Atrial fibrillation prevalence and predictors in patients with diabetes: a cross-sectional screening study

Br J Cardiol 2022;29:21–5doi:10.5837/bjc.2022.008 Leave a comment
Click any image to enlarge
Authors:
First published online March 8th 2022

Prevalence of atrial fibrillation (AF) and diabetes is increasing worldwide. Diabetes is a risk factor for AF and both increase stroke risk. Previous AF screening studies have recruited high-risk patient groups, but not with diabetes as the target group. This study aims to determine whether people with diabetes have a higher prevalence of AF than the general population and investigate whether determinants, such as diabetes duration or diabetes control, add to AF risk.

In a cross-sectional screening study, patients with diabetes were recruited via their GP surgeries or a diabetes centre. A 30-second single-lead electrocardiogram (ECG) was recorded using the Kardia® device, along with physiological measurements and details relating to risk factor variables.

There were 300 participants recruited and 16 patients identified with AF (5.3% prevalence). This demonstrated a significantly greater likelihood of AF than the background population (p=0.043). People with diabetes and AF were significantly older than those who only had diabetes. More people with type 2 diabetes had AF than people with type 1. Prediction of AF diagnosis by age, sex, diabetes type, diabetes duration and level of control revealed only age as a significant predictor.

In conclusion, these findings add to existing data around the association of these chronic conditions, supporting AF screening in this high-risk group, particularly in those of older age. This can contribute to appropriate management of both conditions in combination, not least with regards to stroke prevention.

Advertisement
Angina - BJC Learning programme
For healthcare professionals only

Introduction

Prevalence of atrial fibrillation (AF) and diabetes is increasing worldwide.1,2 AF is a common heart rhythm irregularity and prevalence increases with age. People with AF are up to seven times more likely to have a stroke than the general population,1 with risk increased further in the presence of diabetes.3 AF may exhibit no symptoms and go undiagnosed until patients present with sequelae, such as stroke or heart failure. Stroke secondary to AF is often avoidable with thromboprophylaxis and early identification could lead to stroke prevention.

Diabetes mellitus is a major risk factor for cardiovascular disease,4 and frequency of AF is reported to be 1.4- to 2.1-fold higher in people with diabetes than without.5 The association between AF and diabetes may be through pathological determinants at cellular and molecular level,6 or through shared risk factors.5

Age-adjusted AF incidence and prevalence is larger among men, but women are older at the time of AF diagnosis.7 Men with AF have a larger burden of coronary artery disease, but women tend to have a higher prevalence of heart failure and valvular heart disease.7 Being male is also associated with increased likelihood of developing diabetes,8 and, therefore, effects of sex on the presence of AF in the person with diabetes will be explored.

Finally, the health professional managing the patient’s diabetes (GP or diabetes specialist) may not always relate to diabetes status. GP centres may attract the ‘worried well’ and those who voluntarily seek healthcare. A specialist may see the patient for other reasons, e.g. dietary advice or podiatry. This study will consider whether there is a difference in AF prevalence between screening locations and any impact this could have.

Previous AF screening studies recruited high-risk patient groups, but not with diabetes as the lone, target group.9-11 The aim of this study is to screen people with diabetes for AF using the Kardia® device (Mountainview, California, USA), to determine whether people with diabetes have a higher prevalence of AF than the general population. We also explore whether other determinants, such as age, sex, diabetes stability and duration, impact AF prevalence in this patient group, or whether the existence of diabetes alone determines higher AF risk.

