Atrial fibrillation in Middle Eastern Arabs and South Asians: a scoping review

Eastern Arabs. MEDLINE, Embase and CENTRAL databaseswereincludedinoursearch. Afterscreening8995records, 55 studies were selected; 42 from the Middle East and 13 from South Asia. Characteristics of the included studies were tabulated, and theirdataweresummarizedforstudydesign,setting,enrolmentpe-riod, sample size, demographics, prevalence or incidence of AF, co-morbidities,riskfactors,AFtypesandsymptoms,management,out- comes, and risk determinants. Identified literature gaps included a paucity of community or population-based studies that are representative of these two ethnicities/races. In addition, studies that addressed ethnic/racial in-equality and access to treatment were lacking. Our study underscores the urgent need to study cardiovascular disorders, particularly AF, in South Asians and Middle Eastern Arabs aswellasinotherlessrepresentedethnicitiesandraces.


Introduction
There has been an increasing interest in the impact of ethnicity and race on cardiovascular disorders. Patients from the same race and/or ethnicity share common genetics as well as environmental exposures. This may be responsible for the differences observed between different races/ethnicities in the incidence, presentation and outcomes of cardiovascular disorders as well as for the differences in response to therapy [1]. In addition, racial and ethnic inequalities and bias may also play an important role. Variations in access to care, economic status and education have all been linked to race/ethnicity and can affect awareness of the disease, which may lead to underdiagnosis and insufficient treatment.
Atrial fibrillation (AF) is the most common arrhythmia encountered in clinical practice, accounting for approximately one-third of hospitalizations for cardiac rhythm disturbances [2]. Most of the published literature on AF originates from North America and Europe involving mainly Caucasian patients [3]. Other ethnicities are largely underrepresented particularly South Asians and Middle Eastern Arabs. Therefore, our aim was to perform a scoping review to identify and summarize the current evidence on AF in these two ethnicities with the overarching objective to identify research gaps and inform future research in this area.

Methods
A Scoping review is a systematic method of knowledge synthesis that examines and maps evidence on a specific subject matter, thus, identifying key concepts, theories, sources of evidence, and research [4]. This scoping review follows the five steps proposed by Arskey and O'Malley [5]: (1) identifying the research question, (2) identifying relevant studies, (3) selecting studies, (4) collecting data, and (5) mapping, summarizing and describing the results.

Eligibility criteria
Studies that reported various aspects of AF including epidemiology, co-morbidities, risk factors, management and/or clinical outcomes in the Middle East or South Asia were included. Exclusion criteria included studies on sub-clinical AF, post-operative AF, genetics, cost-effectiveness, and smart devices. The countries in this review were categorized according to the world regions as per the World Health Organization (WHO), i.e., South-East Asia and Eastern Mediterranean regions [6]. For the purpose of this review, the terms Middle East(ern) and South Asia(n) will be used.

Search strategy
A comprehensive electronic literature search using MED-LINE, EMBASE, and CENTRAL was performed on September 26, 2020. Boolean terms ("OR" and "AND"), Medical Subject Headings (MeSH), Emtree and broad keywords were used. The search terms were combined and included: "atrial fibrillation", "Middle East", "South Asia", "Arab", individual country and nationality. Search limitations were not applied. Reviews and meta-analyses, international registries and the references' lists of the retrieved articles were manually screened to identify additional relevant studies. The detailed search strategy is described in Supplementary Table 1.

Incidence and prevalence of AF
The incidence and prevalence of AF was reported in association with other diseases. The overall incidence of AF during hospitalization was 1.8% in patients admitted with acute myocardial infarction (AMI) in Qatar [48], with the incidence in Arabs being higher than that in South Asians (3.1% vs. 0.8%). From the earlier data of the same Qatari registry (1991 to 2002) [45], the overall annual incidence among Qataris was higher than that of other nationalities (12% vs. 8.0%). However, the rate (8.0-9.3%) remained consistent over years for the overall patient population, i.e., those hospitalized with acute cardiac illness [30,46,47,50]. The incidence of AF in the Gulf RACE-2 registry was 2.7% among patients hospitalized with ACS [27]. Two studies from Pak- istan [15,18] reported similar frequency (9.1%) in patients hospitalized with AMI. The prevalence of AF among general hospital admissions was 3.4% in Bahrain [13], 4.2% in Kuwait [19], and 5.8% in Pakistan [24]. Whereas, a lower prevalence (2.8%) was reported in the only study that was conducted in a primary health care center in Iran [14]. The prevalence of AF was 7.8% among patients on maintenance HD in Jordan [22], and 10.6% in patients with chronic heart failure in Morocco [29]. In Iran, the prevalence of AF among patients with stroke was 11.1% [20], while in Qatar, it was 6.4%. In the latter Qatari study, 2.6% of stroke patients were newly diagnosed with AF [28]. The highest AF prevalence reported was 17.8% in patients hospitalized with heart failure [26], and 23.9% in patients with RHD [39] (Table 2 and Supplementary Table 3).

