Obesity is a risk factor for atrial fibrillation (AF); however, obesity is associated with lower mortality in patients with established AF, a phenomenon known as the obesity paradox. Previous studies reported inconsistent results regarding effects of body weight on risk of cardiogenic embolism in patients with AF. To determine relation between body mass index (BMI) and prognosis among Japanese patients with nonvalvular AF (NVAF), a post hoc analysis was conducted using observational data in the J-RHYTHM Registry. Subjects were categorized as underweight (BMI <18.5), normal (18.5 to 24.9), overweight (25.0 to 29.9), or obese (≥30 kg/m 2 ). End points included thromboembolism, major hemorrhaging, all-cause mortality, and cardiovascular mortality. Of the 7,406 patients with NVAF, 6,379 patients (70 ± 10 years old; BMI, 23.6 ± 3.9 kg/m 2 ) having baseline BMI data constituted the study group. During the 2-year follow-up period, 111 patients had thromboembolism, 124 experienced major hemorrhage, and 159 died. Multivariate analysis with the Cox proportional hazards model showed that none of the BMI categories were independent predictors of thromboembolism. However, being underweight was an independent predictor of all-cause mortality (hazard ratio [HR] 2.45; 95% confidence interval [CI] 1.62 to 3.69; p <0.001) and cardiovascular mortality (HR 3.00, 95% CI 1.52 to 5.91, p = 0.001) when normal weight was used as the reference. Additionally, being overweight was a predictor of lower all-cause mortality (HR 0.60, 95% CI 0.37 to 0.95, p = 0.029). In conclusion, being underweight is associated with higher risks of all-cause and cardiovascular mortality compared with having a normal weight. Being overweight or obese is not associated with increased mortality among Japanese patients with NVAF.
Atrial fibrillation (AF) is one of the most common arrhythmias and is associated with mortality and morbidities including cardiogenic embolism. Among the general population, obesity is a risk factor for incidental AF and is associated with increased mortality and ischemic stroke. However, obesity is not associated with poor outcomes among subjects with established AF but rather is associated with lower mortality, a phenomenon that is known as the obesity paradox. This phenomenon was not supported in some studies. Recent studies from Asian countries have indicated that underweight subjects with AF suffer higher all-cause mortality, but studies from Western countries excluded underweight subjects from their analyses. Therefore, the present post hoc analysis was conducted to determine the relation between body mass index (BMI) and prognosis among Japanese patients with nonvalvular AF (NVAF) using the prospective data in the J-RHYTHM Registry.
Methods
The study design and baseline characteristics of the patients included in the J-RHYTHM Registry (UMIN Clinical Trials Registry, UMIN 000001569) have been reported elsewhere. Briefly, the study protocol conformed to the Declaration of Helsinki and was approved by each of the participating institutions. The subjects consisted of a consecutive series of outpatients with AF of any type. All the patients gave written informed consent. Patients were excluded from the present analysis if they had mitral stenosis, had undergone mechanical valve replacement, or were lost to follow-up during the 2-year follow-up period, as were patients for whom baseline BMI data were not available. Antithrombotic drugs and dosages were selected at the discretion of the treating physicians.
BMI was calculated as weight (kg)/[height (m)] 2 . The patients were divided into 4 BMI groups according to the criteria outlined in the World Health Organization guidelines, as was described in previous studies : the underweight group (BMI <18.5 kg/m 2 ), normal weight group (18.5 to 24.9 kg/m 2 ), overweight group (25.0 to 29.9 kg/m 2 ), and obesity group (≥30 kg/m 2 ). Because no direct oral anticoagulants (DOACs) were available when this registry was started in 2009, all anticoagulation therapy was performed with warfarin in the present study. Anticoagulation intensity was determined with the international normalized ratio (INR) of prothrombin time in patients who were receiving warfarin at the baseline, and the time in the therapeutic range (TTR) was determined using the method developed by Rosendaal et al. The target INR level was set at 1.6 to 2.6 for elderly patients aged ≥70 years and at 2.0 to 3.0 for patients aged <70 years according to the Japanese guidelines.
