Introduction: We investigated sex and ethnic differences in the incidence, clinical characteristics and 1-year mortality of patients with newly diagnosed AF in a multi-ethnic population.
Method: This retrospective cohort study of patients diagnosed with AF from 2008 to 2015 was based on medical claims, casemix and subvention data submitted to the Ministry of Health. Patients with AF were matched with controls without AF for age (3-year bands), sex and ethnicity, and categorised as middle-aged (45–64 years) or elderly (≥65 years) among major ethnic groups in Singapore (Chinese, Malay and Indian).
Results: Among 40,602 adults with AF (elderly 74%), Malays had the highest age-standardised incidence rate of AF, followed by Chinese and Indians; and the rate was higher in men. Despite having the worst cardiovascular risk profile, Indians had the lowest prevalence and incidence of AF. The 1-year mortality rate after newly diagnosed AF was 22–26 deaths per 100 people. Newly diagnosed AF was independently associated with increased 1-year all-cause mortality among middle-aged (adjusted odds ratio [AOR] 9.08, 95% confidence interval [CI] 7.36–11.20) and elderly adults (AOR 3.60, 95% CI 3.40–3.80) compared with those without AF. Sex differences in mortality among patients with AF were limited to elderly adults (men: AOR 1.17, 95% CI 1.11–1.24), while Indians were associated with a 30% increased odds of mortality compared with Chinese regardless of age (middle-aged: AOR 1.27, 95% CI 1.09–1.548 elderly: AOR 1.33, 95% CI 1.22–1.45).
Conclusion: Variations in incidence, clinical profile and 1-year mortality of patients with AF in a nationwide cohort were influenced by sex and ethnicity. Newly diagnosed AF portends a worse prognosis and is a marker of high mortality within the first year.
Atrial fibrillation (AF) is the most common clinically significant arrhythmia and is associated with increased risks of stroke, dementia, heart failure (HF) and death.1 Globally, 33.5 million people were reported to have AF in 2010,2 with numbers expected to increase exponentially by 2050.3-7 While in part due to ageing populations,8 increasing cardiovascular risk factors and the detection of silent AF further contribute to the growing epidemic.9
Apart from modifiable clinical risk factors, non-modifiable factors, including sex and ethnicity, exert varying influences on the prevalence of AF.10-22 Significant heterogeneities in AF prevalence were observed within Asian populations.23-25 While not fully understood, these differences may be contributed by the complex interactions between genetic and environmental factors.26,27 The effects of sex and ethnicity on mortality from AF were, however, more variable, with sex differences not observed10-13 but ethnic differences among Asians with HF reported.24
Despite the high prevalence of hypertension, diabetes and HF in Southeast Asia,28 there is a paucity of data on AF characteristics and mortality in this region.29-31 Therefore, we investigated sex and ethnic differences in the incidence, clinical characteristics and 1-year mortality of patients with newly diagnosed AF in a nationwide, multiethnic population in Singapore.
This retrospective cohort study on newly diagnosed non-valvular AF from 2008 to 2015 included subjects aged ≥45 years from the 3 major ethnic groups in Singapore (Chinese 74%, Malay 13% and Indian 9%). The sources of diagnosis data comprised anonymised individual patient-level data from medical claims, casemix and subvention data submitted to the Ministry of Health by primary care clinics (public and private) and public hospitals (emergency departments and inpatient units). These cover more than 95% of Singapore’s population with minimal selection bias. It was expected that the majority of AF would be identified through available data sources as most Singapore residents would be treated in public healthcare institutions owing to greater accessibility, extensive coverage and higher subsidy. Ethnicity was based on registered ethnicity at birth. Housing type served as a surrogate of socioeconomic status, with private housing (condominiums or private houses) corresponding to higher socioeconomic status and public housing of 1- to 2-room flats or studio apartments corresponding to lower socioeconomic status. Identification of AF and other diseases was based on codes from the International Classification of Diseases (ICD) 9th Revision, Clinical Modification (prior to 2012); and ICD 10th Revision, Australian Modification (from 2012) (Supplementary Table S1 in online Supplementary Materials). The survival status of patients was ascertained from the Singapore Registry of Birth and Death. A matched cohort without AF was obtained by matching each patient with AF to a control without AF who was alive in the index year (year of diagnosis of the matched patient) for age (3-year bands), sex and ethnicity. Population numbers were obtained from the Singapore Department of Statistics’ mid-yearly population estimates of Singapore residents for the calculation of AF incidence and prevalence. The study complied with the Declaration of Helsinki and patient consent was waived as only anonymised data were obtained from an administrative database.
