Table 2.
Description of studies investigating the longitudinal association of DM with AF types
| First author, year (Reference) | Country | N | Population | Mean age (years) | Females (%) | DM assessment | Adjustment for potential confounders | Outcome (Ref) | Follow-up time | Result | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| A. Studies providing hazard ratios (Method: Cox regression) | |||||||||||
| 1 | Tsang, 2007 | United States | 3248 | Patients with first episode of paroxysmal AF | 71 ± 15 | 46 | DM medical diagnosis or treatment with antidiabetic medications | Age, sex |
Permanent AF (Ref: no AF recurrence, recurrent paroxysmal AF, recurrent persistent AF) |
Median (IQR), 5.1 (1.2–9.4) y |
aHR (CI), 1.17 (0.94–1.47) |
| 2 | Pappone, 2008 | Italy | 106 | Patients with first episode of paroxysmal AF | 57.5 | 35.8 | NR | Age, HF |
Permanent AF (Ref: no AF recurrence, recurrent paroxysmal AF, recurrent persistent AF) |
Maximum, 5 y | aHR (CI), 17.37 (3.75–80.43) |
| 3 | Kawara, 2010 | Japan | 64 | Paroxysmal or persistent AF | 61 ± 10 | 29.6 | NR | Age, sex, HTN, organic heart disease, HF, severity of symptoms |
Permanent AF (Ref: non-permanent AF) |
Median (IQR), 4.9 (2.4–8.9) y |
aHR (CI), 3.13 (0.46–21.2) |
| 4 | Thacker, 2013 | United States | 1385 | Patients whose initial AF episode terminated within 6 months | 69.2 | 48.7 |
DM medical diagnosis, and current use of insulin or oral hypoglycemic medication |
Age, sex | Permanent AF (Ref: non-permanent AF) |
Mean (range), 7 (5–8) y |
Model 1 aHR (CI), 0.99 (0.72–1.36) |
| Age, sex, BMI, HTN, SBP, DBP |
Model 2 aHR (CI), 0.93 (0.67–1.29) |
||||||||||
| Age, sex, BMI, HTN, SBP, DBP, CHD, valvular heart disease, HF, stroke |
Model 3 aHR (CI), 0.94 (0.67–1.32) |
||||||||||
| 5 | Senoo, 2014 | Japan | 1176 | Paroxysmal AF | 61.4 ± 13.1 | 25.6 | NR | - | Recurrent AF (Ref: non-recurrent AF) | Mean (sd), 3.3 ± 2.5 y | uHR (CI), 1.47 (0.89–2.44) |
| 6* | Sandhu, 2014 | United States | 1039 | Patients who developed AF in a cohort free of AF | 58.9 | 100 | NR | Age, aspirin, vitamin E, beta‐carotene, BMI, HTN, cholesterol, alcohol, smoking, exercise |
Non-paroxysmal AF [i.e., persistent or permanent] (Ref: paroxysmal AF) |
Median (IQR), 16.4 (15.6–16.8) y** |
Model 1 aHR (CI), 1.04 (0.72 to 1.41)*** |
| Age, MI, stroke, revascularization, HF |
Model 2 aHR (CI), 1.06 (0.69–1.5)*** |
||||||||||
| 7 | Blum, 2019 | Switzerland | 2869 | Paroxysmal or persistent AF | 70 ± 9 | 32.3 | NR | Age, sex |
AF Progression [i.e., paroxysmal to persistent or permanent, persistent to permanent] (Ref: paroxysmal or persistent AF) |
Median (IQR), 3 (2–5) y |
Model 1 aHR (CI), 1.14 (0.88–1.48) |
| Age, sex, BMI, heart rate, SBP, coronary artery disease, HTN, stroke/TIA, HF, hyperthyroidism, history of renal failure, physical activity, smoking, history of pulmonary vein isolation, AF‐related symptoms, amiodarone |
Model 2 aHR (CI), 0.92 (0.69–1.21) |
||||||||||
| B. Studies providing odds ratios (Method: Logistic regression) | |||||||||||
| 8*† | Sakamoto, 1995 | Japan | 137 | Paroxysmal AF | 64 | 24.8 | Use of antidiabetic therapy | – | Chronic AF (Ref: paroxysmal AF) | Mean, 1 y | uOR (CI), 2.52 (1.05–6.05) |
