Effect of Obesity on the Prognostic Impact of Atrial Fibrillation in Heart Failure With Preserved Ejection Fraction

Abstract

Background: Although obesity is associated with left ventricular hypertrophy, diastolic dysfunction, and occurrence of atrial fibrillation (AF), obese heart failure (HF) patients have a more favorable clinical outcome (obesity paradox). The clinical impact of AF on obese or lean HF patients has not been fully elucidated.

Methods and Results: We analyzed 1,681 patients who were enrolled in the West Tokyo Heart Failure Registry (WET-HF Registry), a multicenter, prospective cohort registry from 2005 through 2014. We assigned them to 3 categories based on body mass index (BMI): low, BMI <18.5; medium, BMI ≥18.5 and <25; and high, BMI ≥25 (n=182/915/400). The clinical endpoint was all-cause death or readmission for acute decompensated HF. During 406 days of follow-up (IQR, 116–739 days), AF was associated with a higher risk of the endpoint in the HF with preserved ejection fraction (HFpEF) group (P<0.001, log-rank test), but not in the HF with reduced EF (HFrEF) group. AF was associated with a higher risk of the endpoint in low and medium BMI patients with HFpEF (P=0.016 and 0.009, respectively). On Multivariate Cox proportional hazards analysis, AF was an independent predictor of the endpoint in patients with BMI <25 from the HFpEF group (hazard ratio, 1.74; 95% CI: 1.21–2.54, P=0.003), but not in the other subgroups.

Conclusions: AF had a negative impact on clinical outcome in non-obese patients with HFpEF.

Heart failure (HF) is a worldwide health priority that is placing an increasing socioeconomic burden on developed countries with aging populations. Atrial fibrillation (AF) plays a pivotal role in the pathophysiology of HF due to uncontrolled tachycardia, irregular heart rate, shortening of the filling period, and loss of atrial contraction.^1–6 Although many previous studies have addressed the prognostic impact of AF, the results have been inconsistent.^7–9

In contrast, the prognosis of obese HF patients is better than that of lean HF patients, even though obesity itself is a risk factor for the onset of HF,^10–13 which is often referred to as the obesity paradox. Given that obesity is associated with left ventricular hypertrophy (LVH), diastolic dysfunction,¹⁴ and occurrence of AF,¹⁵ it is possible that the impact of AF on hemodynamics and even on clinical outcome might vary in obese and lean subjects.

We hypothesized that AF might have a stronger influence on clinical outcome in patients with HF with preserved ejection fraction (HFpEF) than in patients with HFrEF, and that the effect of AF on prognosis might vary in obese and non-obese HF patients. Therefore, we investigated the potential differences in the prognostic impact of AF between HFpEF and HFrEF patients, as well as between obese and non-obese HF patients, in a multicenter observational study. The findings in the present study will facilitate better understanding of the involvement of AF in HF pathophysiology by clarifying the at-risk population.

Methods

Study Design

The West Tokyo Heart Failure Registry (WET-HF Registry) is a multicenter, prospective cohort registry study that enrolled all patients who were hospitalized with a diagnosis of acute decompensated HF (ADHF) according to Framingham acute HF criteria. Patients presenting with acute coronary syndrome were not included. The 4 study centers are located in Tokyo, Japan, and include 2 university hospitals (Keio University and Kyorin University) and 2 tertiary referral hospitals (Sakakibara Heart Institute and St. Luke’s International Hospital). Informed consent was obtained from each subject before the study. Nearly all of the patients enrolled were Japanese. This study was approved by the ethics review committee of each center.

Subjects

The data for patients enrolled from October 2005 through April 2014 were analyzed. AF was defined as atrial tachyarrhythmia, including AF, flutter, or atrial tachycardia, detected on electrocardiogram (ECG) recorded at the time of hospital admission for ADHF. We divided the subjects into 2 groups according to the presence or absence of AF (AF group and non-AF group, respectively). LV ejection fraction (LVEF) was measured on echocardiography in all subjects during hospitalization. HFpEF was defined as LVEF ≥40% and HFrEF was defined as LVEF <40% based on the MAGGIC study, which showed that mortality increased when LVEF was <40%.¹⁶ The body mass index (BMI) of each subject was calculated using the following formula: body weight (kg)/height (m)². The subjects were also stratified into 3 categories according to BMI at hospital admission based on National Heart, Lung, and Blood Institute and World Health Organization recommendations.^17,18

Clinical Parameters

We collected conventional clinical variables including age, gender, prior admissions for HF, etiology of HF, risk factors (such as hypertension, dyslipidemia, diabetes mellitus, and smoking), heart rate and blood pressure on admission. We also collected laboratory data, echocardiographic findings, and concomitant medications. The primary endpoint was defined as all-cause death or re-hospitalization for ADHF.