The Kardia® single-lead screening device

The Kardia® device, a smartphone-based heart monitor, is activated by placing fingers on pocket-sized metal electrodes, producing a single-lead electrocardiogram (ECG). The device is a clinically validated screening tool, CE (Conformité Européene) marked and FDA (Food and Drug Administration) cleared.12 Validity is high with sensitivities >98% and specificities >90%.13,14 A recent systematic review found it to be a convenient and feasible means of monitoring for AF, easily implemented into opportunistic and systematic screening.14 The Kardia® device is an event-type monitor recommended for use in England when episodes are more than 24 hours apart.15,16

Study hypotheses

This study aims to discover the prevalence of AF in a diabetic population and whether screening in this group is effective. The hypotheses are:

  • Presence of diabetes will be a predictor of the presence of AF, controlling for age and sex.
  • There will be a difference between male and female patients in the frequency with which AF is identified.
  • Screening patients with diabetes will detect a higher prevalence of AF than screening in the general population.
  • Duration of diabetes and level of glycaemic control will be more important predictors of presence of AF than diagnosis of diabetes alone, or age and sex.
  • There will be a difference between the proportion/percentage of people detected as having AF in the two screening locations.

If the null versions of the hypotheses are supported by the data, e.g. presence of diabetes is not a predictor of the presence of AF, then this would indicate that targeted screening is not indicated.

Method

Design and setting

This cross-sectional screening study screened people with diabetes for AF using the Kardia® device. The research was conducted in Jersey, Channel Islands (population approximately 100,000) in an outpatient diabetes centre and a central community clinic where GP patients attended.

Participants

Sample size

Sample size calculation estimated 351 participants to be representative of the island’s population with diabetes (local figures suggest 4,000 people).17 This was based on a confidence interval of 5%, a confidence level of 95% and response distribution of 50%. Data collection was completed over 14 months, with 300 participants recruited. Further recruitment was affected by delays incurred by the COVID-19 pandemic, and low response rates from the initial GP practice, resulting in two further GP practices being recruited.

Eligibility

Table 1. Eligibility criteria

Inclusion criteria
Existing diagnosis of diabetes
≥18 years of age
Capacity to consent
Able to understand English
Able to get to the screening location
Does not have a pacemaker or internal cardiac defibrillator
Exclusion criteria
Does not have diabetes
<18 years of age
Lack capacity to consent
Unable to understand English
Unable to get to the screening location
Has a pacemaker or internal cardiac defibrillator

Eligibility criteria are set out in table 1. Patients were still included if they disclosed existing AF, as this was relevant when calculating prevalence. Patients were excluded if they had a cardiac pacing or defibrillation device due to the short pulse widths transmitted being difficult to detect with ambulatory ECG machines, including the Kardia® device.12

Screening procedure

Patients attending the diabetes centre were invited to participate while waiting for their scheduled appointment. Participating GP surgeries sent invitations to patients on their diabetes database and invited them to call the lead researcher to arrange a screening appointment.

Patients were provided with an information sheet and had the opportunity to ask questions, their right to withdraw was explained and they were informed that declining participation would not affect treatment. If they agreed, patients signed a consent form.

The 30-second Kardia® ECG was saved to the user’s phone then transferred to the lead researcher’s phone for saving in an encrypted folder accessible through two passwords, then deleted. ECGs were documented as ‘normal’, ‘AF’ or ‘unclassified’. All Kardia® ECGs were reviewed by the lead researcher and when ‘unclassified’, a 12-lead ECG performed. If this remained unclear, the Consultant Cardiologist was consulted. In the presence of AF, the patient was given an AF information sheet, a letter for their GP and a copy of the ECG.

Statistical analyses

One-way ANOVA (analysis of variance) were used to determine whether there were statistically significant differences between groups (AF or sinus [normal] rhythm), for the dependent variables age, blood pressure (BP), body mass index (BMI), heart rate, diabetes duration and level of control (glycosylated haemoglobin – HbA1c). Group comparisons for categorical variables (diabetes type, sex) were analysed using the Chi-square test of independence. To determine whether the presence of diabetes predicts AF, logistic regression was applied, controlling for age and sex. To test whether screening patients with diabetes will detect a higher prevalence of AF than screening in the general population, a t-test was used to compare the percentage of patients with diabetes between this study and the general population (previous AF screening studies). To examine whether duration of diabetes and glycaemic control are important predictors of presence of AF (in addition to the variance predicted by diabetes alone, or age and gender), these variables were incorporated into the logistic regression analysis. Finally, the Chi-square test of independence was applied to examine differences in the proportion of people detected as having AF in the two screening locations.

