07 |
| Diabetes mellitus | 26 (18.2%) | 69 (16.2%) | 149 (17.6%) | 199 (18%) | 192 (22.5%) | 105 (27.4%) | 51 (27.6%) | <0.01 |
| Pacemaker | 17 (11.9%) | 31 (7.3%) | 56 (6.6%) | 67 (6.1%) | 49 (5.7%) | 17 (4.4%) | 8 (4.3%) | 0.05 |
| Aspirin | 39 (27.3%) | 123 (28.9%) | 216 (25.4%) | 284 (25.7%) | 214 (25.1%) | 109 (28.5%) | 52 (28.1%) | 0.68 |
| Warfarin | 124 (86.7%) | 372 (87.3%) | 712 (83.9%) | 962 (86.9%) | 715 (83.7%) | 317 (82.8%) | 151 (81.6%) | 0.11 |
| Lipid lowering drug | 33 (23.1%) | 102 (23.9%) | 200 (23.6%) | 269 (24.3%) | 175 (20.5%) | 76 (19.8%) | 39 (21.1%) | 0.34 |
| Beta blocker | 42 (35%) | 146 (41.6%) | 242 (36.5%) | 325 (37.9%) | 257 (38.1%) | 105 (36.6%) | 58 (40.6%) | 0.72 |
| ACE inhibitors | 67 (46.9%) | 153 (35.9%) | 287 (33.8%) | 405 (36.6%) | 343 (40.2%) | 181 (47.3%) | 106 (57.3%) | <0.01 |
| Diuretics | 74 (51.8%) | 181 (42.5%) | 335 (39.5%) | 435 (39.3%) | 374 (43.8%) | 180 (47%) | 96 (51.9%) | 0.001 |
| Other antihypertensive | 48 (33.6%) | 132 (31%) | 236 (27.8%) | 355 (32.1%) | 278 (32.6%) | 139 (36.3%) | 69 (37.3%) | 0.04 |
| CCBs | 30 (25%) | 84 (23.9%) | 219 (33.1%) | 295 (34.4%) | 219 (32.4%) | 97 (33.9%) | 53 (37.3%) | 0.006 |
| Digoxin (%) | 87 (60.8%) | 227 (53.3%) | 446 (52.6%) | 593 (53.6%) | 434 (50.8%) | 216 (56.4%) | 89 (48.1%) | 0.18 |
| Amiodarone (%) | 29 (20.3%) | 83 (19.5%) | 136 (16%) | 208 (18.8%) | 150 (17.6%) | 78 (20.4%) | 37 (20%) | 0.46 |
| Sotalol (%) | 18 (12.6%) | 56 (13.2%) | 121 (14.3%) | 156 (14.1%) | 140 (16.4%) | 76 (19.8%) | 33 (17.8%) | 0.06 |
| Class 1 antiarrhythmic | 12 (8.4%) | 43 (10.1%) | 104 (12.3%) | 150 (13.6%) | 122 (14.3%) | 50 (13.1%) | 25 (13.5%) | 0.25 |
| Ejection fraction ∗ | <0.01 | |||||||
| >50% | 37 (37.8%) | 193 (66.1%) | 490 (76.8%) | 632 (75.6%) | 519 (78.8%) | 209 (74.9%) | 114 (78.1%) | |
| 40%–49% | 15 (15.3%) | 33 (11.3%) | 73 (11.4%) | 120 (14.4%) | 71 (10.8%) | 45 (16.1%) | 22 (15.1%) | |
| 30%–39% | 21 (21.4%) | 32 (11%) | 41 (6.4%) | 59 (7.1%) | 50 (7.6%) | 19 (6.8%) | 8 (5.5%) | |
| <30% | 25 (25.5%) | 34 (11.6%) | 34 (5.3%) | 25 (3%) | 19 (2.9%) | 6 (2.2%) | 2 (1.4%) | |
| Ventricular septal thickness (mm) ∗ | 1.0 ± 0.2 | 1.0 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | <0.01 |
| LV posterior wall thickness (mm) ∗ | 1.0 ± 0.2 | 1.0 ± 0.1 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 0.002 |
| Left atrial diameters (mm) ∗ | 4.4 ± 0.7 | 4.4 ± 0.7 | 4.3 ± 0.7 | 4.3 ± 0.6 | 4.3 ± 0.6 | 4.3 ± 0.6 | 4.2 ± 0.6 | 0.05 |
∗ Echocardiographic data available for 2,948 patients (74.7% of entire study population).