The end points were as follows: thromboembolic events, major hemorrhagic events, all-cause mortality, and cardiovascular mortality. The thromboembolic events included symptomatic ischemic stroke, transient ischemic attack, and systemic embolism. Major hemorrhagic events included intracranial hemorrhaging and other hemorrhagic events that required hospitalization. The diagnostic criteria for each event and methods of data collection have been described elsewhere.
Data are shown as percentage or mean ± SD values. Mean values were compared using the Student’s t test or ANOVA, and frequency values were compared using the chi-square test. The Cox proportional hazards model was used to calculate hazard ratios (HRs) and 95% confidence intervals for the risk factors for each end point. In the multivariate analysis, BMI category and variables that exhibited p values <0.10 in the univariate analysis were included as explanatory variables. A p value <0.05 was considered significant. All statistical analyses were performed with the software SPSS, version 15.0 (SPSS Inc., Chicago, Illinois).
Results
A total of 7,937 patients with AF were enrolled into the study from January 2009 to July 2009. Of these, 421 patients had mitral stenosis or had undergone mechanical valve replacement, and another 110 patients were lost to follow-up. Therefore, follow-up data were available for 7,406 patients with NVAF. Of these, baseline BMI data were missing for 1,027 patients (14%). The remaining 6,379 patients (mean age 70 ± 10 years; men 71%; mean BMI 23.6 ± 3.9 kg/m 2 ) constituted the study group.
The baseline characteristics of the 4 BMI groups are listed in Table 1 . Significant differences among the BMI groups were observed with respect to the baseline frequencies of several co-morbidities. The frequency of warfarin treatment did not differ but that of treatment with some drugs differed among the 4 groups ( Table 2 ). There were significant differences in some baseline characteristics between patients with NVAF with and without baseline BMI data ( Supplementary Table 1 , online only); however, frequency of warfarin use at baseline and baseline INR values did not differ between the 2 groups ( Supplementary Table 2 , online only).
| Variable | Body mass index (kg/m 2 ) | P value for trend | |||
|---|---|---|---|---|---|
| <18.5 (n=386) | 18.5–24.9 (n=3979) | 25.0–29.9 (n=1739) | ≥30.0 (n=275) | ||
| Body mass index (kg/m 2 ) | 17.0 ± 1.6 | 22.2 ± 1.7 | 26.9 ± 1.3 | 32.8 ± 8.0 | <0.001 |
| Age (years) | 75 ± 10 | 70 ± 10 | 68 ± 10 | 65 ± 11 | <0.001 |
| Female | 188 (49%) | 1126 (28%) | 419 (24%) | 93 (34%) | <0.001 |
| Type of atrial fibrillation | |||||
| Paroxysmal | 126 (33%) | 1584 (40%) | 643 (37%) | 81 (30%) | 0.001 |
| Persistent | 64 (16%) | 554 (14%) | 234 (13%) | 47 (17%) | |
| Permanent | 196 (51%) | 1841 (46%) | 862 (50%) | 147 (53%) | |
| Comorbidities | |||||
| Coronary artery disease | 47 (12%) | 397 (10%) | 202 (12%) | 43 (16%) | 0.010 |
| Cardiomyopathy | 37 (10%) | 328 (8%) | 158 (9%) | 30 (11%) | 0.328 |
| Hypertrophic | 17 (4%) | 136 (3%) | 71 (4%) | 11 (4%) | 0.529 |