Results

ECG screenings

Single-lead ECGs were recorded for the 300 participants (diabetes centre n=156, GP practices n=144; one recording was not saved so this information is unavailable). The majority (diabetes centre n=150, GP practices n=134) demonstrated normal rhythm via automated analysis, and 16 (diabetes centre n=6, GP practices n=10) showed AF, both confirmed on manual review. The rhythm could not be accurately analysed using the incorporated algorithm in seven cases (diabetes centre n=3, GP practices n=4) and a 12-lead ECG was recorded on these seven ‘unclassified’ cases. Of these, six were diagnosed as normal following manual review and just one remained unclear, requiring further adjudication by the cardiologist.

Descriptive statistics and classifications are presented in table 2. Interval data are presented as mean ± standard deviation and categorical data as numbers and percentages. Average age of participants was 63 years (± 13) and the majority were male (n=169). One-way between-subjects ANOVA showed a statistically significant age difference between groups (F[1,298]=8.928, p=0.003), with the AF group being older than the diabetes only group, and a difference in heart rate (F[1,298]=12.035, p=0.001) such that the diabetes-only group (sinus rhythm [SR] in table 2) had a lower heart rate than the AF group. There was a statistically significant difference in AF detection between diabetes types (X2=4.696, p=0.030) with more people having AF with type 2 diabetes.

Table 2. Comparison of demographic data

Characteristic Total
N=300
AF
N=16
SR†
N=283
F p value
Gender, n (%) 0.125
Male 169 (56) 12 (75) 157 (55)
Female 131 (43.5) 4 (25) 127 (45)
Measurements, mean ± SD
Age, years 63 ± 13 72.4 ± 7.7 62.5 ± 13 8.928 0.003*
Weight, kg 87.8 ± 19.9 90.5 ± 23.8 87.7 ± 19.7 0.310 0.578
Height, cm 169.3 ± 9.6 170 ± 8.5 169.3 ± 9.7 0.347 0.556
BMI, kg/m2 30.5 ± 6.1 30.8 ± 6.4 30.5 ± 6.1 0.039 0.843
SBP, mmHg 131.7 ± 16.8 130.4 ± 18 131.8 ± 16.8 0.109 0.742
DBP, mmHg 70.9 ± 10.6 74.3 ± 14.3 70.7 ± 10.4 1.718 0.191
HR, mmHg 70.9 ± 13.6 90.7 ± 22 78.7 ± 12.8 12.035 0.001*
Diabetes type, n (%) – 1 missing 0.030*
Type 1 65 (21.6) 0 65 (100)
Type 2 234 (78.2) 16 (100) 218 (77)
Diabetes measures, mean ± SD
Diabetes duration, years 13.2 ± 11.1 12.9 ± 7.6 13.2 ± 11.3 0.012 0.911
HbA1c 7.62 ± 1.3 7.2 ± 1.1 7.6 ± 1.3 1.473 0.226
Other risk factors, n (%)
Hypertension 178 (59) 10 (62.5) 122 (43.1) 0.442
Smoker 25 (8.3) 0 25 (8.8) 0.216
Heart failure 18 (6) 1 (6.25) 17 (6) 0.968
Obesity 147 (49) 8 (66.6) 153 (54) 0.677
TIA/CVA 24 (8) 0 24 (8.4) 0.789
PVD 22 (7) 0 22 (7.7) 0.248
†Rhythm missing on 1 participant.
*p≤0.05 denotes a significant difference.
Key: AF = atrial fibrillation; BMI = body mass index; DBP = diastolic blood pressure; F = F statistic; HbA1c = glycosylated haemoglobin; HR = heart rate; PVD = peripheral vascular disease; SBP = systolic blood pressure; SD = standard deviation; SR = sinus rhythm; TIA/CVA = transient ischaemic attack/cerebrovascular accident