| Variable | Diastolic Blood Pressure (mm Hg) | p Value | ||||
|---|---|---|---|---|---|---|
| ≤60 (n = 90) | >60 to ≤70 (n = 804) | >70 to ≤80 (n = 1986) | >80 to ≤90 (n = 964) | >90 (n = 103) | ||
| Age (years) | 72.5 ± 8.2 | 71.8 ± 7.1 | 70.1 ± 7.6 | 66.8 ± 8.6 | 62.8 ± 9.0 | 0.0001 |
| Male gender | 52 (57.8%) | 491 (61.1%) | 1,178 (59.3%) | 605 (62.8%) | 76 (73.8%) | 0.03 |
| Body mass index (kg/m 2 ) | 26.2 ± 6.0 | 28.0 ± 5.6 | 28.8 ± 6.0 | 30.1 ± 6.0 | 30.6 ± 6.5 | 0.0001 |
| Hypertension | 57 (63.3%) | 469 (58.3%) | 1,390 (70%) | 795 (82.5%) | 97 (94.2%) | <0.01 |
| Coronary artery diseases | 57 (63.3%) | 397 (49.4%) | 722 (36.4%) | 300 (31.1%) | 29 (28.2%) | <0.01 |
| Myocardial infarction | 34 (37.8%) | 206 (25.6%) | 316 (15.9%) | 121 (126%) | 12 (11.6%) | <0.01 |
| Revascularization | 37 (41.1%) | 221 (27.5%) | 335 (16.9%) | 137 (14.2%) | 8 (7.8%) | <0.01 |
| Heart failure | 52 (57.8%) | 243 (30.2%) | 417 (21%) | 167 (17.3%) | 22 (21.4%) | <0.01 |
| NYHA HF class | <0.01 | |||||
| 0 | 48 (53.3%) | 585 (72.8%) | 1,624 (81.8%) | 826 (85.8%) | 87 (84.5%) | |
| 1 | 19 (21.1%) | 113 (14.1%) | 210 (10.6%) | 79 (8.2%) | 6 (5.8%) | |
| 2 | 17 (18.9%) | 79 (9.8%) | 122 (6.1%) | 53 (5.5%) | 8 (7.8%) | |
| 3 | 6 (6.7%) | 27 (3.4%) | 30 (1.5%) | 5 (0.5%) | 2 (1.9%) | |
| Peripheral vascular diseases | 10 (11.1%) | 77 (9.6%) | 128 (6.5%) | 47 (4.9%) | 5 (4.9%) | 0.001 |
| Prior stroke | 15 (16.7%) | 112 (13.9%) | 266 (13.4%) | 123 (12.8%) | 8 (7.8%) | 0.39 |
| Smoker | 12 (13.3%) | 101 (12.6%) | 225 (11.3%) | 120 (12.5%) | 23 (22.3%) | 0.02 |
| Diabetes mellitus | 27 (30%) | 196 (24.4%) | 390 (19.6%) | 164 (17%) | 14 (13.6%) | <0.01 |
| Pacemaker | 11 (12.2%) | 55 (6.8%) | 121 (6.1%) | 53 (5.5%) | 5 (4.9%) | 0.12 |
| Aspirin | 31 (34.4%) | 228 (28.4%) | 524 (26.4%) | 235 (24.4%) | 19 (18.5%) | 0.04 |
| Warfarin | 75 (83.3%) | 675 (84%) | 1,680 (84.6%) | 836 (86.7%) | 87 (84.5%) | 0.5 |
| Lipid lowering drug | 21 (23.3%) | 214 (26.6%) | 443 (22.3%) | 202 (21%) | 14 (13.6%) | 0.008 |
| Beta blocker | 20 (26.7%) | 223 (34.4%) | 603 (38.6%) | 297 (40.5%) | 32 (41.6%) | 0.04 |
| ACE inhibitors | 42 (46.7%) | 308 (38.3%) | 723 (36.4%) | 408 (42.3%) | 61 (59.2%) | <0.01 |
| Diuretics | 64 (71.1%) | 373 (46.4%) | 804 (40.5%) | 378 (39.2%) | 56 (54.4%) | <0.01 |
| Other antihypertensive | 35 (38.9%) | 312 (38.8%) | 615 (31%) | 268 (27.8%) | 27 (26.2%) | <0.01 |
| CCBs | 21 (28%) | 204 (31.5%) | 508 (32.5%) | 240 (32.8%) | 24 (31.6%) | 0.91 |
| Digoxin | 69 (76.7%) | 432 (53.8%) | 1,059 (53.4%) | 476 (49.4%) | 56 (54.4%) | <0.01 |