Co-morbidities and risk factors
The most common co-morbidities associated with AF were hypertension, diabetes mellitus, heart failure and stroke which were comparable between the Middle Eastern and the South Asian registries (hypertension 52.0-63.0% and 31.4-68.5%, diabetes mellitus 30.0-48.0% and 16.1-36.2%, heart failure 27.0-31.7% and 15.5-27.2%, stroke 9.0-13.0% and 9.1-17.8%, respectively). Analyses from the Gulf SAFE registry [33] showed similar frequencies of the first three afore-mentioned co-morbidities in patients with stroke [57], and similar frequencies of hypertension and heart failure in patients with diabetes mellitus [54]. In non-AF registries (stroke and heart failure), AF patients with stroke [28] and acute heart failure [26] had higher frequencies of both hypertension (72.3% and 70.8%) and diabetes (67.7%, including pre-diabetics, and 55.9%), than that for AF patients with chronic heart failure (55.0% and 65.9%) [29] and ACS (39.0% and 44.2%) [27], respectively. Valvular heart diseases were more prominent in South Asia (40.7-50.3% versus 8.0-24.0%) than in the Middle East. In the non-registry studies, there were noticeable variations in the reported frequencies of co-morbidities accompanied with variations in the terms of definitions used for the associated conditions. The mean HAS-BLED score ranged between 1.1 and 1.9 across the included studies, while mean CH 2 ADS 2 -VASc score between 2.3 and 4.1, and CHADS 2 score between 1.6 and 2.3, with the majority of patients having high stroke risk, i.e., CHADS 2 ≥2 in 36.7 to 71.0% of patients (Table 3, Ref [12, 15, 20, 22, 33-38, 42-44, 55] and Supplementary Table 4).

Types and symptoms of AF
There were few studies that reported the types and symptoms of AF. The frequency of new onset AF was reported in two Middle Eastern studies as 27.0% [24] and 59.0% [13], which was higher than that reported in the AF registries  [32,33,41,42,44]. The proportion of patients with permanent AF was higher in AF registry studies as compared with that in non-AF registry studies. The proportion of patients with persistent AF was higher in the latter studies. The most common symptoms at presentation, i.e., palpitation and dyspnea, were reported in more than half of the patients across the studies (Table 4, Ref [12,13,15,20,22,24,32,34,36,38,42], and Supplementary Table 5). A single-centre study from Nepal [35] reported AF types and symptoms in patients with valvular and non-valvular AF. Paroxysmal AF was more common in non-valvular AF (55.2% versus 7.8%), while permanent AF was more common in valvular AF (51.0% versus 10.4%). Symptoms did not differ between the two groups (Supplementary Table 5).