Diabetes as a predictor for AF

A binary logistic regression analysis to investigate whether the presence of diabetes predicted AF, controlling for age and sex, was conducted. Block 1 contained the heart rhythm classification (dependent variable AF encoded to 1 and SR to 0). Block 2 contained age and sex as predictors. Analyses were undertaken on 274 participants (27 contained missing data, leaving 91% of the sample for analysis). The model was statistically significant (X²=12.58, p=0.013) with explained variation in AF presence being 14.1% (Nagelkerke R²=0.141). Results demonstrate that age is the strongest predictor for AF.

Difference between men and women in terms of frequency with which AF is identified

AF was identified in 12 men (7.1%) and four women (3.1%). The Chi-square test of independence showed no effect of sex on the frequency with which AF is observed AF (X²[2,299]=3.641, p=0.162).

Screening patients with diabetes will detect a higher prevalence of AF than does screening in the general population

AF prevalence was 5.3% in this study (n=16/299). Prevalence is expected to increase 2.3-fold between 2016 and 2060,1 with an observed 0.5% rise seen over five years in England (2% in 2014 to 2.5% in 2019).16,18 Therefore, 2.7% was selected as a test value to compare with our data for this diabetes sample. The diabetes population in this study did show a significantly greater likelihood of AF than the background population in the one-sample t-test (t[298]=2.034, p=0.043).

Diabetes duration and glycaemic control

To investigate the role of diabetes duration and glycaemic control in the likelihood of AF being diagnosed, these variables were added to the above logistic regression in Block 2 along with age and sex. The model was statistically significant (X²=12.58, p=0.013). The explained variation in the dependent variable based on our model is 14.1% (Nagelkerke R²=0.141). Neither diabetes duration (p=0.649) nor glycaemic control (p=0.349) added significantly to the model, with diabetes duration showing the least contribution. Therefore, age is the only predictor of AF in this study.

Screening locations

There was no significant association between screening location (diabetes centre and GP practices) and likelihood of detecting AF (X2[df 2,299]=3.641, p=0.314). That is, there was no effect of where screening took place.

Discussion

Diabetes as a predictor for AF (controlling for age and gender)

Age is the only significant predictor of AF in this study. This study was open to people over 18 years of age (range 22 to 90 years). The mean age of people with AF was higher than those in a normal rhythm. Previous research shows that AF prevalence shows a strong age dependence varying from 0.5% in patients under 40 years, 5% over 65 years and 10% in octogenarians.1 Ageing is accompanied by atrial structural remodelling, and this is associated with conduction abnormalities.19 These anatomical changes, along with comorbid conditions, enhance the risk of developing AF. These findings, therefore, support screening for AF in older age groups and can help focus screening resource and approach. Opportunistic screening in those over 65 years is recommended by clinical practice guidelines and expert consensus.1,20,21 A systematic approach, whereby those who are older or at higher stroke risk are targeted, is an alternative recommendation.1 Therefore, if focus can be directed to people with diabetes over the age of 65 years, this might offer a feasible and effective approach to AF screening.

Difference between men and women in terms of frequency with which AF is identified

This research demonstrated no significant difference in the likelihood of men and women with diabetes being diagnosed with AF but a trend in that direction. Previous research has shown that AF is twice as common in men than women, and men develop AF, on average, 10 years earlier.22 Women, however, live longer and so the cumulative lifetime risk of AF is similar.22 The lack of significant difference in this study may be due to the small numbers with AF. Men may have a higher risk of AF because of higher risk factors, but in this population already selected for risk factors (diabetes), less difference may be anticipated.

