Table 1. Basic characteristics of the 14 articles included in the meta-analysis.
| Author, publication year, country | Population, Study design, follow-up duration | Source of participant, Cases/N | Definition of AF, Measurement H1Ac1, Mean age (years), male (%) | H1Ac1 expose level | RR (95%CI) | Adjustment for confounders |
|---|---|---|---|---|---|---|
| Halkos, 2008, USA [15] | CABG, Retrospective cohort, NA | Emory University Hospital, 549/3089 | ECG, NA, 63, 73 | Per 1% | 0.89(0.80–0.98) | age, female, caucasian, renal failure, stroke, NYHA class IV,CCS class, main disease, No. diseased vessels, chronic lung disease, arrhythmia, peripheral vascular disease, perioperative factors. intraoperative glucose, postoperative glucose, arterial grafts, vein grafts, total grafts, CPB time, CPB used, LITA or BITA used |
| Matsuura, 2009, Japan [19] | CABG, Retrospective cohort, 2.1 years | Chiba University Hospital, 26/101 | NA, ion capture assay, 65, 79 | Per 1% | 0.87(0.64–1.19)# | NA |
| Tsuruta, 2011, Japan [21] | CABG, Prospective cohort, 3.6 years | Juntendo University Hospital of Japan, 36/305 | NA, NA, 60, 79 | <6.5%; | Ref | NA |
| 6.5%-7.5% | 0.89(0.36–2.21) | |||||
| ≥7.5% | 1.25(0.58–2.71) | |||||
| Per 1%. | 1.12(0.76–1.65) | |||||
| Kinoshita, 2012, Japan [14] | CABG, Case-control, NA | Shiga University of Medical Science, 159/805 | ECG, high-performance liquid chromatography, 68, 80 | 3.8–5.6% | Ref | age, sex, BMI, chronic kidney disease, chronic pulmonary disease, hypertension, triple vessel disease, ejection fraction of <40%, left atrial dimension, preoperative beta blockers, preoperative statins, inotropic support for >24 h, and transfusion |
| 6.8–11.4% | 0.55 (0.35–0.88) | |||||
| per 1% | 0.78(0.63–0.95) | |||||
| Surer, 2016,Turkey [20] | CABG, Retrospective cohort, NA | Diskapi Yildirim Beyazit Training & Research Hospital, 12/72 | ECG, liquid chromatography, 63, 56 | Per 1% | 3.92(1.92–7.99) | NA |
| Abbaszadeh, 2017, Iran [17] | CABG, Prospective cohort, 1 years | Tehran University of Medical Sciences, 109/708 | ECG, immunoturbidimetric, 61, 61 | Per 1% | 1.06(0.93–1.2) | age, duration of diabetes, COPD, renal failure and hypertension, postoperative use of beta-blockers and calcium channel blockers, left atrial size, cardiopulmonary pump duration, and cross clamp time |
| Dublin, 2010, USA [6] | Population-based, Prospective case-control, 9.4 years | Group Health database of Research Institute of US, 1410/3613 | GH electronic data, calculated by total glycosylated hemoglobin, 70, 41 | 5%* | Ref | age, sex, calendar year, treated hypertension, and BMI |
| ≤7% | 1.06(0.74–1.51) | |||||
| 7–8% | 1.48(1.09–2.01) | |||||
| 8–9% | 1.46(1.02–2.08) | |||||
| >9% | 1.96(1.22–3.14) | |||||
| Per 1% | 1.14(0.96–1.35) | |||||
| Huxley, 2012, USA [5] | DM or non-DM, Prospective cohort study, 14.5 years | ARIC study, 1311/13025 | ECG; high-performance liquid chromatography, 57, 44 | Per 1% (non-DM) | 1.05(0.96–1.15) | age, study site, education, income, prevalent CHD, BMI, systolic BP, antihypertensive medications, and smoking |
| Per 1% (DM) | 1.13(1.07–1.20) | |||||
| Iguchi, 2012, Japan [18] | Population-based, Prospective cohort, NA | Kurashiki city Public Health Center, 1161/52448 | ECG, NA, 72, 34 | Per 1% | 1.18(1.09–1.28) | age, sex, BMI, chronic kidney disease, COPD, hypertension, triple vessel disease, ejection fraction of <40%, left atrial dimension, preoperative beta blockers, preoperative statins, inotropic support for >24 h, and transfusion |
