Gender‐specific association between body mass index and all‐cause mortality in patients with atrial fibrillation

Abstract Background Elevated body mass index (BMI) is related with reduced mortality in various cardiovascular diseases. Hypothesis Gender‐specific association between BMI and mortality exists in atrial fibrillation (AF). Methods In this multicenter observational study with a mean follow‐up of 1 year, a total of 1991 AF patients were enrolled and divided into two groups based on the gender. The primary endpoint was all‐cause mortality while the secondary endpoints were defined as cardiovascular mortality, stroke, and major adverse events during 1‐year follow‐up. Cox regression was performed to identify the association between BMI and clinical outcomes according to gender. Results Female patients with AF tended to be older (P = .027) and thinner (P < .001) than male patients with AF. They were more likely to have heart failure, hyperthyroidism, and valvular AF (all P < .05), but less likely to have coronary artery disease and prior myocardial infarction (all P < .01). Multivariate analysis revealed that overweight (HR(95%CI): 0.55(0.41‐0.75), P < .001) and obese patients (HR(95%CI): 0.56(0.34‐0.94), P = .028) were associated with significant lower all‐cause mortality compared with normal weight patients for the entire cohort. Similar association between elevated BMI and reduced all‐cause mortality were only identified in female patients with AF (overweight vs normal weight: HR(95%CI): 0.43(0.27‐0.70); obesity vs normal weight: HR(95%CI): 0.46(0.22‐0.97)), but not in male patients with AF. Conclusion This study indicates that overweight and obesity were related with improved survival in patients with AF. The association between elevated BMI and reduced mortality was dependent on gender, which was only significant in female patients, rather than male patients.


| INTRODUCTION
As a global health issue, the prevalence of overweight and obesity has increased rapidly over the years due to lifestyle changes. 1 Elevated body mass index (BMI) is a well-established risk factor for various cardiovascular diseases such as hypertension, diabetes mellitus, heart failure, and coronary artery disease. 2 Atrial fibrillation (AF) is one of the most common arrhythmia and is related with notably higher incidence of substantial complications such as stroke and heart failure. 3,4 In patients with AF, overweight and obesity are prevalent and are verified to increase the risk of AF occurrence and recurrence. 5,6 It is well established that overweight and obesity are associated with increased mortality in the general population. 7 On the contrary, a plenty of studies have demonstrated that higher BMI might be associated with lower mortality in the setting of various cardiovascular diseases including hypertension, 8 diabetes, 9 coronary artery disease, 10 heart failure, 11,12 stroke or transient ischemic attack (TIA). 13 In patients with AF, the counterintuitive association between elevated BMI and reduced mortality has also been detected, which is referred to as "obesity paradox." 14 Significant disparities exist in the presentation and mechanism of obesity between male and female patients. 15 Previous studies have suggested that the relation between obesity and mortality might be dependent on gender, not only in general population, 16 but also in patients with heart failure, 11,12 coronary artery disease, 17 and chronic kidney disease. 18 In patients with AF, the clinical characteristics and prognosis are also quite different between male and female patients. 19,20 However, little is known about whether gender has an impact on the association between obesity and mortality in AF patients. Therefore, a multicenter observational study conducted in Chinese AF patients was utilized to explore this issue.

| Baseline
Baseline data about demographics information, medical histories, admission vital signs, and treatments were collected by interviewing the patients, consulting their treating physicians and reviewing their medical records. BMI was recorded on admission and calculated by dividing weight in kilograms by the square of height in meters. AF was classified to paroxysmal, persistent, and permanent according to guidelines. 3

| Follow-up and outcomes
Follow-up was carried out by trained research personnel via telephone interview, outpatient visit, or medical records procurement, with a mean duration of 1 year. The primary endpoint was all-cause mortality, including cardiovascular and non-cardiovascular mortality.
The secondary endpoints were defined as cardiovascular mortality, stroke, and major adverse events (MAEs) during 1-year follow-up. All outcomes were adjudicated by an independent committee blinded to the patients according to standardized definitions. Deaths and its causes were determined by medical records obtained and reports of the participants' relatives or physicians. Cardiovascular deaths included deaths due to heart failure, myocardial infarction, sudden/arrhythmic death, stroke, pulmonary embolus, peripheral embolus, aortic dissection, or other cardiovascular disorders. MAEs referred to composite endpoint events of all-cause mortality, stroke, non-central nervous system (CNS) embolism, and major bleeding. Stroke was defined as focal neurological deficits lasting more than 24 hours and confirmed by imaging. Non-CNS embolism was defined as a vascular occlusion due to embolism confirmed by imaging or surgery. Major bleeding was defined as life-threatening bleeding, and/or symptomatic bleeding in a critical area or organ, such as intra-cranial, or pericardial, or intra muscular with compartment syndrome, and/or bleeding causing a fall in hemoglobin level of 20 g/L (1.24 mmol/L) or more, or leading to transfusion of two or more units of whole blood or red cells.

| Statistical analysis
Continuous variables were presented as medians with interquartile ranges and compared by Mann-Whitney U test for the data were not normally distributed. Categorical variables were presented as frequencies and percentages and compared by Chi-square test. BMI was evaluated both as a continuous variable and as a categorical variable.