Patients were stratified into 2 age groups: middle-aged (45–64 years) and elderly (≥65 years). An age cut-off of 45 years was chosen as 5% of patients with AF were below 45 years, in order to investigate patients between the 5th and 95th age percentiles. Chronic diseases at baseline were defined as a diagnosis prior to or within 6 months of first diagnosis of AF. The clinical risk factors included hypertension, hyperlipidaemia, renal impairment, end-stage renal failure, ischaemic heart disease (IHD), HF, stroke or transient ischaemic attack, haemorrhagic stroke, systemic embolism, peripheral vascular disease, dementia, depression, hyperthyroidism and cancer. CHA2DS2VASc scores were calculated for all patients.32
Crude AF incidence was estimated by the number of newly diagnosed AF patients among people at risk in the same year (with the number of diagnosed AF cases in the previous year removed from the population number in the current year). Crude AF prevalence was estimated by dividing the number of patients diagnosed with AF before and still alive on 1 January of each year by the population number in the same year. To account for the change in population structure over time, adjusted incidence and prevalence rates were calculated using direct age standardisation and the 2008 resident population as the reference population by 5-year age bands stratified by sex and ethnicity. Crude 1-year mortality was estimated by dividing the number of patients who died in the next calendar year by the number of patients in that year. Baseline characteristics between ethnicities and sexes were compared using the Wilcoxon rank sum test or Kruskal-Wallis test for continuous variables and chi-square test for categorical variables. As specific dates of death were unavailable from the National Death Registry (only the years of death were available to prevent re-identification of patients), logistic regression analyses were performed to compare the odds ratios with 95% confidence intervals (CIs) for mortality within 1 year, between the sex and ethnic groups within the cohort of AF patients, and between patients with and without AF. To determine if 2-way interactions existed between age and sex, age and ethnicity, and sex and ethnicity, these interaction terms were included in logistic regression models. All statistical analyses were performed using STATA/SE version 16 (StataCorp, College Station, US), at 5% level of significance.
From 2008 to 2015, a total of 40,602 adults were newly diagnosed with AF. The middle-aged group made up 10,671 (26%) of the patients with a mean age of 57±5 years, 32% women and 76% Chinese; the rest or 29,931 (74%) were in the elderly group with a mean age of 78±8 years, 52% women and 84% Chinese (online Supplementary Fig. S1). Temporal trends in the prevalence and incidence rates of AF are shown in Fig. 1. In both the middle-aged and elderly groups, a linear increase in the age-sex-ethnicity-adjusted prevalence rates was observed from 2008 to 2015 (middle-aged group: 59.9 to 75.2 per 10,000 people; elderly group: 525.4 to 696.5 per 10,000 people). The age-sex-ethnicity-adjusted incidence rate of AF, however, remained fairly constant between 11 and 14 per 10,000 people in the middle-aged group, while the corresponding rate increased from 101.5 to 107.4 per 10,000 people in the elderly group.Ethnic and sex differences in incidence of AF
Malays had the highest age-standardised incidence rate (ASIR) of AF, followed by Chinese and Indians, while men had a higher ASIR of AF than had women in both age groups (Fig. 1). Malay men had the highest ASIR of AF in both age groups, increasing from 20.8 per 10,000 in 2008 to 23.1 per 10,000 in 2015 in middle-aged adults, and 124.5 to 155.9 per 10,000 in elderly adults, while Indian women had the lowest ASIR of AF in both age groups. The ASIR of AF increased from 2008 to 2015 in men of all ethnicities in both age groups, while the ASIR decreased slightly among middle-aged Chinese and Indian women, but increased sharply among Malay elderly women. The temporal trend of AF by sex and ethnicity in both age groups is shown in the online Supplementary Fig. S2.