| Dilated | 20 (5%) | 192 (5%) | 87 (5%) | 19 (7%) | 0.494 |
| Congenital heart disease | 6 (2%) | 65 (2%) | 18 (1%) | 4 (2%) | 0.385 |
| Chronic obstructive pulmonary disease | 23 (6%) | 66 (2%) | 25 (1%) | 2 (1%) | <0.001 |
| Hyperthyroidism | 10 (3%) | 64 (2%) | 29 (2%) | 6 (2%) | 0.492 |
| Risk factors for stroke | |||||
| Heart failure | 161 (42%) | 1069 (27%) | 451 (26%) | 97 (35%) | <0.001 |
| Hypertension | 183 (47%) | 2278 (57%) | 1199 (69%) | 207 (75%) | <0.001 |
| Age (≥75 years) | 227 (59%) | 1428 (36%) | 457 (26%) | 52 (19%) | <0.001 |
| Age (65–74 years) | 96 (25%) | 1541 (39%) | 692 (40%) | 102 (37%) | <0.001 |
| Diabetes mellitus | 45 (12%) | 667 (17%) | 383 (22%) | 115 (42%) | <0.021 |
| Stroke/Transient ischemic attack | 59 (15%) | 564 (14%) | 238 (14%) | 28 (10%) | 0.251 |
| CHADS 2 score | |||||
| 0 | 41 (11%) | 681 (17%) | 237 (13%) | 19 (7%) | |
| 1 | 118 (30%) | 1349 (34%) | 621 (36%) | 95 (35%) | |
| ≥2 | 227 (59%) | 1949 (49%) | 881 (51%) | 161 (58%) | |
| Mean | 1.9 ± 1.2 | 1.7 ± 1.2 | 1.7 ± 1.2 | 1.9 ± 1.2 | <0.001 |
| CHA 2 DS 2 -VASc score | |||||
| 0 | 8 (2%) | 271 (7%) | 126 (7%) | 11 (4%) | |
| 1 | 36 (9%) | 637 (16%) | 291 (17%) | 34 (12%) | |
| 2 | 75 (19%) | 894 (22%) | 395 (23%) | 69 (25%) | |
| ≥3 | 267 (69%) | 2177 (55%) | 927 (53%) | 161 (59%) | |
| Mean | 3.4 ± 1.6 | 2.8 ± 1.6 | 2.7 ± 1.6 | 3.0 ± 1.6 | <0.001 |
| Variable | Body mass index (kg/m 2 ) | P value for trend | |||
|---|---|---|---|---|---|
| <18.5 (n=386) | 18.5 – 24.9 (n=3979) | 25.0 – 29.9 (n=1739) | ≥30.0 (n=275) | ||
| Warfarin | 333 (86%) | 3444 (87%) | 1523 (88%) | 249 (91%) | 0.217 |
| Dosage (mg/day) | 2.4 ± 0.9 | 2.8 ± 1.1 | 3.1 ± 1.2 | 3.1 ± 1.2 | <0.001 |
| International normalized ratio | |||||
| <1.6 | 88 (26%) | 908 (26%) | 394 (26%) | 68 (27%) | <0.001 |
| 1.6–1.99 | 112 (34%) | 1288 (37%) | 546 (36%) | 97 (39%) | |
| 2.0–2.59 | 98 (29%) | 951 (28%) | 471 (31%) | 60 (24%) | |
| 2.6–2.99 | 23 (7%) | 198 (6%) | 81 (5%) | 16 (6%) | |
| ≥3.0 | 12 (4%) | 99 (3%) | 31 (2%) | 8 (3%) | |
| Mean | 1.94 ± 0.49 | 1.91 ± 0.50 | 1.90 ± 0.47 | 1.90 ± 0.48 | 0.585 |
| Time in therapeutic range (%) ∗ | 60 ± 27 (n=310) | 60 ± 29 (n=3269) | 57 ± 30 (n=1429) | 56 ± 28 (n=237) | <0.001 |
| Any antiplatelets | 102 (26%) | 1028 (26%) | 488 (28%) | 75 (27%) | 0.368 |
| Aspirin | 83 (22%) | 878 (22%) | 423 (24%) | 71 (26%) | 0.151 |
| Others | 26 (7%) | 241 (6%) | 118 (7%) | 8 (3%) | 0.090 |
| Warfarin + antiplatelets | 63 (16%) | 752 (19%) | 360 (21%) | 54 (20%) | 0.185 |
| ARB/ACE-I | 171 (44%) | 2019 (51%) | 1026 (59%) | 186 (68%) | <0.001 |
| Statins | 59 (15%) | 889 (22%) | 505 (29%) | 105 (38%) | <0.001 |
| Digitalis | 69 (18%) | 455 (11%) | 168 (10%) | 20 (7%) | <0.001 |
| Na channel blockers | 54 (14%) | 862 (22%) | 376 (22%) | 49 (18%) | 0.002 |
| β-blockers | 59 (15%) | 645 (16%) | 291 (17%) | 54 (20%) | 0.443 |
| K channel blockers † | 51 (13%) | 611 (15%) | 258 (15%) | 50 (18%) | 0.343 |
| Ca channel antagonists | 40 (10%) | 285 (7%) | 108 (6%) | 20 (7%) | 0.039 |
∗ Target international normalized ratio was 2.0 to 3.0 (<70 years old) or 1.6 to 2.6 (≥70 years old).