Outcomes in AF
The main adverse outcomes reported by the studies included mortality (in-hospital, one-year), cerebrovascular events, and bleeding. In-hospital mortality was reported in 0.79% of patients in Bahrain [13] and was doubled in Thailand (1.6%) [37]. The rate was 3.6% in Qatar among patients admitted with general cardiac conditions [30], with a rate of 4.8% in Middle Eastern and 1.2% in South Asian patients as defined in the study [47]. The rates were higher in  patients with acute heart failure (6.7%) [26], chronic kidney disease (CKD) (11.7%) [51], ACS (14.7%) [27], and acute ischemic stroke (9.0% [20] and 26.8% [36]). One-year mortality rate ranged from 6.5% to 18.1% in four South Asian studies [37,40,43,44]. Gulf SAFE registry reported a rate of 15.0% [58] with a significantly lower mortality rate in patients who were compliant with the AF Better Care (ABC) pathway (7.3% versus 13.1%, p = 0.033) [55]. In the same registry, diabetic patients had significantly higher one-year mortality rate than non-diabetics (14.4% versus 9.6%, p = 0.003) [53]. Similarly, diabetic patients who were compliant with the ABC pathway had lower mortality rate as well (5.8% versus 15.9, p = 0.0014) [54]. Higher one-year mortality rates (>20%) were reported in AF patients presenting with ACS [27] or acute heart failure [26]. The observed rate of cerebrovascular accident (CVA) and/or transient ischemic attach (TIA) in the South Asian AF registries ranged from 0.85% to 3.0% [43,44,60], while a single-centre study found much higher rates; 17.2% in non-valvular and 23.1% in valvular AF patients [35]. The Gulf SAFE reported a rate of 6.4% [58], while the Qatari registry reported a lower rate of 1.1% in patients admitted with cardiac condition [46]. Additionally, the rate was higher in Middle Eastern (1.4%) than South Asian (0.6%) population in the Qatari registry [47]. The rates of CVA/TIA in special AF patient populations were 2.3%, 2.0%, 1.8-2.8%, and 6.7% in CKD [51], acute heart failure, [26] AMI/ACS [27,48], and RHD [39], respectively. Two single-center studies from Iran reported higher rates in patients presenting with AMI (12.5% [18] and 13.6% [15]). Finally, the rate of major bleeding in studies from the Middle East [26,27,58] ranged from 1.1% to 1.7% including those in patients presenting with ACS [27] and acute heart failure [26]. However, one-single center study from KSA reported a rate of 35.6% in those with high risk of bleeding [7]. Studies from South Asia reported rates ranged from 0.31% to 4.4% [35,43,44,60]. A study from UAE found that the use of inappropriate NOACs doses subjected the patients to higher ischemic stroke and major bleeding rates compared to those on appropriate doses (9.9% versus 3.1%, p < 0.01 and 11.7% versus 6.0%, p = 0.04; respectively) [8] (Supplementary Table 7).

Discussion
The objective of this scoping review was to assess the extent of the available literature of AF in South Asians and Middle Eastern Arabs, then identify and summarize various related aspects such as clinical features and management. A preliminary search of MEDLINE and the Cochrane Database of Systematic Reviews was conducted and no current or underway systematic reviews or scoping reviews on the topic were identified. The reason we chose a scoping review over a systematic review for this study was the broad research question. While a rigorous and detailed in-depth analysis of data is not normally required in scoping reviews, a systematic inquiry was planned and conducted with the synthesis of available evidence into several themes to comply with the study objectives. The study identified significant research gaps in the reporting of this important arrhythmia in the two prespecified ethnicities/races.
Previous reports have suggested a lower incidence and prevalence of AF in the individuals of non-European ancestry. Data from the United Kingdom-based West Birmingham Atrial Fibrillation Project [62,63] suggested a low prevalence of AF (0.6%), in Indo-Asians aged >50 years, compared with 2.4% in the general study population. The incidence of AF in individuals of European ancestry was higher than that in individuals of African, Chinese, or Japanese backgrounds as suggested by analysis of data from the multinational ASSERT trial [64]. We observed considerable variations in the epidemiological data reported in the included studies, with incidence/prevalence rates ranging from 2.8 to 17.8% in the Middle Eastern region and 3.4 to 23.9% in South Asian region ( Table 2). In addition, in our study, data of AF in these two ethnicities were reported in association with other diseases such as heart failure, ACS and stroke. Furthermore, most of the available data originates from hospital-based studies which may under-or over-estimate the true prevalence of the disease in the community. Moreover, the reported studies focused on data from the urban populations, with underpresentation of patients living in remote or rural areas. Such patient populations could have different risk factors for AF and limited access to healthcare compared with their counterparts from the urban populations. Data on the accurate incidence and prevalence of AF in South Asians and Arabs, in longitudinal epidemiological studies that are representative of the two ethnicities are therefore needed.
In our study the age range of AF patients was 58.8-68.4 years in the Middle Eastern region, and 51.2-64.6 years in the South Asian region. Compared with our findings, most of AF patients (>70%) of European ancestry were older, aged more than 65 years [65]. Moreover, the leading prevalent AF-associated co-morbidities in our study included hypertension, diabetes mellitus, heart failure, valvular heart diseases and stroke. This was consistent with studies in other ethnicities from around the world [65]. In one study examining secular trends between the two ethnicities with AF over 20 years (1991-2010), Salam et al. [47] reported significant variations in the prevalence of underlying cardiac aetiologies in both ethnicities. There was an increase in the prevalence of ischemic heart disease and a decrease in that of valvular heart disease in the latter years of the study. In addition, there was an increase in the prevalence of diabetes mellitus and hypertension in both groups. Consistent with our findings, valvular heart disease used to be a major cause of AF in the Western world [66], but because of the wide availability of early treatments for rheumatic fever and streptococcal infection these diseases are now rare in the developed world.