Prevalence of AF in the diabetes screened population compared with prevalence of AF in the general population

AF prevalence was 5.3% in this study compared with 2–4% in the general population. Existing literature has shown between 5.2% and 47.1% of people with diabetes also have AF.23,24 There are limitations to existing studies, including lack of adjustment of common risk factors, smaller samples, varying methodologies or where prevalence may be underestimated through less rigorous approaches to detection.25-27 Detection of paroxysmal AF is an issue when monitoring is of short duration, as in this research. A meta-analysis on over 108,000 patients indicated people with diabetes had a 40% greater risk of AF compared with those unaffected.28 The outcomes, therefore, in this study, while higher than the general population, do not reflect the higher prevalence seen in other studies. This may be due to sample size, the voluntary nature of participation and missed paroxysmal AF. Performing larger screening studies with diabetes as the target group would be advantageous while utilising repeated monitoring through interval screening or consecutive home readings using a portable ECG application.

Diabetes duration and glycaemic control

Neither diabetes duration nor stability were significant predictors for AF in this study. The mean HbA1c was comparable in both groups. There were, however, only 16 people identified as having AF, so this could impact findings. Diabetes duration has been suggested as relevant in the development of AF, with risk increasing by 3% for each year of treatment.29 The same study also highlighted higher AF risk with increased glucose levels. Higher HbA1c levels had a significant association with incident AF in prospective cohort studies but not in retrospective case-control studies.29 Poor glycaemic control has been further reported to increase AF risk in a recent subanalysis of patients with diabetes and AF.30 While this was not replicated in this research, it is worthy of consideration for future research with larger samples, and where recruitment was not, in part, dependent on patient initiation.

Numbers of people detected as having AF in the two screening locations

There was no difference in the number of people detected with AF between screening locations. Patients from GP surgeries required motivation to attend an appointment, perhaps representing a more engaged approach to healthcare with less symptoms. Furthermore, the latter part of screening recruitment occurred during a pandemic and involvement could have been influenced by other factors (isolation, health focus, availability).

Participants from the diabetes centre were invited consecutively and this direct approach may have influenced recruitment. Diabetes and health stability in these participants should be considered, but while this group were attending a specialist centre, screening times were random, and reason for attendance varied (routine monitoring, podiatry, nurse advice, specialist input). This, therefore, may contribute to there being no significant difference in AF detection between the two locations.

Limitations

The sample size was slightly lower than the sample size calculation, as it became challenging to recruit further due to aforementioned reasons. As a single-point in time screening study, paroxysmal AF may have been missed, however, this approach still offers value in AF detection for persistent and permanent AF. Prevalence was lower than some other studies where a higher prevalence of AF in people with diabetes has been detected. Contributing factors may also include recruitment approach between locations.

Conclusion

Findings from this study have shown that age is the only predictor for AF in people with diabetes. There was a significantly greater prevalence of AF in this patient group, than in the general population. There was no difference in AF detection between recruitment locations or sex. This adds to existing data around the association of the two chronic conditions and assists in guiding the importance of AF screening in people with diabetes, particularly older patients. Larger screening studies would be advantageous to explore the variables within this study further. This can then inform and contribute to appropriate management of both conditions when in combination, not least with regards to stroke prevention. Up until now, research into screening high-risk patient groups for AF has been approached in combination, rather than as individual risk-factor groups. This can confuse findings, if not controlled for within analyses. Understanding more about the risks imposed by individual risk factors is valuable when allocating resources for AF screening.

Key messages

  • Atrial fibrillation (AF) and diabetes are increasing in prevalence and are risk factors for stroke. Screening for AF can help identify and reduce stroke risk
  • Screening of people with diabetes rather than mixed high-risk patient groups has not been concentrated on as the lone, target group in previous research. This study also considers the impact of diabetes stability and duration
  • Identifying populations at high-risk can help when planning screening approaches and programmes in order to ensure they are effective in terms of cost and resource while reducing stroke risk

Conflicts of interest

None declared.