The event rates and age-adjusted HRs are summarized in Table 3 . Underweight patients were at higher risk of all-cause mortality and cardiovascular mortality compared with the normal weight group. Interestingly, the overweight patients were found to be at lower risk of all-cause mortality than the normal weight patients. The 2-year incidence rates of events did not differ between patients who were excluded from the present study because of a lack of BMI data and those included in the present study ( Supplementary Table 3 , online only). In Tables 4 and 5 , unadjusted and adjusted HRs for each event are shown, respectively. None of the BMI categories emerged as independent predictors of thromboembolic events ( Table 5 ). However, being underweight was found to be an independent predictor of all-cause mortality and cardiovascular mortality, and being overweight was demonstrated as a predictor of lower all-cause mortality. These findings were also true when the analyses were performed in the propensity score–matched groups ( Supplementary Tables 4 and 5 , online only). Female gender emerged as a predictor of lower all-cause mortality ( Table 5 ). Underweight male patients were at higher risk of all-cause and cardiovascular mortality, but female patients did not show any association between BMI categories and the end points ( Supplementary Table 6 , online only).
| Variable | Body mass index (kg/m 2 ) | P value for trend | |||
|---|---|---|---|---|---|
| <18.5 (n=386) | 18.5 – 24.9 (n=3979) | 25.0 – 29.9 (n=1739) | ≥30.0 (n=275) | ||
| Event rates (per 2 years) | |||||
| Thromboembolism | 10 (2.6%) | 71 (1.8%) | 27 (1.6%) | 3 (1.1%) | 0.133 |
| Major hemorrhage | 14 (3.6%) | 79 (2.0%) | 26 (1.5%) | 5 (1.8%) | 0.035 |
| All-cause death | 32 (8.3%) | 96 (2.4%) | 22 (1.3%) | 9 (3.3%) | <0.001 |
| Cardiovascular death | 13 (3.4%) | 27 (0.7%) | 12 (0.7%) | 3 (1.1%) | 0.017 |
| Hazard ratio (95% CI) ∗ | |||||
| Thromboembolism | 1.18 (0.61 – 2.30) | 1.00 (reference) | 0.96 (0.62 – 1.50) | 0.78 (0.24 – 2.47) | |
| Major hemorrhage | 1.62 (0.91 – 2.87) | 1.00 (reference) | 0.80 (0.52 – 1.25) | 1.08 (0.43 – 2.69) | |
| All-cause death | 2.56 (1.70 – 3.85) | 1.00 (reference) | 0.60 (0.38 – 0.95) | 1.88 (0.95 – 3.74) | |
| Cardiovascular death | 3.81 (1.84 – 7.48) | 1.00 (reference) | 1.15 (0.58 – 2.27) | 2.17 (0.65 – 7.18) | |
| Explanatory variables | Thromboembolism | Major hemorrhage | All-cause death | Cardiovascular death | ||||
|---|---|---|---|---|---|---|---|---|
| HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | |
