Abstract
Our knowledge of the association between abdominal obesity (AO) and the risk of atrial fibrillation (AF) after adjusting for body mass index (BMI) is limited. We included 11,617 Black and White participants (mean age 63.0 ± 8.4 years) from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) national cohort study who were free of AF at baseline. A multivariable logistic regression model was used to estimate the odds ratio (OR) with 95% confidence interval (CI) of incident AF associated with AO. We also evaluated the association between waist circumference (WC) and incident AF. Over a median follow-up of 9.4 years, 999 participants developed AF. AO was associated with an increased risk of AF in a multivariable model adjusted for sociodemographic, lifestyle, and cardiovascular risk factors (OR 1.43, 95% CI 1.24 to 1.65, p <0.001). The association was attenuated after adjusting for BMI (OR 1.13, 95% CI 0.95 to 1.35, p = 0.16). There was no evidence of interaction between AO and incident AF by age category (age >65 vs age ≤65), gender, race, obesity, or BMI category. Conversely, a 10cm increase in WC was associated with a higher incidence of AF after controlling for BMI (OR 1.18 95% CI 1.09 to 1.29, p <0.001), in both nonobese (OR 1.14, 95% CI 1.03 to 1.28, p = 0.02) and obese (OR 1.26, 95% CI 1.11 to 1.42, p <0.001) people. In conclusion, there was an association between AO and incident AF, but the association was weakened after adjusting for BMI. There was a significant association between WC and incident AF, after taking other AF risk factors and BMI into account. WC is a potentially modifiable risk factor for AF, and further research is warranted to explore the effect of decreasing WC on the population AF burden.
Previous studies demonstrated that subjects with obesity have a higher risk of atrial fibrillation (AF) compared with nonobese counterparts.1,2 Several anthropometric measures, including body mass index (BMI), waist circumference (WC), and waist-to-hip ratio, have been used to identify obese subjects. Obesity defined by BMI is widely used, but increasing evidence shows that it can miss a significant portion of subjects with increased cardiovascular risk.3,4 Studies have suggested that abdominally defined obesity (AO) has a stronger association with cardiometabolic disease than BMI-defined obesity.3 Previous studies attempting to elucidate the association of AO and incident AF were limited by relatively short follow-up, lack of proper adjustment for BMI, and a homogenous ethnic population.5,6 Therefore, we propose to examine the association between AO and incident AF in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) population. We also assess the association between WC and incident AF, stratified by the presence of BMI-defined obesity.
Methods
Details of the study design and methodology of REGARDS have been published previously.7 The REGARDS study is a prospective cohort study designed to identify contributors to the Black-White and regional disparities in stroke mortality. Between January 2003 and October 2007, 30,239 participants aged 45 years or older were recruited from the continental United States. Written informed consent was obtained from all participants and the study was approved by the institutional review boards of all participating universities.
Of the 30,239 participants initially enrolled, 15,521 participants were still active in the study and completed a follow-up examination conducted between 2013 and 2016, approximately 10 years after the first in-home visit. Of those, we excluded 3,904 participants with baseline AF or missing data on AF, BMI, WC, or other covariates. The final sample included 11,617 participants with complete data at baseline and follow-up visits.
Incident AF was identified by a resting electrocardiogram obtained during the home visit and a self-reported history of AF diagnosis by a physician during the computer-assisted telephone interview system surveys. Self-reported data were used for the following variables: age, gender, race, education (college graduate and above, some college, high school graduate, less than high school), income (less than $20k per year, $20k–$34k per year, $35k–$74k per year, $75k and above per year, refused to answer), region of residence (belt, buckle, nonbelt), smoking status (current, never, past), physical activity (none, 1 to 3 times per week, 4 or more times per week), alcohol intake (current, never, past), and history of stroke. History of coronary heart disease was ascertained by a self-reported history of myocardial infarction, coronary artery bypass grafting, angioplasty or stenting, or electrocardiographic evidence of previous myocardial infarction. AO was defined as WC >102 cm in men and >88 cm in women. Underweight, normal weight, overweight, and obesity were defined as BMI of <18.5, 18.5 to 24.9, 25 to 29.9, and ≥30 kg/m2, respectively.