Funding

None.

Study approval

The study was approved by the University of Lancaster Faculty of Health and Medicine Research Ethics Committee (Ref: FHMREC18070) and the Health and Community Service Research Ethics Committee in Jersey.

Patient consent

Patients were provided with an information sheet and had the opportunity to ask questions, their right to withdraw was explained and they were informed that declining participation would not affect treatment. If they agreed, patients signed a consent form.

Data availability

The raw data required to reproduce these findings are available on request, held on a saved file held by the lead author. This is, therefore, not available on a public repository.

References

1. 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). Eur Heart J 2000;42:373–498. https://doi.org/10.1093/eurheartj/ehaa612

2. Wild S, Roglic G, Green A et al. Global prevalence of diabetes: estimates for year 2000 and projections for 2030. Diabetes Care 2004;21:1047–53. https://doi.org/10.2337/diacare.27.5.1047

3. Marini C, De Santis F, Sacco S et al. Contribution of atrial fibrillation to incidence and outcome of ischaemic stroke: results from a population-based study. Stroke 2005;36:1115–19. https://doi.org/10.1161/01.STR.0000166053.83476.4a

4. Du X, Ninomiya T, de Galan B et al. Risks of cardiovascular events and effects of routine blood pressure lowering among patients with type 2 diabetes and atrial fibrillation: results of the ADVANCE study. Eur Heart J 2009;30:1128–35. https://doi.org/10.1093/eurheartj/ehp055

5. Movahed MR, Hashemzadeh M, Jamal MM. Diabetes mellitus is a strong, independent risk factor for atrial fibrillation and flutter in addition to other cardiovascular disease. Int J Cardiol 2005;105:315–18. https://doi.org/10.1016/j.ijcard.2005.02.050

6. De Sensi F, De Potter T, Cresti A et al. Atrial fibrillation in patients with diabetes: molecular mechanisms and therapeutic perspectives. Cardiovasc Diagn Ther 2015;5:364–73. https://doi.org/10.3978/j.issn.2223-3652.2015.06.03

7. Kavousi M. Differences in epidemiology and risk factors for atrial fibrillation between women and men. Front Cardiovasc Med 2016;7:3. https://doi.org/10.3389/fcvm.2020.00003

8. International Diabetes Federation: IDF Diabetes Atlas, 5th edition. Brussels: IDF, 2011. Available from: https://diabetesatlas.org/atlas/fifth-edition/

9. Samol A, Masin M, Gellner R et al. Prevalence of unknown atrial fibrillation in patients with risk factors. Europace 2013;15:657–62. https://doi.org/10.1093/europace/eus366

10. Turakhia M, Ullal A, Hoang D et al. Feasibility of extended ambulatory electrocardiogram monitoring to identify silent atrial fibrillation in high-risk patients: the screening study for undiagnosed atrial fibrillation (STUDY-AF). Clinical Card 2015;38:285–92. https://doi.org/10.1002/clc.22387

11. Davis R, Hobbs R, Kenkre J et al. Prevalence of atrial fibrillation in the general population and in high-risk groups: the ECHOES study. Europace 2012;14:1553–9. https://doi.org/10.1093/europace/eus087

12. AliveCor. AliveCor’s heart monitor for iPhone receives FDA clearance. Press release – 3 December 2012. Available at: https://www.kardia.com/press/press_release/alivecors-heart-monitor-for-iphone-receives-fda-clearance/

13. Lau J, Lowres N, Neubeck L et al. iPhone ECG application for community screening to detect silent atrial fibrillation: a novel technology to prevent stroke. Int J Cardiol 2013;165:193–4. https://doi.org/10.1016/j.ijcard.2013.01.220

14. Hall A, Mitchell A, Wood L et al. Effectiveness of a single lead AliveCor electrocardiogram application for the screening of atrial fibrillation: a systematic review. Medicine 2020;99:e21388. https://doi.org/10.1097/MD.0000000000021388