| Heart failure | 1.17 (0.78–1.76) | 0.441 | 1.80 (1.25–1.57) | 0.001 | 4.41 (3.20–6.08) | <0.001 | 9.60 (5.06–18.29) | <0.001 |
| Hypertension | 1.15 (0.78–1.70) | 0.471 | 1.65 (1.12–2.44) | 0.012 | 0.77 (0.56–1.05) | 0.094 | 0.58 (0.34–0.99) | 0.045 |
| Age (≥75 years) ∗ | 2.67 (1.59–4.48) | <0.001 | 3.19 (1.87–5.45) | <0.001 | 6.63 (3.80–11.58) | <0.001 | 5.08 (2.14–12.07) | <0.001 |
| Age (65–74 years) ∗ | 1.29 (0.74–2.27) | 0.373 | 1.93 (1.10–3.37) | 0.022 | 2.02 (1.10–3.74) | 0.024 | 1.74 (0.67–4.53) | 0.256 |
| Diabetes mellitus | 1.31 (0.85–2.04) | 0.223 | 1.43 (0.95–2.15) | 0.083 | 1.49 (1.05–2.13) | 0.027 | 1.76 (0.98–3.15) | 0.057 |
| Stroke/Transient ischemic attack | 1.75 (1.12–2.75) | 0.015 | 2.02 (1.34–3.05) | <0.001 | 1.87 (1.29–2.71) | 0.001 | 1.42 (0.72–2.84) | 0.313 |
| Coronary artery disease | 1.13 (0.63–2.02) | 0.679 | 1.94 (1.23–3.05) | 0.004 | 3.28 (2.32–4.65) | <0.001 | 3.22 (1.78–5.83) | <0.001 |
| Female | 0.96 (0.64–1.46) | 0.860 | 0.73 (0.48–1.11) | 0.135 | 0.73 (0.50–1.06) | 0.094 | 1.11 (0.63–1.97) | 0.714 |
| Warfarin | 0.46 (0.30–0.70) | <0.001 | 2.46 (1.45–5.27) | 0.021 | 0.75 (0.49–1.14) | 0.179 | 0.86 (0.41–1.82) | 0.336 |
| Antiplatelets | 1.52 (1.03–2.24) | 0.035 | 1.54 (1.07–2.23) | 0.021 | 1.99 (1.46–2.73) | <0.001 | 2.34 (1.38–3.98) | 0.002 |
| Cardiomyopathy | 1.04 (0.54–1.99) | 0.913 | 0.92 (0.48–1.75) | 0.797 | 1.50 (0.94–2.40) | 0.088 | 3.23 (1.74–6.03) | <0.001 |
| Chronic obstructive pulmonary disease | 1.52 (0.48–4.78) | 0.475 | 1.82 (0.67–4.94) | 0.237 | 1.78 (0.73–4.34) | 0.203 | 1.02 (0.14–7.39) | 0.983 |
| Persistent atrial fibrillation † | 1.05 (0.52–2.11) | 0.885 | 1.06 (0.58–1.92) | 0.853 | 1.51 (0.90–2.54) | 0.123 | 1.44 (0.58–3.61) | 0.436 |
| Permanent atrial fibrillation † | 2.06 (1.33–3.19) | 0.001 | 1.50 (1.01–2.22) | 0.044 | 2.02 (1.39–2.92) | <0.001 | 2.16 (1.14–4.08) | 0.018 |
| ARB/ACE-I | 1.20 (0.12–1.75) | 0.343 | 1.60 (1.11–2.32) | 0.012 | 1.18 (0.86–1.62) | 0.297 | 1.55 (0.89–2.68) | 0.121 |
| Statins | 0.94 (0.61–1.47) | 0.798 | 1.08 (0.72–1.61) | 0.724 | 0.60 (0.40–0.92) | 0.019 | 0.45 (0.20–0.996) | 0.049 |
| Digitalis | 1.66 (1.01–2.71) | 0.045 | 1.28 (0.76–2.13) | 0.352 | 1.87 (1.26–2.78) | 0.002 | 1.17 (0.53–2.59) | 0.694 |
| Na channel blockers | 0.72 (0.44–1.19) | 0.203 | 0.89 (0.57–1.39) | 0.618 | 0.53 (0.33–0.85) | 0.009 | 0.29 (0.11–0.81) | 0.017 |
| β-blockers | 1.33 (0.84–2.10) | 0.230 | 1.34 (0.87–2.07) | 0.179 | 0.77 (0.49–1.22) | 0.270 | 0.62 (0.27–1.45) | 0.272 |
| K channel blockers | 0.61 (0.33–1.14) | 0.122 | 0.77 (0.45–1.31) | 0.331 | 0.94 (0.60–1.46) | 0.779 | 1.23 (0.62–2.44) | 0.552 |