International registries
Middle Eastern Arabs and South Asians were included as part of five AF international registries [67][68][69][70][71] (Table 6). The age of patients (i.e., range 57.9-70.0 years) and female distribution (i.e., range 41.0-57.6%) were in the ranges reported in the registries of this review. The South Asian registries reported wider variabilities in the rate of heart failure, hypertension and diabetes, but not stroke, compared with those in the Middle Eastern ones. The rates of the aforementioned co-morbidities were similar to those in the international registries except for stroke in the RE-LY AF registry, which reported a higher rate in the Middle East. RHD was the highest in India (31.5%) [67]. Similarly, there was a wide range in the rates of AF types in the South Asian registries of this study. The rates of paroxysmal and permanent AF were higher in the international registries. It is important to mention that the regions and distribution of countries in the international registries were different from the WHO regions which were used in our study. The RE-LY AF reg-istry [67] concluded with large regional variations in age, comorbidities, risk factors, and management. Finally, substantial regional differences in stroke prevention have been found in the GLORIA-AF registry during its phase II, with an oral anticoagulation use rate of 87.4% in Africa/Middle East [63], which was higher than the rate reported in this review (i.e., range 38.5-51.9%).

Literature gaps and future directions
Our review identified several gaps in AF literature related to Middle Eastern Arabs and South Asians. Importantly, most of the available evidence was based on studies in the hospital setting with a paucity of informative communitybased or population-based studies representative of these two ethnicities/races. This is particularly important in correctly identifying the AF disease prevalence and burden. Socioeconomic difficulties are an important barrier for the conduct of these types of studies but can be overcome by joining internationally funded studies with samples of these and other ethnicities representative of the population. Furthermore, studies that address ethnic/racial in-equality and access to treatment, particularly oral anticoagulation, and the impact on outcomes, are lacking and should be baldly addressed in populations with mixed ethnicities and races including Middle Eastern Arabs and South Asians. In addition, how ethnicity and race affect AF presentation, symptomatology and interpretation needs to be examined. The Flow-AF registry is currently ongoing to recruit 1446 patients with a new non-valvular AF diagnosis across four Middle Eastern countries (Egypt, Lebanon, KSA, UAE). The registry will examine patients' characteristics, patterns of treatment, clinical outcomes, and the economic aspects of treatment [72].

Limitations
Although our study was conducted using rigorous and systematic search strategy, it is possible that some articles were not identified by our search. The unidentified papers could contribute to a potential bias in the presented results with unknown impact. Moreover, the gap in the time between the literature search and completing results synthesis may have contributed to missing recently published relevant papers. However, we believe that this limitation may be balanced by the meticulousness of our methodology and the included literature.

Conclusions
Although in this scoping study we identified and summarized studies addressing clinical features and management of AF in Middle Eastern Arabs and South Asians, the main body of evidence comes from hospital-based studies. Thus, considered a limited and incomplete evidence. There is, additionally, an evidence gap of knowledge of AF burden that should be addressed in representative community or populationbased studies. Our study highlights the need to further examine the effects of ethnicity and race in heart diseases particularly in AF.

Author contributions
Authors AS & RK-Conceptualization, literature search, study selection, data extraction, preparing figures/tables and manuscript writing. Authors AS and SA made substantial contributions to the conception or design of the work; analysis and interpretation of data for the work. Author VS critically revised the paper for important intellectual content and final editing of the version to be published. All authors read and approved the final manuscript.

Ethics approval and consent to participate
Not applicable.