| DISCUSSION
In the present study, there were differences between male and female patients with AF in some aspect of clinical characteristics, comorbidities, medications, and prognosis. In the entire cohort, overweight and obesity were associated with significant lower all-cause mortality, which was in-line with the so-called "obesity paradox." However, when divided by gender, the association between elevated BMI and reduced all-cause mortality was only significant in female patients, rather than in male patients.
Significant differences exist in the clinical characteristics and prognosis between male and female patients with AF. 19 In this cohort, female patients with AF tended to have a lower rate of coronary artery disease and a higher rate of heart failure, hyperthyroidism, and valvular AF. During 1-year follow-up, they were more likely to suffer from stroke, which was consistent with previous reports. 19 Obesity and overweight are well-established risk factors for cardiovascular diseases, AF included. 1,2 Although elevated BMI was associated with increased mortality in general population, 16 plentiful studies have demonstrated a converse relation between BMI and allcause mortality in the setting of numerous cardiovascular diseases. [8][9][10][11][12][13] 13 As to patients with AF, the relation between BMI and mortality was controversial in previous studies. 14,22 In our study, AF patients in overweight and obesity groups had significantly lower risk of mortality than patients in normal weight group, which was in accordance with the obesity paradox.
The exact mechanisms of obesity paradox in AF have not been fully elucidated yet and might be multifactorial. One explanation is that obesity is a well-established risk factor for heavier symptom burdens 6 and more cardiovascular comorbidities in patients with AF. 14,22 These could lead to earlier diagnosis and better management. As a result, cardioprotective medications are adopted more frequently and aggressively in obese patients. 14,22 In addition, there might exist a "healthy survivor effect." The prevalence of AF increases significantly with age. 3,4 Severe obese patients might die before developing AF due to comorbidities, leaving the rest obese AF patients with relatively favorable prognosis. In view of the above reasons, confounding effects should not be ignored in obesity paradox. 14,22 However, after adjustment for potential confounders, the relation between BMI and mortality still existed in our study. Another explanation might be that obese patients have better metabolic reserves to cope with increased metabolic stress in the setting of various diseases. 23 AF is a kind of chronic disease with elevated energy and protein consumption, which is associated with worse prognosis. 3 Adipose tissues could serve as energy reserves and play a positive role in delaying malnutrition and energy wastage caused by illnesses. 23 On the other hand, female patients with AF in the present study had a significantly higher rate of valvular AF, which was a chronic consumptive disease and might benefit more from metabolic reserves of excessive adipose tissues. 3 These confounding factors might partly account for the sex-specific association between BMI and mortality.
However, the result remained the same after adjustment for these potential confounding factors in the present study. Second, there exist sex-related differences in the clinical characteristics of obese patients.
BMI, a parameter combining both fat and lean body mass, is used to quantify overweight and obesity. 30 Gender differences in body composition should not be ignored. Women usually have a higher percentage of fat mass than men with an equivalent BMI. 15,30 Since high percentage of fat mass is supposed to be protective for favorable survival due to better metabolic reserves, the relation between obesity and reduced mortality in female patients with chronic diseases is reasonable. On the other hand, the condition of fat distribution could not be reflected by BMI, either. 30 Excessive fat stored in visceral fat deposits is more common in men, while in women, it is usually distributed in peripheral subcutaneous tissue. 15,31 Excessive visceral fat could increase the risk of developing metabolic syndrome and cardiovascular diseases, whereas femoral-gluteal fat might be beneficial as a "sink" for lipid. 32 Finally, greater myocardial fatty acid metabolism and lower myocardial glucose utilization have been observed in obese women, which might be partly attributed to the effect of estrogen. 33 Estrogen can increase the activity of lipoprotein lipase and fatty acid oxidation enzyme, thus improving the myocardial fatty acid metabolism. 34,35 On the other hand, estrogen could reduce glucose oxidation, gluconeogenesis and glycogenolysis in other organs and decrease glucose transporter 4 translocation to the cell surface, thereby inhibiting glucose utilization. 35,36 Similar changes have also been detected in postmenopausal women with chronic estrogen replacement. 37 In all, female myocardium might be more dependent on fatty acids F I G U R E 2 Multivariate adjusted hazard ratios of all-cause mortality categorized by sex and body mass index (BMI) metabolism for energy utilization than male myocardium. This could be one potential explanation for the survival advantages in female patients with obesity. In addition, excessive adipose tissue could increase the circulating levels of estrogen in obese women and exert further beneficial impacts on their prognosis. 38,39 The study had several limitations need to be pointed out. First, as one of the most common parameter of obesity, BMI has defects in reflecting the percentage and distribution of adipose tissues. However, our study lacked data of determinants for central obesity, such as waist circumference and waist-to-hip ratio, which have been confirmed to be better predictors of clinical outcomes. 40 Second, only baseline BMI was available in the present study. The relation between BMI changes and outcomes was unknown due to lack of serial BMI during follow-up. Third, some potentially relative factors, such as socioeconomic situation, cardiorespiratory fitness, nutritional status, unintentional and intentional weight loss, have not been collected and adjusted in this study. These residual confounders might have influenced the accuracy of our results. Finally, this is a post hoc observational study with inherent defects. The gender-specific relation between BMI and mortality could only be interpreted as associative but not causal. Therefore, large prospective multicenter studies with rational design are needed to confirm our results.

| CONCLUSION
Our study has demonstrated that overweight and obesity were related with decreased mortality in patients with AF. The association between elevated BMI and reduced mortality was dependent on gender. The phenomenon of obesity paradox could only be detected in female patients with AF. The exact mechanisms of the gender-specific association between BMI and mortality have not been fully elucidated and require further investigation.