| Ca channel antagonists | 1.59 (0.88–2.90) | 0.128 | 1.41 (0.78–2.56) | 0.258 | 1.27 (0.73–2.20) | 0.392 | 0.76 (0.24–2.44) | 0.646 |
| Explanatory variables | Thromboembolism | Major hemorrhage | All-cause death | Cardiovascular death | ||||
|---|---|---|---|---|---|---|---|---|
| HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | |
| Underweight (<18.5 kg/m 2 ) ∗ | 1.22 (0.63–2.38) | 0.561 | 1.71 (0.96–3.05) | 0.069 | 2.40 (1.59–3.63) | <0.001 | 2.91 (1.47–5.75) | 0.002 |
| Overweight (25.0–29.9 kg/m 2 ) ∗ | 0.94 (0.60–1.46) | 0.770 | 0.74 (0.47–1.16) | 0.192 | 0.60 (0.37–0.95) | 0.031 | 1.16 (0.58–2.30) | 0.680 |
| Obese (≥30.0 kg/m 2 ) ∗ | 0.71 (0.22–2.27) | 0.563 | 0.87 (0.35–2.19) | 0.771 | 1.70 (0.84–3.44) | 0.137 | 1.83 (0.54–6.20) | 0.332 |
| Heart failure | – | – | 1.42 (0.97–2.09) | 0.072 | 3.13 (2.21–4.43) | <0.001 | 5.99 (3.02–11.88) | <0.001 |
| Hypertension | – | – | 1.39 (0.89–2.16) | 0.148 | 0.77 (0.56–1.07) | 0.115 | 0.60 (0.35–1.05) | 0.072 |
| Age (≥75 years) † | 2.35 (1.38–4.01) | 0.002 | 2.27 (1.30–3.95) | 0.004 | 4.86 (2.73–8.65) | <0.001 | 3.59 (1.46–8.82) | 0.005 |
| Age (65–74 years) † | 1.26 (0.71–2.22) | 0.434 | 1.60 (0.91–2.82) | 0.106 | 1.90 (1.02–3.53) | 0.043 | 1.75 (0.66–4.61) | 0.259 |
| Diabetes mellitus | – | – | 1.21 (0.80–1.85) | 0.368 | 1.23 (0.87–1.83) | 0.227 | 1.53 (0.83–2.82) | 0.173 |
| Stroke/Transient ischemic attack | 1.65 (1.04–2.62) | 0.033 | 1.69 (1.11–2.56) | 0.015 | 1.56 (1.07–2.27) | 0.021 | – | – |
| Coronary artery disease | – | – | 1.30 (0.79–2.16) | 0.305 | 2.16 (1.44–3.23) | <0.001 | 2.06 (1.03–4.10) | 0.041 |
| Female | – | – | – | – | 0.57 (0.39–0.84) | 0.004 | – | – |
| Warfarin | 0.35 (0.22–0.57) | <0.001 | 2.17 (0.99–4.78) | 0.055 | – | – | – | – |
| Antiplatelet | 1.00 (0.65–1.54) | 0.993 | 1.32 (0.87–2.01) | 0.197 | 1.41 (0.99–2.00) | 0.057 | 1.98 (1.10–3.56) | 0.023 |
| Cardiomyopathy | – | – | – | – | 1.08 (0.66–1.77) | 0.771 | 1.79 (0.92–3.51) | 0.088 |
| Persistent atrial fibrillation ‡ | 1.16 (0.57–2.34) | 0.682 | 0.95 (0.52–1.73) | 0.865 | 1.15 (0.66–1.98) | 0.623 | 0.97 (0.38–2.50) | 0.951 |
| Permanent atrial fibrillation ‡ | 2.13 (1.35–3.36) | 0.001 | 1.14 (0.75–1.71) | 0.545 | 1.14 (0.75–1.74) | 0.378 | 0.95 (0.48–1.89) | 0.879 |
| ARB/ACE-I | – | – | 1.25 (0.82–1.90) | 0.294 | – | – | – | |
| Statins | – | – | – | – | 0.52 (0.33–0.81) | 0.004 | 0.34 (0.15–0.79) | 0.012 |
| Digitalis | 1.28 (0.77–2.12) | 0.339 | – | – | 1.24 (0.82–1.88) | 0.303 | – | – |
| Na channel blockers | – | – | – | – | 0.97 (0.57–1.64) | 0.907 | 0.56 (0.19–1.67) | 0.302 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