Ethnic and sex differences in risk profile of AF
Among middle-aged adults, Chinese had the lowest prevalence of hypertension, hyperlipidaemia, diabetes, renal impairment, IHD, HF and peripheral vascular disease compared with Malays and Indians at AF diagnosis (P<0.05) (Table 1). Indian men had the highest prevalence of hyperlipidaemia, diabetes, IHD and vascular disease but the lowest prevalence of hyperthyroidism. Among elderly adults, men had a higher prevalence of IHD, HF, haemorrhagic stroke and cancer (online Supplementary Table S2). When stratified by ethnicity, Indians had the highest prevalence of hypertension, hyperlipidaemia, diabetes and IHD in both men and women (Table 2), while Chinese again had the lowest prevalence of hyperlipidaemia, diabetes, renal impairment, IHD and HF but the highest cancer prevalence (P<0.05) (online Supplementary Table S3). Compared with patients without AF, those with AF had a higher prevalence of chronic conditions including hypertension, hyperlipidaemia, diabetes, IHD, HF, stroke, renal impairment, cancer, depression, dementia and thyrotoxicosis (P<0.05) (online Supplementary Table S4).
Between 2008 and 2015, the 1-year mortality after AF diagnosis improved but remained high, from 26 to 22 deaths per 100 people (temporal trends stratified by sex, ethnicity and age in Fig. 2). One-year mortality after AF diagnosis was higher in elderly adults but was on the downward trend from 31 to 26 per 100 patients, while remaining fairly constant between 11 and 14 per 100 middle-aged adults (online Supplementary Fig. S3). However, the majority of deaths were non-cardiovascular in both middle-aged and elderly adults (Tables 1 and 2).Among patients with AF and after adjusting for age, sex, ethnicity, housing type and clinical risk factors, sex differences in 1-year all-cause mortality were limited to elderly adults, with men having a 17% increased odds of mortality compared with women (Table 3). Ethnic differences in 1-year all-cause mortality were observed only between Indians and Chinese, with Indians associated with a 30% increased odds of mortality compared with Chinese in both age groups (Table 3). With respect to cardiovascular mortality, Indians were noted to have 36% and 54% increased odds of 1-year mortality in the middle-aged and elderly groups, respectively, compared with Chinese in the corresponding groups (Table 3). Age, diabetes, IHD, renal impairment, HF, stroke/transient ischaemic attack, haemorrhagic stroke, systemic embolism, peripheral vascular disease, depression and cancer were associated with all-cause mortality in both age groups, while end-stage renal failure and dementia were also associated with all-cause mortality among elderly adults (online Supplementary Table S5).
The presence of AF in patients conferred a substantially higher risk of mortality compared with matched patients without AF (28% vs 7%, P<0.001), with a 9-fold increased odds of mortality among middle-aged adults (adjusted odds ratio [AOR] 9.08, 95% CI 7.36–11.20) and a greater than 3-fold increased odds of mortality among elderly adults (AOR 3.60, 95% CI 3.40–3.80) within a year of diagnosis after adjusting for clinical risk factors (Table 4). Interactions in all-cause mortality were present only among elderly adults and were between age and sex (P for age and male = 0.023), age and ethnicity (P for age and Indian = 0.012), and sex and ethnicity (P for male and Malay <0.001; P for male and Indian = 0.015). Among middle-aged adults, AF affected mortality differently in the sex and ethnic categories. AF conferred a greater increase in mortality in women than in men; while Chinese also had a greater increase in mortality compared with Malays and Indians (Table 4). In the elderly population, the increased odds of mortality associated with AF were generally 2- to 3-fold higher across the sexes and ethnicities.