Statistical Analysis

Data with a normal distribution are expressed as mean±SD. Differences between groups were assessed using unpaired t-test or Mann-Whitney U-test for unpaired data, and with chi-squared test for discrete variables. Differences between the 3 groups were compared using analysis of variance (ANOVA) or Kruskal-Wallis test. The subjects were stratified by presence/absence of AF, LVEF (≥40% or <40%), and BMI (<18.5, ≥18.5–<25, and ≥25). Kaplan-Meier survival curves were constructed for each group and differences between groups were analyzed on log-rank test. Multivariate Cox regression analysis was also carried out using the stratifications LVEF (≥40% or <40%) and BMI (<25 or ≥25), to identify variables including AF that predicted clinical outcome during follow-up. P<0.05 was considered to indicate statistical significance. All statistical analysis was performed using JMP 12.0.1 (SAS, Cary, NC, USA).

Results

A total of 1,681 subjects were included in the present study, of whom, 184 were excluded because of missing ECG or BMI data at the time of admission (Figure 1), leaving 1,497 patients for investigation (AF group, n=631; non-AF group, n=866). When the subjects were divided into 2 categories according to baseline LVEF, 821 patients had HFpEF and 676 patients had HFrEF; and the 3 BMI groups were as follows: BMI <18.5, n=182; BMI≥18.5–<25, n=915; and BMI ≥25, n=400.

Figure 1.

Flowchart of subject selection. AF, atrial fibrillation; BMI, body mass index; ECG, electrocardiogram; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction.

With regard to the AF groups, mean age was higher, ischemic etiology was less common and valvular etiology was more common in the AF group than the non-AF group (Table 1). Prior admission for ADHF was more frequent and the number of prior admissions was greater in the AF group. On echocardiography, LVEF was higher and LV dimensions were smaller in the AF group than the non-AF group (Table 1).

Table 1. Baseline Characteristics vs. Presence of AF

	Non-AF group (n=866)	AF group (n=631)	P value
Age (years)	71±15	75±11	<0.0001
Female	331 (38)	241 (38)	0.99
Etiology
Ischemic	248/795 (31)	115/598 (19)	<0.0001
Valvular	170/761 (22)	206/602 (34)	<0.0001
History of prior admission	278/853 (33)	272/616 (44)	<0.0001
No. prior HF admissions	0.9±1.3	1.2±1.5	0.005
Hypertension	604/859 (70)	417/622 (67)	0.18
Dyslipidemia	376/854 (44)	231/619 (37)	0.01
Diabetes mellitus	345/865 (40)	206/628 (33)	0.005
Smoking	326/821 (40)	269/605 (44)	0.07
Heart rate (beats/min)	91±25	97±30	0.007
SBP (mmHg)	141±35	134±30	0.0005
LVEF (%)	41.6±15.8	44.9±15.6	<0.0001
LVEDD (mm)	54±10	52±10	0.002
LVESD (mm)	42±13	40±12	0.001
Hemoglobin (g/dL)	12.0±2.5	12.1±2.3	0.46
Serum Na (mEq/L)	139.0±4.3	138.9±6.1	0.63
eGFR (mL/min/1.73 m2)	49.9±27.4	51.6±23.8	0.11
Pre-discharge medications
Loop diuretics	507/779 (65)	437/588 (74)	0.0002
ACEI/ARB	507/786 (65)	371/607 (61)	0.19
β-blockers	565/795 (71)	445/599 (74)	0.18
Aldosterone antagonists	234/815 (29)	200/603 (33)	0.07
LOH (days)	20±30	18±23	0.09
In-hospital mortality	18/866 (2)	15/631 (2)	0.72

Data given as mean±SD or n (%). ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin receptor blocker; eGFR, estimated glomerular filtration rate; HF, heart failure; LOH, length of hospitalization; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; SBP, systolic blood pressure.

With regard to EF, AF was more frequent in the HFpEF group (377/821, 46%) than in the HFrEF group (254/676, 38%; P=0.001; Table 2). In both groups, prior admission for ADHF was more common among patients with AF than in those without AF, and systolic blood pressure (SBP) was lower in patients with AF. These differences between AF and non-AF patients were more evident in the HFpEF group than in the HFrEF group. In the HFpEF group, ischemic etiology was less common and valvular etiology was more common in the AF group than non-AF group, while loop diuretics, β-blockers, and aldosterone antagonists were more commonly prescribed at discharge in the AF group.