| Amiodarone | 22 (24.4%) | 183 (22.8%) | 338 (17%) | 162 (16.8%) | 16 (15.5%) | 0.002 |
| Sotalol | 10 (11.1%) | 117 (14.6%) | 290 (14.6%) | 165 (17.1%) | 18 (17.5%) | 0.27 |
| Class 1 antiarrhythmic | 12 (13.3%) | 78 (9.7%) | 266 (13.4%) | 135 (14%) | 15 (14.6%) | 0.06 |
| Ejection fraction ∗ | <0.01 | |||||
| >50% | 31 (45.6%) | 414 (70.2%) | 1,122 (75.9%) | 567 (77.6%) | 60 (74.1%) | |
| 40%–49% | 13 (19.1%) | 77 (13.1%) | 187 (12.7%) | 89 (12.2%) | 13 (16.1%) | |
| 30%–39% | 10 (14.7%) | 49 (8.3%) | 114 (7.7%) | 52 (7.1%) | 5 (6.2%) | |
| <30% | 14 (20.6%) | 50 (8.5%) | 55 (3.7%) | 23 (3.2%) | 3 (3.7%) | |
| Ventricular septal thickness (mm) ∗ | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 1.2 ± 0.2 | 1.2 ± 0.2 | 0.0001 |
| LV posterior wall thickness (mm) ∗ | 1.0 ± 0.2 | 1.0 ± 0.2 | 1.0 ± 0.2 | 1.1 ± 0.2 | 1.1 ± 0.2 | 0.0005 |
| Left atrial diameters (mm) ∗ | 4.5 ± 0.6 | 4.3 ± 0.7 | 4.3 ± 0.6 | 4.3 ± 0.6 | 4.2 ± 0.7 | 0.08 |
∗ Echocardiographic data available for 2,948 patients (74.7% of entire study population).
C indexes of average and baseline SBP models were 0.73 (95% confidence interval [CI] 0.71 to 0.74) and 0.71 (95% CI 0.69 to 0.73), respectively. Similarly, C indexes of average and baseline DBP models were 0.72 (95% CI 0.70 to 0.74) and 0.71 (95% CI 0.69 to 0.73), respectively. The C indexes of the models with average on-treatment BP were significantly better than those of models with baseline BP (p values of 0.02 and 0.007 for SBP and DBP, respectively). The nadir BP (SBP and/or DBP) for the complete cohort resulting in the lowest ACM was calculated to be 140/78 mm Hg.
Of the 3,947 participants, ACM was observed in 614 participants (15.6%). The relation between average SBP and ACM followed a U-shaped relation with increased event rates at the low and high SBP ranges ( Figure 1 and Table 3 ). After adjusting for baseline covariates compared with the reference group (average SBP = 130 to 140 mm Hg), the risk of ACM (complete cohort) increased by 3.9 fold (p <0.001) in the group with SBP ≤110 mm Hg and by 1.9 fold (p <0.001) in the group with SBP >160 mm Hg. This U-shaped relation was not observed only when the baseline SBP was used ( Supplementary Figure 1 ). This relation remained unchanged when the analysis was repeated for patients after accounting for LVEF ( Supplementary Table 1 ). The group with average SBP <110 mm Hg actually had greater mortality than the group with average SBP >160 mm Hg (HR 3.9, p <0.001).