Table 4. Odds ratio of 1-year all-cause mortality in patients with atrial fibrillation compared with those without atrial fibrillation (reference) in subgroups of patients stratified by age group, sex and ethnicity
In this nationwide multiethnic cohort, we found significant sex and ethnic differences in the incidence, clinical profile and 1-year all-cause mortality among patients with newly diagnosed AF. Despite having the worst cardiovascular risk profile, Indians had the lowest incidence of AF. The development of AF conferred substantially higher risk of mortality within a year of diagnosis compared with no development of AF, with a mortality rate of up to 22–26 per 100 people. Mortality in newly diagnosed AF was influenced by sex and ethnicity, particularly among elderly men and Indians who were more susceptible.
Prevalence and incidence of AF
The prevalence of AF in Asia is approximately 1% but with significant geographical variations.2,5-7,9 We found an almost 10-fold higher prevalence of AF in elderly adults compared with middle-aged adults, a finding consistent with known age-related effects.2,8 Within the Asia-Pacific region, the prevalence of AF in Singapore mirrors that of Australia and New Zealand (1.4–4.9%) and is higher than those of other Southeast Asian countries (<2%).5,29-31 Based on national population data, we report a prevalence of 0.6% among middle-aged adults and 7% in elderly adults, and a steady temporal increase in the prevalence of AF. The higher rates of AF observed in Singapore are not unexpected and are likely due to an ageing population with increased life expectancy, a higher burden of cardiovascular disease and increasing detection of AF.28 Additionally, Singapore is a high-income country, and the adoption of lifestyles similar to those in affluent Western nations may partly account for AF rates closer to those of Western nations.5 Despite the steady incidence of AF, mortality rates were on the downward trend from 2010, possibly explained by rapid improvements in the standard of living, healthcare access and increasing emphasis on cardiovascular risk factor prevention and management, as well as the increasing longevity of population in Singapore.
Regardless of age, women consistently had lower prevalence and incidence of AF. Sex differences in baseline characteristics among elderly adults with AF in Singapore were consistent with previous reports10-18 and have been attributed to the protective effect of oestrogen on atrial electrophysiological properties, while body mass index had varying effects on atrial mechanical function by sex.16,33,34 In contrast, disparities in AF risk factors did not correspond to the observed rates of AF among the major ethnic groups in Singapore. Despite the higher prevalence of cardiovascular disease in Indians and Malays compared with Chinese regardless of age or sex, Indians had the lowest rates of AF while Malays had the highest, with higher rates of AF even among Malay women than among Indian men. Low rates of AF among Indians were previously reported within and outside Asia and were attributed to genetic variations. These variations might lead to atrial structural and electrophysiological properties that result in Indians being less susceptible to AF.24-26 Differences in AF rates between Malays and Chinese are likely a reflection of the higher cardiovascular risk profile among Malays, while differences in local dietary habits and activity levels may also contribute, with the effects of lifestyle modifications on maintenance of sinus rhythm being highlighted previously in a study.35 The higher prevalence of inpatient diagnoses of AF among Malays may also suggest a worse comorbid status. Despite shared environmental factors within the same country, ethnic differences in AF rates in this population are likely multifactorial, involving clinical risk factors, genetic susceptibility, lifestyle and socioeconomic status.9,27
One-year mortality after newly diagnosed AF
The 1-year mortality rate of 22–26 per 100 person-years is far higher than the previously reported rate of 18 per 1,000 patient-years in a meta-analysis,36 but is similar to that in the Framingham Heart Study, which reported a lower 1-year mortality rate of 11–15% among younger patients aged 55–64 years and a graded increase per decade in those aged above 65 years.37 The lower mortality previously reported may be due to survival bias, as those studies included patients with known AF who were likely to have survived past the first year of diagnosis compared with our cohort with newly diagnosed AF. Hence, those studies may have included more patients with fewer comorbidities who were generally in better health. For example, patients in our study when compared with those in the Framingham study had higher rates of diabetes (41.9% vs 11.3%), hypertension (80.2% vs 58.4%), myocardial infarction (47.1% vs 28.0%), HF (28.5% vs 21.8%) and stroke (25.2% vs 14.0%).37 Another reason for the high mortality may be that over two-thirds of our patients were first diagnosed with AF during hospital admissions and not during outpatient clinic encounters; hence, a higher-risk cohort was selected compared to previous studies, with AF possibly picked up at later stages of the disease. The majority of the initial presentations during hospital admissions were likely multifactorial, considering the lack of a systematic community screening programme for AF. Furthermore, most Singapore residents prefer to seek medical care at public hospitals, especially through emergency departments, in order to have easy access to highly subsidised acute care. Hence, AF is often only diagnosed when acutely unwell patients seek emergency care.