Table 2. Baseline Characteristics vs. EF Type and Presence of AF

	HFpEF			HFrEF
	Non-AF (n=444)	AF (n=377)	P value	Non-AF (n=422)	AF (n=254)	P value
Age (years)	68±16	72±13	0.002	74±13	77±10	0.0015
Female	197/444 (44)	153/377 (41)	0.27	134/422 (32)	88/254 (35)	0.44
Etiology
Ischemic	120/397 (30)	46/360 (13)	<0.0001	128/398 (32)	69/238 (29)	0.40
Valvular	118/367 (32)	163/366 (45)	0.0006	52/394 (13)	43/236 (18)	0.09
History of prior admission	113/438 (26)	152/366 (42)	<0.0001	165/415 (40)	120/250 (48)	0.04
No. prior HF admissions	0.7±1.2	1.1±1.4	<0.0001	1.2±1.4	1.3±1.7	0.88
Hypertension	342/442 (77)	259/374 (69)	0.009	262/417 (63)	158/248 (64)	0.82
Dyslipidemia	202/442 (46)	125/373 (34)	0.0004	174/412 (42)	106/246 (43)	0.83
Diabetes mellitus	191/444 (43)	118/375 (31)	0.0007	154/421 (37)	88/253 (35)	0.64
Smoking	139/426 (33)	153/366 (42)	0.008	187/395 (47)	116/239 (49)	0.77
Heart rate (beats/min)	89±27	95±31	0.04	93±24	99±30	0.04
SBP (mmHg)	147±37	137±29	0.0001	135±33	130±30	0.07
LVEF (%)	54.5±9.4	55.9±8.6	0.02	28.0±7.4	28.5±6.9	0.46
LVEDD (mm)	48±8	48±8	0.67	60±9	58±9	0.02
LVESD (mm)	34±8	33±8	0.61	51±10	50±10	0.04
Hemoglobin (g/dL)	11.3±2.4	11.8±2.2	0.006	12.8±2.4	12.6±2.4	0.64
Serum Na (mEq/L)	138.8±4.5	139.2±4.1	0.14	139.2±4.1	138.4±8.3	0.37
eGFR (mL/min/1.73 m2)	49.2±28.8	54.3±24.5	0.002	50.7±25.8	47.7±22.1	0.17
Pre-discharge medications
Loop diuretics	230/386 (60)	266/355 (75)	<0.0001	277/393 (70)	171/233 (73)	0.43
ACEI/ARB	241/391 (62)	213/363 (59)	0.41	266/395 (67)	158/244 (65)	0.50
β-blockers	249/396 (63)	257/358 (72)	0.009	316/399 (79)	188/241 (78)	0.72
Aldosterone antagonists	81/412 (20)	105/363 (29)	0.003	153/403 (38)	95/240 (40)	0.68
LOH (days)	20±34	18±26	0.55	21±26	18±16	0.10
In-hospital mortality	7/444 (1.6)	7/377 (1.9)	0.76	11/422 (2.6)	8/254 (3.2)	0.68

Data given as mean±SD or n (%). HFpEF, HF with preserved ejection fraction; HFrEF, HF with reduced ejection fraction. Other abbreviations as in Table 1.

For the 3 BMI categories, patients with higher BMI were younger and less likely to be female, have valvular etiology, or prior admission. And patients with higher BMI had higher SBP, hemoglobin, estimated glomerular filtration rate (eGFR), and prescription rate of angiotensin-converting enzyme inhibitors/angiotensin receptor antagonists, β-blockers, and aldosterone antagonists. Higher BMI was associated with lower in-hospital mortality rate (Table S1).

Furthermore, each of the 3 BMI categories was stratified by the presence or absence of AF. In all 3 BMI categories, patients with AF were older than those without AF (Table 3). These subgroups were then further stratified into HFpEF (Table S2) and HFrEF (Table S3). In all subgroups, ischemic etiology was less common and valvular etiology was more common for patients with AF compared with the non-AF group, although the difference was not significant in the majority of subgroups.