| Subgroups | All-Cause Mortality (Hazards Ratio, (CI) p Value) | Secondary Outcome (Hazards Ratio, (CI) p Value) | ||||||
|---|---|---|---|---|---|---|---|---|
| ∗ SBP < 110 | ∗ SBP > 160 | ∗ DBP < 60 | ∗ DBP > 90 | SBP < 110 | SBP > 160 | DBP < 60 | DBP > 90 | |
| Complete cohort (3,947) | 3.9 (2.9–5.4), p <0.001 | 1.9 (1.4–2.7), p <0.001 | 3.9 (2.8–5.3), p <0.001 | 1.8 (1.1–3.0), p = 0.02 | 2.5 (1.9–3.3), p <0.001 | 1.4 (1.0–1.9), p = 0.03 | 2.5 (1.9–3.3), p <0.001 | 1.5 (1.0–2.3), p = 0.04 |
| Rate control (1,974) | 4.2 (2.7–6.7), p <0.001 | 2.8 (1.7–4.6), p <0.001 | 4.2 (2.6–6.8), p <0.001 | 3.5 (1.8–6.7), p <0.001 | 2.5 (1.7–3.7), p <0.001 | 1.5 (1.0–2.3), p = 0.08 | 2.2 (1.4–3.5), p <0.001 | 2.1 (1.2–3.6), p = 0.01 |
| Rhythm control (1,973) | 3.6 (2.3–5.7), p <0.001 | 1.4 (0.9–2.3), p = 0.13 | 3.6 (2.4–5.6), p <0.001 | 0.9 (0.4–2.2), p = 0.8 | 2.4 (1.6–3.7), p <0.001 | 1.4 (0.9–2.1), p = 0.09 | 2.7 (1.8–4.0), p <0.001 | 1.2 (0.6–2.2), p = 0.61 |
| CAD absent (2,442) | 3.0 (1.7–5.3), p <0.001 | 1.6 (0.9–2.6), p = 0.09 | 4.8 (2.7–8.5), p <0.001 | 1.5 (0.7–3.3), p = 0.28 | 1.9 (1.1–3.1), p = 0.01 | 1.2 (0.8–1.9), p = 0.33 | 2.7 (1.6–4.6), p <0.001 | 1.2 (0.6–2.1), p = 0.62 |
| CAD present (1,505) | 4.7 (3.1–7.0), p <0.001 | 2.3 (1.4–3.7), p <0.001 | 3.9 (2.6–5.6), p <0.001 | 2.1 (1.0–4.3), p = 0.03 | 2.9 (2.0–4.2), p <0.001 | 1.6 (1.0–2.4), p = 0.02 | 2.6 (1.8–3.7), p <0.001 | 2.0 (1.1–3.6), p = 0.01 |
| Hypertension absent (1,139) | 3.3 (1.8–5.9), p <0.001 | 7.2 (3.2–16), p <0.001 | 2.9 (1.6–5.5), p <0.001 | 2.0 (0.3–15), p = 0.5 | 2.5 (1.5–3.9), p <0.001 | 4.6 (2.1–9.8), p <0.001 | 1.9 (1.1–3.4), p = 0.02 | 5.3 (1.6–17), p = 0.01 |
| Hypertension present (2,808) | 4.7 (3.2–6.9), p <0.001 | 1.6 (1.1–2.3), p = 0.02 | 4.6 (3.1–6.6), p <0.001 | 1.9 (1.1–3.2), p = 0.02 | 2.8 (1.9–4.0), p <0.001 | 1.2 (0.9–1.7), p = 0.2 | 3.0 (2.1–4.2), p <0.001 | 1.4 (0.9–2.2), p = 0.01 |
| CHF absent (3,046) | 3.3 (1.9–5.7), p <0.001 | 1.9 (1.2–2.9), p = 0.01 | 4.7 (2.9–7.9), p <0.001 | 1.8 (0.9–3.4), p = 0.07 | 2.1 (1.3–3.3), p <0.001 | 1.3 (0.9–1.9), p = 0.12 | 2.4 (1.5–3.9), p <0.001 | 1.3 (0.8–2.2), p = 0.28 |
| CHF present (901) | 4.2 (2.7–6.4), p <0.001 | 2.0 (1.1–3.4), p = 0.01 | 4.0 (2.6–6.1), p <0.001 | 1.8 (0.7–4.3), p = 0.16 | 2.8 (1.9–4.1), p <0.001 | 1.5 (0.9–2.5), p = 0.08 | 2.9 (2.0–4.2), p <0.001 | 2.1 (1.0–4.3), p = 0.03 |
| EF ≥ 30% (2,803) | 4.0 (2.6–6.4), p <0.001 | 1.9 (1.2–2.8), p <0.001 | 3.7 (2.4–5.6), p <0.001 | 1.8 (1.0–3.4), p = 0.07 | 2.9 (1.9–4.2), p <0.001 | 1.5 (1.0–2.1), p = 0.04 | 2.4 (1.6–3.6), p <0.001 | 1.7 (1.0–2.7), p = 0.04 |
| EF < 30% (145) | 1.8 (0.7–4.5), p = 0.22 | 0.7 (0.1–6.7), p = 0.75 | 5.0 (2.0–13), p <0.001 | 6.5 (1.1–36), p = 0.04 | 1.8 (0.8–4.2), p = 0.16 | 0.8 (0.1–6.9), p = 0.84 | 5.2 (2.2–12), p <0.001 | 3.6 (0.7–19), p = 0.12 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