Notably, more than half of the deaths in our study were non-cardiovascular in nature, and the predictive effects of AF on mortality were independent of cardiovascular risk factors. AF may be more of a risk marker than a risk factor in this population, occurring in patients at high risk of mortality from other causes. This view is further supported by excess mortality in men within 30 days of AF diagnosis in the Framingham Heart Study, suggesting the role of AF as a marker of terminal illness.37 Our study extends current evidence by demonstrating ethnic-specific mortality risks among patients with shared environments. Despite having the lowest incidence of AF, Indians had a 30% increased 1-year mortality after diagnosis of AF regardless of age. Although the Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) study demonstrated higher adjusted 1-year mortality rates in India compared with Southeast Asia,38 the higher mortality rate in Indians compared with Chinese in this study was among patients who lived in the same country with less environmental differences. Given the genetic basis that potentially renders Indians less susceptible to AF,25-27 it is possible that the development of AF among Indians may indicate more advanced stages of systemic disease with an increased likelihood of early mortality. The significantly increased mortality within a year of AF diagnosis, especially among certain patient subgroups, suggests that AF should be considered a marker of poor prognostic outcomes and a harbinger of early mortality. Early aggressive management of AF and underlying comorbidities should be considered in these high-risk patient groups.
Strengths and limitations
The large study population and long study period provide statistically stable estimates and trends. However, we acknowledge the possibility of underestimation or overestimation of number of AF cases as cases included in this study were based on coded diagnoses submitted by healthcare institutions rather than direct inspection of 12-lead electrocardiograms. Furthermore, ethnic differences in perceptions of healthcare access and thresholds to seeking medical attention might potentially impact the incidence of newly diagnosed AF, although we think it is likely a small effect due to the lack of significant geographical and financial barriers to healthcare in Singapore. Newly diagnosed AF does not imply new-onset AF and may have been detected only late in the course of the disease. However, it is not possible to reliably ascertain the date of AF onset, with current guidelines recommending opportunistic screening for AF only in elderly patients ≥65 years of age.9 Data collection was made primarily for an administrative database on disease prevalence within Singapore, which precluded in-depth analyses on treatment efficacy and differences in cardiac structure and function. Specific dates of death were unavailable, and only a small proportion of patients would have a lag in death certification and these lags are not expected to affect the time to death significantly. The retrospective nature of this study allowed only inferences to be drawn about the association between AF and death, and may be affected by unknown confounders despite our comprehensive adjustment of clinical covariates in the multivariable models.
Among patients from a multiethnic population, variations in incidence, clinical profile and 1-year mortality of patients with AF were influenced by age, sex and ethnicity. Newly diagnosed AF portends a worse prognosis and is a marker of high mortality within the first year.
The contributions of all staff at the Singapore Ministry of Health and Department of Statistics for this study are duly acknowledged.
- Staerk L, Sherer JA, Ko D, et al. Atrial fibrillation: Epidemiology, pathophysiology, and clinical outcomes. Circ Res 2017;120:1501-17.