Table 3. Baseline Characteristics vs. BMI and Presence of AF

	Low BMI (<18.5 kg/m2)			Medium BMI (18.5–24.9 kg/m2)			High BMI (≥25 kg/m2)
	Non-AF (n=106)	AF (n=76)	P value	Non-AF (n=527)	AF (n=388)	P value	Non-AF (n=233)	AF (n=167)	P value
Age (years)	75±15	81±9	0.002	73±14	76±11	0.006	65±16	71±12	0.0002
Female	55/106 (52)	31/76 (41)	0.14	203/527 (39)	152/388 (39)	0.84	73/233 (31)	58/167 (35)	0.48
Etiology
Ischemic	26/92 (28)	14/72 (19)	0.19	161/478 (34)	61/364 (17)	<0.0001	61/225 (27)	40/162 (25)	0.59
Valvular	29/90 (32)	30/69 (43)	0.15	101/455 (22)	134/371 (36)	<0.0001	40/216 (19)	42/162 (26)	0.09
History of admission	47/103 (46)	38/73 (52)	0.4	167/519 (32)	175/378 (46)	<0.0001	64/231 (28)	59/165 (36)	0.09
No. prior HF admissions	1.2±1.9	1.5±1.8	0.17	0.9±1.2	1.2±1.5	0.004	0.8±1.1	0.9±1.3	0.61
Hypertension	71/104 (68)	51/74 (69)	0.92	357/522 (68)	240/384 (63)	0.07	176/233 (76)	126/164 (77)	0.77
Dyslipidemia	28/104 (27)	23/74 (31)	0.55	225/518 (43)	130/380 (34)	0.005	123/232 (53)	78/165 (47)	0.26
Diabetes mellitus	29/106 (27)	22/76 (29)	0.81	209/526 (40)	109/387 (28)	0.0003	107/233 (46)	75/165 (45)	0.93
Smoking	28/100 (28)	22/73 (30)	0.76	189/501 (38)	159/369 (43)	0.11	109/220 (50)	88/163 (54)	0.39
Heart rate (beats/min)	92±27	98±32	0.36	90±25	95±29	0.09	91±26	100±33	0.04
SBP (mmHg)	133±32	131±27	0.92	140±35	132±29	0.0004	148±38	142±30	0.23
LVEF (%)	41.3±17.1	41.5±1.9	0.97	41.2±15.4	45.3±15.8	<0.0001	42.6±15.9	45.4±14.5	0.07
LVEDD (mm)	52±11	50±10	0.49	54±10	52±11	0.02	56±10	54±9	0.048
LVESD (mm)	41±13	40±11	0.56	42±13	40±12	0.007	43±13	41±11	0.096
Hemoglobin (g/dL)	10.7±2.3	11.5±1.9	0.01	11.9±2.3	12.0±2.4	0.51	13.0±2.6	12.7±2.3	0.26
Serum Na (mEq/L)	138.1±4.9	137.4±13.4	0.22	138.9±4.4	139.0±4.4	0.91	139.6±3.8	139.3±3.8	0.52
eGFR (mL/min/1.73 m2)	45.9±31.7	47.6±26.7	0.39	49.2±26.7	51.6±23.9	0.16	53.3±26.5	53.5±21.8	0.58
Pre-discharge medications
Loop diuretics	56/86 (65)	46/71 (65)	0.97	307/477 (64)	276/362 (76)	0.0002	144/216 (67)	115/155 (74)	0.12
ACEI/ARB	48/93 (52)	32/75 (43)	0.25	304/476 (64)	228/372 (61)	0.44	155/217 (71)	111/160 (69)	0.67
β-blockers	57/96 (59)	50/73 (68)	0.22	343/481 (71)	271/368 (74)	0.45	165/218 (76)	124/158 (78)	0.53
Aldosterone antagonists	13/94 (14)	22/72 (31)	0.009	141/501 (28)	131/370 (35)	0.02	80/220 (36)	47/161 (29)	0.14
LOH (days)	26±43	25±51	0.32	21±32	18±15	0.64	18±14	16±14	0.047
In-hospital mortality	5/106 (4.7)	3/76 (4.0)	0.80	12/527 (2.3)	11/388 (2.8)	0.60	1/233 (0.4)	1/167 (0.6)	0.81

Data given as mean±SD or n (%). BMI, body mass index. Other abbreviations as in Table 1.

The median follow-up period was 406 days (IAR, 116–739 days). During follow-up, 248 subjects died; the cause was cardiac in 118 (48%), non-cardiac in 82 (33%), and unknown in the remainder (48; 19%). On Kaplan-Meier analysis, patients with AF reached the clinical endpoint more frequently than those without AF (P=0.002, log-rank test, Figure 2A). Comparison between the HFpEF and HFrEF groups showed that event-free survival was slightly lower in the latter group with borderline significance (P=0.12, log-rank test; P=0.047, Wilcoxon test). AF on admission was associated with a higher risk of the endpoint in the HFpEF group (P=0.0009, log-rank test, Figure 2B), but not in the HFrEF group (P=0.16, log-rank test, Figure 2C).

Figure 2.

Kaplan-Meier curves for event-free survival (i.e. without all-cause death or re-hospitalization for HF) stratified by the presence of AF in (A) the overall subject group (AF/non-AF, n=487/644), (B) HFpEF (AF/non-AF, n=279/311); and (C) HFrEF (AF/non-AF, n=208/333). AF was associated with a higher incidence of the endpoint in (A) the overall group (P=0.002, log-rank test) and (B) the HFpEF group (P=0.0009, log-rank test), but not in (C) the HFrEF group (P=0.16, log-rank test). HF, heart failure. Other abbreviations as in Figure 1.

In the overall subject group, the clinical endpoint was reached more frequently by patients with lower BMI (P<0.0001, log-rank test, Figure S1). AF was associated with a higher risk of the endpoint in patients with medium BMI (P=0.046, log-rank test, Figure S2B), but this association did not reach significance in patients with low or high BMI (Figure S2A, 2B). When the subjects were further divided into HFpEF and HFrEF groups, AF was associated with a higher risk of the endpoint for patients with low or medium BMI from the HFpEF group (Figure 3A,B), but differences in clinical outcome between patients with or without AF were not statistically significant in the other subgroups (Figure 3C–F). On Cox proportional hazards analysis, AF was an independent predictor of the endpoint after adjusting for age, gender, valvular etiology, pre-discharge eGFR, hemoglobin, and serum sodium in patients from the HFpEF group with BMI <25 (hazard ratio, 1.74; 95% CI: 1.21–2.54, P=0.003), but not in the other subgroups (Figure 4).

Figure 3.

Kaplan-Meier curves for all-cause death or re-hospitalization due to HF stratified by AF in the HFpEF and HFrEF groups according to BMI. AF/non-AF: (A) n=25/28; (B) n=170/188; (C) n=84/95; (D) n=31/37; (E) n=126/210; (F) n=51/86. AF was associated with a higher incidence of the endpoint in (A,B) HFpEF patients with low or medium BMI (A, P=0.016, log-rank test; B, P=0.009, log-rank test), and in (F) HFrEF patients with high BMI at borderline significance (P=0.057, log-rank test), but not in (C–E) the other subgroups. Abbreviations as in Figures 1,2.