| 9*† | Kerr, 2005 | Canada | 757 | Paroxysmal AF | 64 | 38.3 | NR | – |
Chronic AF (Ref: paroxysmal AF) |
8 (0–11) y | uOR (CI), 1.07 (0.60–1.92) |
| 10* | Pillarisetti, 2009 | United States | 437 | Paroxysmal AF | 67.9 ± 13.4 | 43 | NR | – |
Non-paroxysmal AF [i.e., persistant or permanent] (Ref: paroxysmal AF) |
Mean (sd), 4.7 ± 4.6 y | uOR (CI), 1.50 (0.90–2.60) |
| 11*† | de Vos, 2010 | 35 European countries | 1219 |
Paroxysmal AF and first detected AF in whom sinus rhythm restored spontaneously or after treatment during admission |
64 | 43 | NR | – | Non-paroxysmal AF [i.e., persistent or permanent] (Ref: paroxysmal AF) | Mean, 1 y | uOR (CI), 1.42 (0.94–2.14) |
| 12*† | de Vos, 2012 | 21 countries in Europe, North and South America, Asia | 2137 | Paroxysmal and first-detected AF | 65 ± 12 | NR | NR | – | Non-paroxysmal AF [i.e., persistent or permanent] (Ref: paroxysmal AF) | Mean, 1 y | uOR (CI), 1.18 (0.86–1.63) |
| 13 | Echouffo-Tcheugui, 2017 | United States | 6575 | Paroxysmal or persistent AF | 75 | 42.6 | Previous medical history or new DM diagnosis during enrollment visit | – | AF progression [i.e., paroxysmal to persistent or permanent, persistent to permanent] (Ref: paroxysmal or persistent AF) | Median (IQR), 2.78 (1.95–3) y |
Model 1 uOR (CI), 1.05 (0.93–1.17) |
| Age, sex, race, medical history, cardiovascular history |
Model 2 aOR (CI), 0.96 (0.85, 1.08) |
||||||||||
| 14 | Schnabel, 2018 | 7 European countries | 2151 | Paroxysmal or persistent AF | 72 | 39.9 | NR | – | AF progression [i.e., paroxysmal to persistent or permanent, persistent to permanent] (Ref: paroxysmal or persistent AF) | Mean, 1 y | uOR (CI), 1.22 (0.98–1.52) |
| Age, sex, country | aOR (CI), 1.23 (0.98–1.53) | ||||||||||
| AF duration, HF, hyperthyroidism, no sinus rhythm, cardioversion, valvular heart disease | aOR (CI), 1.29 (1.01–1.65) | ||||||||||
DM diabetes mellitus, HR hazard ratio, AF atrial fibrillation, BMI body mass index, HTN hypertension, aHR adjusted hazard ratio, CI 95% confidence interval, uHR unadjusted hazard ratio, PAF paroxysmal atrial fibrillation, aOR adjusted odds ratio, uOR unadjusted odds ratio, NR not reported, CHD coronary heart disease, LVH left ventricular hypertrophy, HF heart failure, SBP systolic blood pressure, DBP diastolic blood pressure, CHD coronary heart disease, MI myocardial infarction, y years, IQR interquartile range, Ref reference, sd standard deviation
*The studies investigated the association of diabetes with nonparoxysmal AF (compared to paroxysmal AF)
**Median follow-up time of the entire cohort of 34,720 women without AF at baseline, of which 1039 developed AF over follow-up
***HRs of developing non-paroxysmal AF compared to paroxysmal AF were calculated based on: (i) HR of developing paroxysmal AF compared to no AF; (ii) HR of developing non-paroxysmal AF compared to no AF; (iii) p-value from likelihood ratio tests of the null hypothesis that diabetes has an equal effect on the development of paroxysmal vs non-paroxysmal AF
†Odds ratios for developing non-paroxysmal AF compared to paroxysmal AF were calculated manually