- Chugh SS, Havmoeller R, Narayanan K, et al. Worldwide epidemiology of atrial fibrillation: A Global Burden of Disease 2010 Study. Circulation 2014;129:837-47.
- Bai Y, Wang YL, Shantsila A, et al. The global burden of atrial fibrillation and stroke: A systematic review of the clinical epidemiology of atrial fibrillation in Asia. Chest 2017;152:810-20.
- Guo Y, Tian Y, Wang H, et al. Prevalence, incidence, and lifetime risk of atrial fibrillation in China: New insights into the global burden of atrial fibrillation. Chest 2015;147:109-19.
- Wong CX, Brown A, Tse HF, et al. Epidemiology of atrial fibrillation: The Australian and Asia-Pacific perspective. Heart Lung Circ 2017;26:870-9.
- Kodani E, Atarashi H. Prevalence of atrial fibrillation in Asia and the world. J Arrhythmia 2012;28:330-7.
- Lip GYH, Brechin CM, Lane DA. The global burden of atrial fibrillation and stroke: A systematic review of the epidemiology of atrial fibrillation in regions outside North America and Europe. Chest 2012;142:1489-98.
- Tan ESJ, Lee CH. Obstructive sleep apnea and arrhythmias in the elderly. Curr Sleep Medicine Rep 2021;7:197-205.
- Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J 2021;42:373-498.
- Magnussen C, Niiranen TJ, Ojeda FM, et al. Sex differences and similarities in atrial fibrillation epidemiology, risk factors, and mortality in community cohorts: Results from the BiomarCaRE Consortium (Biomarker for Cardiovascular Risk Assessment in Europe). Circulation 2017;136:1588-97.
- Bose A, O’Neal WT, Wu C, et al. 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;123:1453-7.
- Dagres N, Nieuwlaat R, Vardas PE, et al. Gender-related differences in presentation, treatment, and outcome of patients with atrial fibrillation in Europe: A report from the Euro Heart Survey on Atrial Fibrillation. J Am Coll Cardiol 2007;49:572-7.
- Daneshvar M, Desai N, Andriulli J, et al. Gender differences in presentation, treatment, and in-hospital outcome of patients admitted with heart failure complicated by atrial fibrillation (from the Get With the Guidelines-Heart Failure [GWTG-HF] Registry). Am J Cardiol 2018;121:450-4.
- Schnabel RB, Pecen L, Ojeda FM, et al. Gender differences in clinical presentation and 1-year outcomes in atrial fibrillation. Heart 2017;103:1024-30.
- Westerman S, Wenger N. Gender differences in atrial fibrillation: A review of epidemiology, management, and outcomes. Curr Cardiol Rev 2019;15:136-44.
- Gillis AM. Atrial fibrillation and ventricular arrhythmias: Sex differences in electrophysiology, epidemiology, clinical presentation, and clinical outcomes. Circulation 2017;135:593-608.
- Magnussen C, Ojeda FM, Wild PS, et al. Atrial fibrillation manifestations risk factors and sex differences in a population-based Cohort (from the Gutenberg Health Study). Am J Cardiol 2018;122:76-82.
- Xiong Q, Shantsila A, Lane DA, et al. Sex differences in clinical characteristics and inpatient outcomes among 2442 hospitalized Chinese patients with nonvalvular atrial fibrillation: The Nanchang Atrial Fibrillation Project. Int J Cardiol 2015;201:195-9.
- Rodriguez F, Stefanick ML, Greenland P, et al. Racial and ethnic differences in atrial fibrillation risk factors and predictors in women: Findings from the Women’s Health Initiative. Am Heart J 2016;176:70-7.
- Golwala H, Jackson LR, 2nd, Simon DN, et al. Racial/ethnic differences in atrial fibrillation symptoms, treatment patterns, and outcomes: Insights from Outcomes Registry for Better Informed Treatment for Atrial Fibrillation Registry. Am Heart J 2016;174:29-36.