Figure 4.

Hazard ratios (HR) for all-cause death and re-hospitalization due to HF stratified by the presence of AF in the (A) HFpEF group and (B) HFrEF group categorized by BMI (<25 or ≥25 kg/m²). HR were calculated by the Cox proportional hazards model after adjustment for age, gender, valvular etiology, estimated glomerular filtration rate, hemoglobin, and serum sodium at discharge. Abbreviations as in Figures 1,2.

Each endpoint of death and readmission for ADHF was analyzed separately. No differences were seen in all-cause death in any subgroups, but the AF group had a higher rate of HF readmission in medium BMI with HFpEF (P=0.03) and in high BMI with HFrEF (P=0.01). On Kaplan–Meier analysis for cardiac/non-cardiac death stratifying by AF in the HFpEF and HFrEF groups divided into 3 BMI categories, there were no differences between AF and non-AF in any of those subgroups (data not shown).

Discussion

Although various studies have examined the significance of AF in HF, the results have been inconsistent, possibly because of differences in the severity of HF, the definition of HFpEF/HFrEF, and the observation period.^{1,2,7–9,19,20} In the present study, we analyzed the subset of HF patients most affected by AF, which included a relatively high percentage of patients with low BMI compared with Western studies.^21–23 The main findings were as follows: (1) the risk of the composite primary endpoint (all-cause death and re-hospitalization for ADHF) was higher in patients with AF than in those without AF; (2) the incidence rate of the primary endpoint was higher in patients with AF from the HFpEF group, but not the HFrEF group; (3) the incidence rate of the primary endpoint was higher in patients with AF from the HFpEF group who had low-medium BMI; and (4) AF was an independent predictor of the endpoint only in the HFpEF BMI <25 group. Taken together, the prognostic impact of AF might differ in non-obese and obese HFpEF patients.

Prognostic Impact of AF in HF

AF may adversely affect hemodynamics in patients with HF because increased heart rate results in inadequate diastolic filling, as well as the impact of loss of atrial contraction and onset of atrioventricular valvular regurgitation.^1–6 Although many studies have examined the prognostic significance of AF in HF, the results have been inconsistent.^2,7–9 Previous studies have shown that AF is independently associated with mortality and morbidity in patients with a slight decrease of LVEF, but not in patients with severe reduction of LVEF.^1,19,20

Significance of AF in HFpEF

AF was more frequent in the HFpEF than the HFrEF group (Table 2). In addition, AF was associated with worse mortality and morbidity in the HFpEF group, but not in the HFrEF group (Figure 2B,C). Theoretically, the presence of diastolic dysfunction, indicated by transmitral E/A ratio exceeding the median and increase in the left atrial dimension, is strongly associated with the development of AF.²⁴ Loss of atrial contraction may adversely affect hemodynamics, especially in patients with HFpEF,²⁵ although the prognostic impact of AF on the 2 types of HF is also controversial,^26–30 possibly because of inconsistency in the definition of HFpEF/HFrEF as well as variation of the observation period.

Obesity Paradox and Cardiac Cachexia in HF

Obesity is a well-recognized risk factor for cardiovascular disease. In the general population, higher BMI is associated with an increased risk of cardiovascular events and new onset HF.^10–13 Nonetheless, higher BMI is paradoxically associated with better prognosis in chronic HF³¹ or ADHF.^21,22,32 These findings were supported by a recent meta-analysis.³³ Obese patients have lower plasma B-type natriuretic peptide³⁴ and their prognosis is more favorable, despite a higher prevalence of LVH.¹⁴ This suggests that unknown protective factors might exist in obese patients. For example, lower systemic vascular resistance and lower plasma renin activity have been postulated as possible factors contributing to the better prognosis in obese HF patients.³⁵

Cachexia is frequently observed in HF,³⁶ and this is known as cardiac cachexia. This condition is considered to be attributable to nutritional intake being inadequate to meet the increased energy requirements associated with HF,^37,38 along with alterations of the heart (including a decrease of cardiac mass), a catabolic state, neurohormone imbalance, and inflammation.³⁹ Cachexia is a strong predictor of poor clinical outcome.³⁸ In agreement with these findings, lower BMI was associated with worse outcome in the present study (Figure S1).