We performed descriptive statistics for demographic, socioeconomic, lifestyle, anthropometric, and medical history variables at the baseline assessment according to the presence of AO. Frequencies and percentages were used to describe categorical variables. Means and standard deviations were used to describe continuous variables. We tested for differences in distributions of variables between exposure groups using the chi-square test for categorical variables and analysis of variance test for continuous variables.
Multivariable logistic regression analyses were performed to estimate the odds ratio (OR) with 95% confidence interval (CI) of incident AF associated with AO. Variables were selected based on previously published AF risk factors.8,9 Multivariable models were fit sequentially as follows: Model 1: Sociodemographic: age (as a continuous variable), gender, race, education, household income, and region; Model 2: variables in Model 1 plus lifestyle factors (smoking, physical activity, alcohol intake); Model 3: variables in Model 2 plus cardiovascular disease (CVD) risk factors (hypertension, dyslipidemia, diabetes mellitus, history of coronary artery disease, history of stroke); Model 4: variables in Model 3 plus BMI (as a continuous variable).
Effect modification by age group >65 versus ≤65, race, gender, BMI-defined obesity, and BMI category was evaluated by stratified analysis and comparison of models with and without interaction terms using the likelihood ratio test. Potential effect modifiers were selected based on previous studies assessing the association between AO and incident AF.5
We also evaluated the association between WC as a continuous variable and incident AF. We conducted a multivariable logistic regression analysis to estimate the OR with 95% CI of incident AF associated with WC, with the same multivariable models used for the association between AO and incident AF. We tested whether the association between WC and incident AF differed across BMI-defined obesity status by introducing WC BMI-defined obesity into the logistic regression model. A stratified analysis based on BMI-defined obesity was performed as well. All statistical analyses were performed using R Statistical Software version 4.0.0 (Foundation for Statistical Computing, Vienna, Austria) at α = 0.05 significance level.
Results
A total of 11,617 participants were included in this analysis. Table 1 lists the baseline characteristics of the study participants stratified by AO. Compared with those without AO, participants with AO were more likely to be Black, have lower education levels, and have lower income. There was a higher prevalence of hypertension, hyperlipidemia, diabetes mellitus, and a history of CVD observed in those with AO compared with those without.
Table 1.
Baseline characteristics of participants stratified by abdominal obesity
| Variable | Overall | Abdominal obesity |
p value | |
|---|---|---|---|---|
| (n=11617) | No (n=6248) | Yes (n=5369) | ||
|
| ||||
| BMI category | <0.001 | |||
| <18.5 | 96 (0.8%) | 92 (1.5%) | 4 (0.1%) | |
| 18.5–24.9 | 2699 (23.2%) | 2564(41.0%) | 135 (2.5%) | |
| 25–29.9 | 4474 (38.5%) | 3018(48.3%) | 1456 (27.1%) | |
| ≥ 30 | 4348 (37.4%) | 574 (9.2%) | 3774 (70.3%) | |
| BMI (kg/m2, mean (SD)) | 29.18 (5.87%) | 25.68 (3.35%) | 33.25 (5.55%) | <0.001 |