15. National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation. CG180. London: NICE, 2014. Available from: https://www.nice.org.uk/guidance/cg180

16. National Institute for Health and Care Excellence. KardiaMobile for the ambulatory detection of atrial fibrillation. London: NICE, 2020. Available from https://www.nice.org.uk/guidance/mib232

17. Public Health Statistics Unit. Disease projections 2016–2036. St. Helier: Statistics Jersey (Public Health), 2016. Available from: https://www.gov.je/Government/Pages/StatesReports.aspx?ReportID=3121

18. Public Health England. Public Health Profiles. 2019. Available at: https://fingertips.phe.org.uk/search/atrial [accessed 21 February 2021].

19. Sankaranarayanan R, Kirkwood G, Dibb K, Garratt CJ. Comparison of atrial fibrillation in the young versus that in the elderly: a review. Cardiol Res Pract 2013;2013:976976. https://doi.org/10.1155/2013/976976

20. Freedman B, Camm J, Calkins H et al.; AF-Screen Collaborators. Screening for atrial fibrillation: a report of the AF-SCREEN International Collaboration. Circulation 2017;135:1851–67. https://doi.org/10.1161/CIRCULATIONAHA.116.026693

21. Mairesse G, Moran P, Van Gelder I et al.; ESC Scientific Document Group. Screening for atrial fibrillation: a European Heart Rhythm Association (EHRA) consensus document endorsed by the Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS), and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología (SOLAECE). Europace 2017;19:1589–623. https://doi.org/10.1093/europace/eux177

22. 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. https://doi.org/10.1161/CIRCULATIONAHA.117.028981

23. Davis R, Hobbs R, Kenkre J et al. Prevalence of atrial fibrillation in the general population and in high-risk groups: the ECHOES study. Europace 2012;14:1553–9. https://doi.org/10.1093/europace/eus087

24. Sanmartίn M, Fraga F, Martίn-Santos Ά et al. A campaign for information and diagnosis of atrial fibrillation: ‘’pulse week’’. Rev Esp Cardiol (Engl Ed) 2013;66:34–8. https://doi.org/10.1016/j.rec.2012.05.009

25. Krijthe B, Kunst A, Benjamin E et al. Projections on the number of individuals with atrial fibrillation in the European Union from 2000 to 2060. Eur Heart J 2013;34:2746–51. https://doi.org/10.1093/eurheartj/eht280

26. Nichols G, Reinier K, Chugh S. Independent contribution of diabetes to increased prevalence and incidence of atrial fibrillation. Diabetes Care 2009;32:1851–6. https://doi.org/10.2337/dc09-0939

27. Dublin S, Glazer N, Smith N et al. Diabetes mellitus, glycaemic control, and risk of atrial fibrillation. J Gen Intern Med 2010;25:853–8. https://doi.org/10.1007/s11606-010-1340-y

28. Huxley R, Filion K, Konety S et al. Meta-analysis of cohort and case-control studies of type 2 diabetes mellitus and risk of atrial fibrillation. Am J Cardiol 2011;108:56–62. https://doi.org/10.1016/j.amjcard.2011.03.004

29. Zhang Y, Pan X, Chen J et al. Combined lifestyle factors and risk of incident type 2 diabetes and prognosis among individuals with type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies. Diabetologia 2020;63:21–33. https://doi.org/10.1007/s00125-019-04985-9

30. Ahmadi S, Svensson A, Pivodic A et al. Risk of atrial fibrillation in persons with type 2 diabetes and the excess risk in relation to glycaemic control and renal function: a Swedish cohort study. Cardiovasc Diabetol 2020;19:9. https://doi.org/10.1186/s12933-019-0983-1

THERE ARE CURRENTLY NO COMMENTS FOR THIS ARTICLE - LEAVE A COMMENT