Impact of Overweight/Underweight on AF

In the HFpEF group, patients with AF had a higher event rate than patients without AF from the low and medium BMI subgroups, whereas event rate did not differ according to presence of AF in the high BMI subgroup (Figure 3A–C). This suggests that AF might have a greater negative impact on clinical outcome in patients with lower BMI and HFpEF, but not HFrEF. As noted here, AF has an adverse influence on diastolic function due to loss of the atrial kick and the high heart rate. Furthermore, obesity is associated with LVH and diastolic dysfunction,¹⁴ and is a predictor of new onset AF,¹⁵ although a recent paper showed no association of BMI with new onset AF in HF patients.⁴⁰ Accordingly, AF might have a stronger negative impact on hemodynamics in patients with diastolic dysfunction because of impaired blood filling of the LV chamber. Therefore, AF might have had a greater adverse clinical impact on obese patients, but, as shown in the present study, obese patients were more resistant to the unfavorable effects of AF. Although loss of LV mass has been confirmed on echocardiography⁴¹ or cardiac magnetic resonance⁴² in cachectic patients, no other cardiac abnormalities have been reported so far, and the significance of this morphological change in relation to the pathology of cardiac cachexia remains unknown. So far, we cannot provide any clues on the potential mechanism by which AF has a stronger impact on the clinical outcome in low BMI patients. We cannot exclude the possibility that AF might be a marker of comorbidity (including older age), which negatively affects the clinical outcome of HF, rather than a negative driver for it. AF is associated with frailty in elderly subjects.^43,44 Therefore, in the present study, frailty might be more frequent in patients with AF and low BMI, and it might be associated with poor outcome in this subgroup, although we did not collect data on frailty, physical disability, or cognitive impairment in the present study.

Taken together, AF was associated with worse outcome in HFpEF patients with lower BMI via unknown mechanisms. The nature of the present observational study did not allow us to explore their cause-effect relationship. These findings may, however, provide a clue leading to the development of novel strategies targeting the population more susceptible to negative hemodynamic effect by AF. Further investigation is needed to elucidate the underlying mechanisms.

Study Limitations

BMI is not considered to be an accurate measure of body fat. BMI is known to misclassify patients, especially elderly patients with low-normal BMI but very high body fat percentage, and vice versa.⁴⁵ In the present study, the data on body weight were obtained within 24 h from admission. High body weight might reflect, at least partly, the presence of fluid retention in addition to high body fat. Another limitation is the fairly marked body composition differences between Asia and the US and Europe. Therefore, these data should be interpreted and applied, with caution, to other regions. There were some differences of HF etiology between AF and non-AF patients in the LVEF and BMI subgroups. These differences in etiology might have affected the prognostic impact of AF in this study, although we conducted multivariate analysis with valvular etiology as an independent variable (Figure 4). In addition, newly diagnosed AF is associated with worse in-hospital clinical outcome in ADHF compared with pre-existing AF,⁴⁶ but we were not able to obtain information on duration of AF in this study. Finally, the data on vital signs at discharge were not collected, given that the present data were very limited. This is also a limitation in the present study given that SBP and heart rate at discharge have also been shown to be associated with long-term prognosis.^47,48

Conclusions

In patients with ADHF, AF had a stronger influence on mortality and morbidity in HFpEF than in HFrEF. Moreover, this impact was even greater in patients with low BMI. This suggests that AF might be a key to understanding the pathophysiological mechanisms that lead to poorer prognosis in patients with lower BMI: namely, the obesity paradox.

Acknowledgments

This study was supported by Grant-in-Aid for Scientific Research (C) (23591062, 26461088), Health Labour Sciences Research Grant (14528506) and the Sakakibara Clinical Research Grant for Promotion of Sciences, 2012, 2013, 2014.

Disclosures

The authors declare no conflict of interest.

Supplementary Files

Supplementary File 1

Supplementary Methods

Figure S1. Kaplan-Meier curves for event-free survival (i.e., without all-cause death or re-hospitalization for heart failure) according to body mass index (BMI: <18.5, 18.5–24.9, ≥25 kg/m²).

Figure S2. Kaplan-Meier survival curves for event-free survival (i.e., without all-cause death or re-hospitalization for heart failure) according to body mass index (BMI) and presence of atrial fibrillation (AF).