| Waist circumference (cm, mean (SD)) | 94.99 (14.95%) | 85.69 (10.12%) | 105.82 (12.11%) | <0.001 |
| Age (years, mean (SD)) | 62.99 (8.35%) | 63.13 (8.63%) | 62.83 (8.00%) | 0.053 |
| White | 7452 (64.1%) | 4447 (71.2%) | 3005 (56.0%) | <0.001 |
| Male | 5208 (44.8%) | 3370 (53.9%) | 1838 (34.2%) | <0.001 |
| Education category | <0.001 | |||
| College graduate and above | 5008 (43.1%) | 3011 (48.2%) | 1997 (37.2%) | |
| Some college | 3071 (26.4%) | 1598 (25.6%) | 1473 (27.4%) | |
| High school graduate | 2698 (23.2%) | 1294 (20.7%) | 1404 (26.2%) | |
| Less than high school | 840 (7.2%) | 345 (5.5%) | 495 (9.2%) | |
| Income | <0.001 | |||
| < $20k | 1363 (11.7%) | 518(8.4%) | 829 (15.5%) | |
| $20k–$34.9k | 2530 (21.8%) | 1210 (19.4%) | 1320 (24.6%) | |
| $35k–$74.9k | 3985 (34.3%) | 2227 (35.6%) | 1758 (32.7%) | |
| ≥ $75k | 2533 (21.8%) | 1600 (25.6%) | 933 (17.4%) | |
| Refused | 1206 (10.4%) | 677(10.8%) | 529 (9.9%) | |
| Region | 0.336 | |||
| Belt | 3905 (33.6%) | 2069 (33.1%) | 1836 (34.2%) | |
| Buckle | 2539 (21.9%) | 1359 (21.8%) | 1180 (22.0%) | |
| Non-belt | 5173 (44.5%) | 2820 (45.1%) | 2353 (43.8%) | |
| Smoker | 0.002 | |||
| Current | 1271 (10.9%) | 725 (11.6) | 546 (10.2%) | |
| Never | 5735 (49.4%) | 3125 (50.0) | 2610 (48.6%) | |
| Past | 4611 (39.7%) | 2398 (38.4%) | 2213 (41.2%) | |
| Exercise category | <0.001 | |||
| None | 3361 (28.9%) | 1472 (23.6%) | 1889 (35.2%) | |
| 1 to 3 times per week | 4590 (39.5%) | 2491 (39.9%) | 2099 (39.1%) | |
| 4 or more per week | 3666 (31.6%) | 2285 (36.6%) | 1381 (25.7%) | <0.001 |
| Alcohol Use | ||||
| Current | 6710 (57.8%) | 3920 (62.7%) | 2790 (52.0%) | |
| Never | 3196 (27.5%) | 1507 (24.1%) | 1689 (31.5%) | |
| Past | 1711 (14.7%) | 821 (13.1%) | 890 (16.6%) | |
| Hypertension | 6118 (52.7%) | 2649 (42.4%) | 3469 (64.6%) | <0.001 |
| Hyperlipidemia | 3714 (32.0%) | 1799 (28.8%) | 1915 (35.7%) | <0.001 |
| Diabetes mellitus | 1892 (16.3%) | 530 (8.5%) | 1362 (25.4%) | <0.001 |
| Prior CHD | 1359 (11.7%) | 678 (10.9%) | 681 (12.7%) | 0.002 |
| Prior stroke | 390 (3.4%) | 170 (2.7%) | 220 (4.1%) | <0.001 |
BMI = body mass index; CHD = coronary heart disease.
In a multivariable model adjusted for sociodemographic, lifestyle, and CVD risk factors, AO was associated with a higher risk of AF (OR 1.43, 95% CI 1.24 to 1.65, p <0.001). The association was no longer significant after adjusting for BMI (OR 1.13, 95% CI 0.95 to 1.35, p = 0.16) (Table 2). There was no evidence of interaction between AO and incident AF by age category (age >65 vs age ≤65), gender, race, obesity, or BMI category (Figure 1).
Table 2.
Association between abdominal obesity and incident atrial fibrillation
| Abdominal obesity | # case/# at risk (%) | Odds Ratio (95% Confidence Interval) |
|||
|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | ||
|
| |||||
| Absent | 484/6248 (7.7%) | Reference | Reference | Reference | Reference |
| Present | 515/5369 (9.6%) | 1.60 (1.40–1.83) | 1.59 (1.38–1.82) | 1.43 (1.24–1.65) | 1.13 (0.95–1.35) |
Model 1: adjusted for age, sex, race/ethnicity, education, household income and region
Model 2: model 1 plus smoking, physical activity, alcohol intake
Model 3: model 2 plus hypertension, dyslipidemia, diabetes, history of CVD, history of stroke
Model 4: model 3 plus BMI
Figure 1.