- Ugowe FE, Jackson LR 2nd, Thomas KL. Racial and ethnic differences in the prevalence, management, and outcomes in patients with atrial fibrillation: A systematic review. Heart Rhythm 2018;15:1337-45.
- Magnani JW, Norby FL, Agarwal SK, et al. Racial differences in atrial fibrillation-related cardiovascular disease and mortality: The Atherosclerosis Risk in Communities (ARIC) Study. JAMA Cardiol 2016;1:433-41.
- Tan ESJ, Tay WT, Teng TK, et al. Ethnic differences in atrial fibrillation in patients with heart failure from Asia-Pacific. Heart 2019;105:842-7.
- Tan ESJ, Goh V, Santema BT, et al. Ethnic differences in atrial fibrillation among patients with heart failure in Asia. ESC Heart Fail 2020;7:1419-29.
- Gillott RG, Willan K, Kain K, et al. South Asian ethnicity is associated with a lower prevalence of atrial fibrillation despite greater prevalence of established risk factors: A population-based study in Bradford Metropolitan District. Europace 2017;19:356-63.
- O’Neill J, Tayebjee MH. Why are south Asians seemingly protected against the development of atrial fibrillation? A review of current evidence. Trends Cardiovasc Med 2017;27:249-57.
- Chang SH, Kuo CF, Chou IJ, et al. Association of a family history of atrial fibrillation with incidence and outcomes of atrial fibrillation: A population-based family cohort study. JAMA Cardiol 2017;2:863-70.
- Lam CS, Teng TK, Tay WT, et al. Regional and ethnic differences among patients with heart failure in Asia: the Asian sudden cardiac death in heart failure registry. Eur Heart J 2016;37:3141-53.
- Yap KB, Ng TP, Ong HY. Low prevalence of atrial fibrillation in community-dwelling Chinese aged 55 years or older in Singapore: a population-based study. J Electrocardiol 2008;41:94-8.
- Lim CW, Kasim S, Ismail JR, et al. Prevalence of atrial fibrillation in the Malaysian communities. Heart Asia 2016;8:62-6.
- Kiatchoosakun S, Pachirat O, Chirawatkul A, et al. Prevalence of cardiac arrhythmias in Thai community. J Med Assoc Thai 1999;82:727-33.
- Lip GY, Nieuwlaat R, Pisters R, et al. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: The Euro Heart Survey on Atrial Fibrillation. Chest 2010;137:263-72.
- Tse HF, Oral H, Pelosi F, et al. Effect of gender on atrial electrophysiologic changes induced by rapid atrial pacing and elevation of atrial pressure. J Cardiovasc Electrophysiol 2001;12:986-9.
- Doria de Vasconcellos H, Betoko A, Ciuffo LA, et al. Sex differences in the association of cumulative body mass index from early adulthood to middle age and left atrial remodeling evaluated by three-dimensional echocardiography: The coronary artery risk development in young adults study. J Am Soc Echocardiogr 2020;33:878-87.e3.
- Rienstra M, Hobbelt AH, Alings M, et al. Targeted therapy of underlying conditions improves sinus rhythm maintenance in patients with persistent atrial fibrillation: results of the RACE 3 trial. Eur Heart J 2018;39:2987-96.
- Emdin CA, Wong CX, Hsiao AJ, et al. Atrial fibrillation as risk factor for cardiovascular disease and death in women compared with men: Systematic review and meta-analysis of cohort studies. BMJ 2016;532:h7013.
- Benjamin EJ, Wolf PA, D’Agostino RB, et al. Impact of atrial fibrillation on the risk of death: The Framingham Heart Study. Circulation 1998;98:946-52.
- Healey JS, Oldgren J, Ezekowitz M, et al. Occurrence of death and stroke in patients in 47 countries 1 year after presenting with atrial fibrillation: a cohort study. Lancet 2016;388:1161-9.