| Turgut, 2013, Turkey [9] | DM, Retrospective Case-control, NA | Cumhuriyet University of Turkey, 81/162 | Medical records, turbidimetric immunoinhibition assay, 63, 52 | Per 1% | 1.87(0.747–3.014) | age, male gender, HT, smoking, history of CAD, previous CVA, microalbuminuria, retinopathy, duration of diabetes, BMI, total cholesterol, LDL cholesterol, HDL cholesterol, triglyceride, platelet count, and MPV |
| Latini,2013, Italy [4] | Population with impaired glucose tolerance, Post hoc analysis of RCT, 6.5 years | NAVIGATOR study, 613/8943 | ECG or investigator-reported event, NA, 63, 49 | Per 1% | 1.11 (0.91–1.32) | age, height, abnormal ECG, heart rate, COPD, BP, weight, hypertension, eGFR, fasting plasma glucose, CHF, race, CHD outcomes, platelet |
| Sandhu, 2014, Canada [8] | Population-based(women), Prospective cohort, 16.4 years | Women’s Health Study, 1039/34720 | ECG or medical report, NA, 53, 0 | Paroxysmal AF | interim MI, stroke, revascularization, and HF | |
| Per 1% | 1.08(0.95–1.22)# | |||||
| ≤4.84% | Ref | |||||
| 4.84–5.00% | 0.90(0.69–1.17) | |||||
| 5.00–5.19% | 0.99(0.77–1.27) | |||||
| >5.19% | 0.76(0.58–1.00) | |||||
| Nonparoxysmal AF | ||||||
| ≤4.84% | Ref | |||||
| 4.84–5.00% | 1.30(0.85–1.97) | |||||
| 5.00–5.19% | 1.43(0.96–2.15) | |||||
| >5.19% | 1.48(0.98–2.22) | |||||
| Blasco, 2014, Spain [7] | AMI, Prospective cohort, 8 days | University Clinic Hospital of Valencia, 12/601 | ECG, ion-exchange high-performance liquid chromatography, 62, 78 | <5.5% | Ref | age, sex, classical cardiovascular risk factors, Killip class, history of previous MI, serum creatinine, serum glycaemia, systolic, BP and heart rate on admission. |
| 5.5–6.4% | 1.68(0.64–4.39) | |||||
| >6.4% | 29.74(10.79–81.94) | |||||
| per 1% | 3.24(2.41–4.35)# | |||||
| Dahlqvist, 2017, Sweden [13] | type 1 diabetes, prospective case-control study | Swedish National Diabetes Registry, 1283/ 216852 | ECG, 35.4, 55 | Control | Ref | age, sex, education level, birthplace, country of birth, diabetes duration, level of education, mean systolic BP, BMI, HDL cholesterol, and LDL cholesterol, and time-updated smoking, use of blood pressure-lowering medication, use of lipid-lowering drugs, and insulin delivery method, CAD, HF, valve disease, stroke, and cancer |
| <6.9 | 1.03(0.84–1.27) | |||||
| 7.0–7.8 | 1.11(0.97–1.28) | |||||
| 7.8–8.7 | 1.28(1.12–1.47) | |||||
| 8.8–9.6 | 1.4(1.15–1.69) | |||||
| >9.7 | 2.32(1.8–2.99) | |||||
| Per 1% | 1.14(1.06–1.24) |
Abbreviations: CABG = coronary artery bypass grafting; DM = diabetes mellitus; AMI = acute myocardial infarction; ARIC = The Atherosclerosis Risk in Communities study; H1Ac1 = Hemoglobin A1c; ECG = electrocardiogram; AF = atrial fibrillation; COPD = chronic obstructive pulmonary disease. NYHA = New York Heart Association; BMI = Body Mass Index HDL = high density lipoprotein; CPB = cardiopulmonary bypass; CCS = Canadian Cardiovascular Society; LITA BITA = bilateral internal thoracic artery; eGFR = estimated creatinine clearance; CHD = Coronary heart disease; HF = heart failure; BP = blood pressure; CAD = coronary artery disease; NAVIGATOR = Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research. NA = not available.
* H1Ac1 level of no history of DM
#Calculated from the number of AF occurrence in different HbA1c levels