Association of abdominal obesity and atrial fibrillation in subgroups. AO = abdominal obesity.
When WC was considered as a continuous variable, a 10cm increase in WC was associated with an increased risk of AF (OR 1.27, 95% CI 1.21 to 1.33, p <0.001) in a multivariable model adjusted for sociodemographic factors (Model 1). The association remained significant in a multivariable model adjusted for sociodemographic, lifestyle, CVD risk factors, and BMI (Model 4). Every 10cm increase of WC was associated with an 18% increased odds of incident AF (OR 1.18 95% CI 1.09 to 1.29, p <0.001) (Figure 2). In a stratified analysis based on BMI-defined obesity, WC was associated with AF in both nonobese (OR 1.14, 95% CI 1.03 to 1.28, p = 0.02) and obese (OR 1.26, 95% CI 1.11 to 1.42, p <0.001) subjects after adjusting for variables in Model 4. There was no evidence of interaction by BMI-defined obesity status (p value = 0.09).
Figure 2.
Forest plot of ORs with 95% CI examining the association of relevant covariates with incident atrial fibrillation.
Discussion
In this large prospective cohort study in a biracial population with long-term follow-up, we found that the presence of AO was associated with an elevated risk of AF, but the association was attenuated after adjusting for BMI. When WC was considered as a continuous variable, it was associated with a higher risk of AF both in nonobese and obese participants after taking other AF risk factors and BMI into account.
Our study has important public health implications. It is well established that AO is associated with worse cardiovascular outcomes, including the risk of AF.6,10 However, there have been major debates on whether AO provides an incremental prognostic value beyond BMI.11,12 In a pooled analysis of prospective cohort studies from the BMI and mortality projects, WC was independently associated with increased mortality at all levels of BMI from 20 to 50 kg/m2.10 On the other hand, an individual level pooled analysis from the Emerging Risk Factors Collaboration study showed that a cardiovascular prediction model incorporating AO did not improve the cardiovascular risk prediction accuracy.13 Our results suggest that there is an association between AO and AF, but the association is not independent of BMI.
In our analysis, WC was associated with an increased risk of AF after adjusting for multiple variables including BMI when it is considered as a continuous variable. This association was significant in both the obese and nonobese groups. The cutoff for the diagnosis of AO (WC >102 cm in men and >88 cm in women) defined by the World Health Organization (WHO) is derived mainly from a White cohort.14,15 There is increasing evidence that this definition likely misclassifies people at increased risk in a multiracial population.14 For instance, a cohort study of East Asian participants reported that WC cut points of 85 cm in men and 82 cm in women are better predictors of metabolic syndrome than WHO cut points.16 Another study from the National Health and Nutrition Examination Survey also demonstrated that current cut points defined by WHO had a sensitivity of 37.4% in non-Hispanic Black men to detect the presence of at least 1 cardiovascular risk factor.17 Further studies are indicated to investigate the prognostic utility of adding WC to other established AF risk prediction models.
In a subgroup analysis, we found no evidence of interaction between AO and AF based on age group, gender, race, or BMI category. Previously published epidemiologic studies demonstrated a significant difference in AF incidence based on gender and race. For instance, Black participants have a lower AF incidence than White participants after adjusting for known risk factors.18 Similarly, higher incidence and prevalence of AF are noted in men than in women.8,19 Multiple pathophysiologic risk factors have been suggested to explain these differences, such as left atrial size or visceral adipose tissue.20–22 Our findings imply that AO may not play a major role in the pathogenesis of gender and racial differences in AF.