Table S1. Baseline characteristics vs. BMI

Table S2. Baseline HFpEF patient characteristics vs. BMI and presence of AF

Table S3. Baseline HFrEF patient characteristics vs. BMI and presence of AF

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-16-1130

References

• 1.   Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991; 84: 40–48.
• 2.   Carson PE, Johnson GR, Dunkman WB, Fletcher RD, Farrell L, Cohn JN. The influence of atrial fibrillation on prognosis in mild to moderate heart failure: The V-HeFT Studies: The V-HeFT VA Cooperative Studies Group. Circulation 1993; 87: vi102–vi110.
• 3.   Senni M, Tribouilloy CM, Rodeheffer RJ, Jacobsen SJ, Evans JM, Bailey KR, et al. Congestive heart failure in the community: A study of all incident cases in Olmsted County, Minnesota, in 1991. Circulation 1998; 98: 2282–2289.
• 4.   Deedwania PC, Singh BN, Ellenbogen K, Fisher S, Fletcher R, Singh SN. Spontaneous conversion and maintenance of sinus rhythm by amiodarone in patients with heart failure and atrial fibrillation: Observations from the veterans affairs congestive heart failure survival trial of antiarrhythmic therapy (CHF-STAT): The Department of Veterans Affairs CHF-STAT Investigators. Circulation 1998; 98: 2574–2579.
• 5.   Mahoney P, Kimmel S, DeNofrio D, Wahl P, Loh E. Prognostic significance of atrial fibrillation in patients at a tertiary medical center referred for heart transplantation because of severe heart failure. Am J Cardiol 1999; 83: 1544–1547.
• 6.   Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: Epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003; 91: 2d–8d.
• 7.   Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: A retrospective analysis of the SOLVD trials: Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998; 32: 695–703.
• 8.   Swedberg K, Olsson LG, Charlesworth A, Cleland J, Hanrath P, Komajda M, et al. Prognostic relevance of atrial fibrillation in patients with chronic heart failure on long-term treatment with beta-blockers: Results from COMET. Eur Heart J 2005; 26: 1303–1308.
• 9.   Kober L, Swedberg K, McMurray JJ, Pfeffer MA, Velazquez EJ, Diaz R, et al. Previously known and newly diagnosed atrial fibrillation: A major risk indicator after a myocardial infarction complicated by heart failure or left ventricular dysfunction. Eur J Heart Fail 2006; 8: 591–598.
• 10.   Chen YT, Vaccarino V, Williams CS, Butler J, Berkman LF, Krumholz HM. Risk factors for heart failure in the elderly: A prospective community-based study. Am J Med 1999; 106: 605–612.
• 11.   He J, Ogden LG, Bazzano LA, Vupputuri S, Loria C, Whelton PK. Risk factors for congestive heart failure in US men and women: NHANES I epidemiologic follow-up study. Arch Intern Med 2001; 161: 996–1002.
• 12.   Wilhelmsen L, Rosengren A, Eriksson H, Lappas G. Heart failure in the general population of men: Morbidity, risk factors and prognosis. J Intern Med 2001; 249: 253–261.
• 13.   Kenchaiah S, Evans JC, Levy D, Wilson PW, Benjamin EJ, Larson MG, et al. Obesity and the risk of heart failure. N Engl J Med 2002; 347: 305–313.
• 14.   Lavie CJ, Milani RV, Ventura HO, Cardenas GA, Mehra MR, Messerli FH. Disparate effects of left ventricular geometry and obesity on mortality in patients with preserved left ventricular ejection fraction. Am J Cardiol 2007; 100: 1460–1464.
• 15.   Wang TJ, Parise H, Levy D, D’Agostino RB Sr, Wolf PA, Vasan RS, et al. Obesity and the risk of new-onset atrial fibrillation. JAMA 2004; 292: 2471–2477.
• 16.   Meta-analysis Global Group in Chronic Heart Failure (MAGGIC). The survival of patients with heart failure with preserved or reduced left ventricular ejection fraction: An individual patient data meta-analysis. Eur Heart J 2012; 33: 1750–1757.
• 17.   Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report. National Institutes of Health. Obes Res 1998; 6(Suppl 2): 51s–209s.
• 18.   Obesity: Preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000; 894: i–xii, 1–253.
• 19.   Pedersen OD, Bagger H, Kober L, Torp-Pedersen C. Impact of congestive heart failure and left ventricular systolic function on the prognostic significance of atrial fibrillation and atrial flutter following acute myocardial infarction. Int J Cardiol 2005; 100: 65–71.
• 20.   Corell P, Gustafsson F, Schou M, Markenvard J, Nielsen T, Hildebrandt P. Prevalence and prognostic significance of atrial fibrillation in outpatients with heart failure due to left ventricular systolic dysfunction. Eur J Heart Fail 2007; 9: 258–265.
• 21.   Fonarow GC, Srikanthan P, Costanzo MR, Cintron GB, Lopatin M. An obesity paradox in acute heart failure: Analysis of body mass index and inhospital mortality for 108,927 patients in the Acute Decompensated Heart Failure National Registry. Am Heart J 2007; 153: 74–81.
• 22.   Littnerova S, Parenica J, Spinar J, Vitovec J, Linhart A, Widimsky P, et al. Positive influence of being overweight/obese on long term survival in patients hospitalised due to acute heart failure. PLoS One 2015; 10: e0117142.
• 23.   Padwal R, McAlister FA, McMurray JJ, Cowie MR, Rich M, Pocock S, et al. The obesity paradox in heart failure patients with preserved versus reduced ejection fraction: A meta-analysis of individual patient data. Int J Obes (Lond) 2014; 38: 1110–1114.