Our study has strengths, including a prospective study design with a large, biracial population and long-term follow-up. We were able to adjust for potential confounders and mediators on the association between AO and incident AF. Several potential limitations need to be considered when interpreting our results. First, AF cases were ascertained by follow-up electrocardiograms and self-reported AF, therefore asymptomatic paroxysmal AF could have been missed. There may be attrition bias because of subjects missing the second visit. Although comprehensive adjustments for confounding factors were performed, potential confounders were assessed only at baseline and these parameters could have changed over time. Finally, the generalizability of the study results to other ethnicities such as Hispanics or Asians is unknown.
In conclusion, our analysis of a biracial population with long-term follow-up demonstrates that the presence of AO is associated with incident AF, but the association is not independent of BMI. There is an association between WC and incident AF after controlling for BMI, in both obese and nonobese populations. WC is a potentially modifiable risk factor for AF.
Acknowledgment
The authors thank the investigators, staff, and participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org.
This work is supported by cooperative agreement U01 NS041588 cofunded by the National Institute of Neurological Disorders and Stroke, Bethesda, Maryland and the National Institute on Aging, National Institutes of Health, Bethesda, Maryland.
Disclosures
Dr. Levitan receives research funding from Amgen and has received consulting fees for a scientific project funded by Novartis, unrelated to the current work. All other authors have no conflicts to declare.
References
- 1.Wang TJ, Parise H, Levy D, D’Agostino RB Sr, Wolf PA, Vasan RS, Benjamin EJ. Obesity and the risk of new-onset atrial fibrillation. JAMA 2004;292:2471–2477. [DOI] [PubMed] [Google Scholar]
- 2.Frost L, Hune LJ, Vestergaard P. Overweight and obesity as risk factors for atrial fibrillation or flutter: the Danish Diet, Cancer, and Health Study. Am J Med 2005;118:489–495. [DOI] [PubMed] [Google Scholar]
- 3.Cameron AJ, Zimmet PZ. Expanding evidence for the multiple dangers of epidemic abdominal obesity. Circulation 2008;117:1624–1626. [DOI] [PubMed] [Google Scholar]
- 4.Sun Y, Liu B, Snetselaar LG, Wallace RB, Caan BJ, Rohan TE, Neuhouser ML, Shadyab AH, Chlebowski RT, Manson JE, Bao W. Association of normal-weight central obesity With all-cause and cause-specific mortality Among postmenopausal women. JAMA Netw Open 2019;2:e197337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Baek YS, Yang PS, Kim TH, Uhm JS, Park J, Pak HN, Lee MH, Joung B. Associations of abdominal obesity and new-onset atrial fibrillation in the general population. J Am Heart Assoc 2017;6:e004705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Aronis KN, Wang N, Phillips CL, Benjamin EJ, Marcus GM, Newman AB, Rodondi N, Satterfield S, Harris TB, Magnani JW. Health ABC study. Associations of obesity and body fat distribution with incident atrial fibrillation in the biracial health aging and body composition cohort of older adults. Am Heart J 2015;170. 498–505.e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Howard VJ, Cushman M, Pulley L, Gomez CR, Go RC, Prineas RJ, Graham A, Moy CS, Howard G. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology 2005;25:135–143. [DOI] [PubMed] [Google Scholar]
- 8.Bose A, O’Neal WT, Wu C, McClure LA, Judd SE, Howard VJ, Howard G, Soliman EZ. Sex differences in risk factors for incident atrial fibrillation (from the Reasons for Geographic and Racial Differences in Stroke [REGARDS] Study). Am J Cardiol 2019;123:1453–1457. [DOI] [PubMed] [Google Scholar]
- 9.Schnabel RB, Yin X, Gona P, Larson MG, Beiser AS, McManus DD, Newton-Cheh C, Lubitz SA, Magnani JW, Ellinor PT, Seshadri S, Wolf PA, Vasan RS, Benjamin EJ. Levy D. 50 year trends in atrial fibrillation prevalence, incidence, risk factors, and mortality in the Framingham Heart Study: a cohort study. Lancet 2015;386:154–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cerhan JR, Moore SC, Jacobs EJ, Kitahara CM, Rosenberg PS, Adami HO, Ebbert JO, English DR, Gapstur SM, Giles GG, Horn-Ross PL, Park Y, Patel AV, Robien K, Weiderpass E, Willett WC, Wolk A, Zeleniuch-Jacquotte A, Hartge P, Bernstein L, Berrington de Gonzalez A. A pooled analysis of waist circumference and mortality in 650,000 adults. Mayo Clin Proc 2014;89:335–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ashwell M, Mayhew L, Richardson J, Rickayzen B. Waist-to-height ratio is more predictive of years of life lost than body mass index. PLoS One 2014;9:e103483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Fenger-Grøn M, Overvad K, Tjønneland A, Frost L. Lean body mass is the predominant anthropometric risk factor for atrial fibrillation. J Am Coll Cardiol 2017;69:2488–2497. [DOI] [PubMed] [Google Scholar]
- 13.Emerging Risk Factors Collaboration, Wormser D, Kaptoge S, Di Angelantonio E, Wood AM, Pennells L, Thompson A, Sarwar N, Kizer JR, Lawlor DA, Nordestgaard BG, Ridker P, Salomaa V, Stevens J, Woodward M, Sattar N, Collins R, Thompson SG, Whitlock G, Danesh J. Separate and combined associations of body-mass index and abdominal adiposity with cardiovascular disease: collaborative analysis of 58 prospective studies. Lancet 2011;377:1085–1095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rao G, Powell-Wiley TM, Ancheta I, Hairston K, Kirley K, Lear SA, North KE, Palaniappan L, Rosal MC. American Heart Association Obesity Committee of the Council on Lifestyle and Cardiometabolic Health. Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: A scientific statement From the American Heart Association. Circulation 2015;132:457–472. [DOI] [PubMed] [Google Scholar]
- 15.World Health Organization. Obesity and Overweight Fact Sheet. Available at: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight Accessed on September 14, 2021.
- 16.He J, Ma R, Liu J, Zhang M, Ding Y, Guo H, Mu L, Zhang J, Wei B, Yan Y, Ma J, Pang H, Li S, Guo S. The optimal ethnic-specific waist-circumference cut-off points of metabolic syndrome among low-income rural Uyghur adults in far Western China and implications in preventive public health. Int J Environ Res Public Health 2017;14:158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhu S, Heymsfield SB, Toyoshima H, Wang Z, Pietrobelli A, Heshka S. Race-ethnicity-specific waist circumference cutoffs for identifying cardiovascular disease risk factors. Am J Clin Nutr 2005;81:409–415. [DOI] [PubMed] [Google Scholar]
- 18.Borzecki AM, Bridgers DK, Liebschutz JM, Kader B, Kazis LE, Berlowitz DR. Racial differences in the prevalence of atrial fibrillation among males. J Natl Med Assoc 2008;100:237–245. [DOI] [PubMed] [Google Scholar]
- 19.Ko D, Rahman F, Schnabel RB, Yin X, Benjamin EJ, Christophersen IE. Atrial fibrillation in women: epidemiology, pathophysiology, presentation, and prognosis. Nat Rev Cardiol 2016;13:321–332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Carroll JF, Fulda KG, Chiapa AL, Rodriquez M, Phelps DR, Cardarelli KM, Vishwanatha JK, Cardarelli R. Impact of race/ethnicity on the relationship between visceral fat and inflammatory biomarkers. Obesity (Silver Spring) 2009;17:1420–1427. [DOI] [PubMed] [Google Scholar]
- 21.Marcus GM, Olgin JE, Whooley M, Vittinghoff E, Stone KL, Mehra R, Hulley SB, Schiller NB. Racial differences in atrial fibrillation prevalence and left atrial size. Am J Med 2010;123:375.e1–375.e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lovejoy JC, de la Bretonne JA, Klemperer M, Tulley R. Abdominal fat distribution and metabolic risk factors: effects of race. Metabolism 1996;45:1119–1124. [DOI] [PubMed] [Google Scholar]