• 24.   Rosenberg MA, Manning WJ. Diastolic dysfunction and risk of atrial fibrillation: A mechanistic appraisal. Circulation 2012; 126: 2353–2362.
• 25.   Anter E, Jessup M, Callans DJ. Atrial fibrillation and heart failure: Treatment considerations for a dual epidemic. Circulation 2009; 119: 2516–2525.
• 26.   Olsson LG, Swedberg K, Ducharme A, Granger CB, Michelson EL, McMurray JJ, et al. Atrial fibrillation and risk of clinical events in chronic heart failure with and without left ventricular systolic dysfunction: Results from the Candesartan in Heart failure-Assessment of Reduction in Mortality and morbidity (CHARM) program. J Am Coll Cardiol 2006; 47: 1997–2004.
• 27.   Pai RG, Varadarajan P. Prognostic significance of atrial fibrillation is a function of left ventricular ejection fraction. Clin Cardiol 2007; 30: 349–354.
• 28.   Mamas MA, Caldwell JC, Chacko S, Garratt CJ, Fath-Ordoubadi F, Neyses L. A meta-analysis of the prognostic significance of atrial fibrillation in chronic heart failure. Eur J Heart Fail 2009; 11: 676–683.
• 29.   Eapen ZJ, Greiner MA, Fonarow GC, Yuan Z, Mills RM, Hernandez AF, et al. Associations between atrial fibrillation and early outcomes of patients with heart failure and reduced or preserved ejection fraction. Am Heart J 2014; 167: 369–375.e362.
• 30.   Cheng M, Lu X, Huang J, Zhang J, Zhang S, Gu D. The prognostic significance of atrial fibrillation in heart failure with a preserved and reduced left ventricular function: Insights from a meta-analysis. Eur J Heart Fail 2014; 16: 1317–1322.
• 31.   Horwich TB, Fonarow GC, Hamilton MA, MacLellan WR, Woo MA, Tillisch JH. The relationship between obesity and mortality in patients with heart failure. J Am Coll Cardiol 2001; 38: 789–795.
• 32.   Hamaguchi S, Tsuchihashi-Makaya M, Kinugawa S, Goto D, Yokota T, Goto K, et al. Body mass index is an independent predictor of long-term outcomes in patients hospitalized with heart failure in Japan. Circ J 2010; 74: 2605–2611.
• 33.   Sharma A, Lavie CJ, Borer JS, Vallakati A, Goel S, Lopez-Jimenez F, et al. Meta-analysis of the relation of body mass index to all-cause and cardiovascular mortality and hospitalization in patients with chronic heart failure. Am J Cardiol 2015; 115: 1428–1434.
• 34.   Madamanchi C, Alhosaini H, Sumida A, Runge MS. Obesity and natriuretic peptides, BNP and NT-proBNP: Mechanisms and diagnostic implications for heart failure. Int J Cardiol 2014; 176: 611–617.
• 35.   Lavie CJ, Milani RV, Ventura HO. Obesity, heart disease, and favorable prognosis: Truth or paradox? Am J Med 2007; 120: 825–826.
• 36.   von Haehling S, Lainscak M, Springer J, Anker SD. Cardiac cachexia: A systematic overview. Pharmacol Ther 2009; 121: 227–252.
• 37.   Anker SD, Coats AJ. Cardiac cachexia: A syndrome with impaired survival and immune and neuroendocrine activation. Chest 1999; 115: 836–847.
• 38.   Mustafa I, Leverve X. Metabolic and nutritional disorders in cardiac cachexia. Nutrition 2001; 17: 756–760.
• 39.   Araujo JP, Lourenco P, Rocha-Goncalves F, Ferreira A, Bettencourt P. Nutritional markers and prognosis in cardiac cachexia. Int J Cardiol 2011; 146: 359–363.
• 40.   Kato TS, Di Tullio MR, Qian M, Wu M, Thompson JL, Mann DL, et al. Clinical and echocardiographic factors associated with new-onset atrial fibrillation in heart failure: Subanalysis of the WARCEF Trial. Circ J 2016; 80: 619–626.
• 41.   Florea VG, Henein MY, Rauchhaus M, Koloczek V, Sharma R, Doehner W, et al. The cardiac component of cardiac cachexia. Am Heart J 2002; 144: 45–50.
• 42.   Florea VG, Moon J, Pennell DJ, Doehner W, Coats AJ, Anker SD. Wasting of the left ventricle in patients with cardiac cachexia: A cardiovascular magnetic resonance study. Int J Cardiol 2004; 97: 15–20.
• 43.   Pulignano G, Del Sindaco D, Tinti MD, Di Lenarda A, Alunni G, Senni M, et al. Atrial fibrillation, cognitive impairment, frailty and disability in older heart failure patients. J Cardiovasc Med (Hagerstown) 2016; 17: 616–623.
• 44.   Fumagalli S, Tarantini F, Guarducci L, Pozzi C, Pepe G, Boncinelli L, et al. Atrial fibrillation is a possible marker of frailty in hospitalized patients: Results of the GIFA Study. Aging Clin Exp Res 2010; 22: 129–133.
• 45.   Oreopoulos A, Fonarow GC, Ezekowitz JA, McAlister FA, Sharma AM, Kalantar-Zadeh K, et al. Do anthropometric indices accurately reflect directly measured body composition in men and women with chronic heart failure? Congest Heart Fail 2011; 17: 90–92.
• 46.   Mountantonakis SE, Grau-Sepulveda MV, Bhatt DL, Hernandez AF, Peterson ED, Fonarow GC. Presence of atrial fibrillation is independently associated with adverse outcomes in patients hospitalized with heart failure: An analysis of get with the guidelines-heart failure. Circ Heart Fail 2012; 5: 191–201.
• 47.   Lee DS, Ghosh N, Floras JS, Newton GE, Austin PC, Wang X, et al. Association of blood pressure at hospital discharge with mortality in patients diagnosed with heart failure. Circ Heart Fail 2009; 2: 616–623.
• 48.   Habal MV, Liu PP, Austin PC, Ross HJ, Newton GE, Wang X, et al. Association of heart rate at hospital discharge with mortality and hospitalizations in patients with heart failure. Circ Heart Fail 2014; 7: 12–20.