Association of diabetes with atrial fibrillation types: a systematic review and meta-analysis

Alijla, Fadi; Buttia, Chepkoech; Reichlin, Tobias; Razvi, Salman; Minder, Beatrice; Wilhelm, Matthias; Muka, Taulant; Franco, Oscar H.; Bano, Arjola (2021). Association of diabetes with atrial fibrillation types: a systematic review and meta-analysis. Cardiovascular diabetology, 20(1), p. 230. BioMed Central 10.1186/s12933-021-01423-2

Alijla_CardiovascDiabetol_2021.pdf - Published Version
Available under License Creative Commons: Attribution (CC-BY).

Download (945kB) | Preview

Background: Atrial fibrillation (AF) is a common arrhythmia classified as paroxysmal and non-paroxysmal. Non-paroxysmal AF is associated with an increased risk of complications. Diabetes contributes to AF initiation, yet its role in AF maintenance is unclear. We conducted a systematic review and meta-analysis to summarize the evidence regarding the association of diabetes with AF types.

Methods: We searched 5 databases for observational studies investigating the association of diabetes with the likelihood of an AF type (vs another type) in humans. Study quality was evaluated using the Newcastle-Ottawa Scale. Studies classifying AF types as paroxysmal (reference) and non-paroxysmal were pooled in a meta-analysis using random effects models.

Results: Of 1997 articles we identified, 20 were included in our systematic review. The population sample size ranged from 64 to 9816 participants with mean age ranging from 40 to 75 years and percentage of women from 24.8 to 100%. The quality of studies varied from poor (60%) to fair (5%) to good (35%). In the systematic review, 8 studies among patients with AF investigated the cross-sectional association of diabetes with non-paroxysmal AF (vs paroxysmal) of which 6 showed a positive association and 2 showed no association. Fourteen studies investigated the longitudinal association of diabetes with "more sustained" AF types (vs "less sustained") of which 2 showed a positive association and 12 showed no association. In the meta-analysis of cross-sectional studies, patients with AF and diabetes were 1.31-times more likely to have non-paroxysmal AF than those without diabetes [8 studies; pooled OR (95% CI), 1.31 (1.13-1.51), I2 = 82.6%]. The meta-analysis of longitudinal studies showed that for patients with paroxysmal AF, diabetes is associated with 1.32-times increased likelihood of progression to non-paroxysmal AF [five studies; pooled OR (95% CI), 1.32 (1.07-1.62); I2 = 0%].

Conclusions: Our findings suggest that diabetes is associated with an increased likelihood of non-paroxysmal AF rather than paroxysmal AF. However, further high quality studies are needed to replicate these findings, adjust for potential confounders, elucidate mechanisms linking diabetes to non-paroxysmal AF, and assess the impact of antidiabetic medications on AF types. These strategies could eventually help decrease the risk of non-paroxysmal AF among patients with diabetes.

Item Type:

Journal Article (Original Article)


13 Central Units > Administrative Director's Office > University Library of Bern
04 Faculty of Medicine > Department of Cardiovascular Disorders (DHGE) > Clinic of Cardiology
04 Faculty of Medicine > Pre-clinic Human Medicine > Institute of Social and Preventive Medicine (ISPM)

Graduate School:

Graduate School for Health Sciences (GHS)

UniBE Contributor:

Alijla, Fadi, Buttia, Chepkoech, Reichlin, Tobias Roman, Minder, Beatrice, Wilhelm, Matthias, Muka, Taulant, Franco Duran, Oscar Horacio, Bano, Arjola


000 Computer science, knowledge & systems > 020 Library & information sciences
600 Technology > 610 Medicine & health
300 Social sciences, sociology & anthropology > 360 Social problems & social services




BioMed Central




Matthias Wilhelm

Date Deposited:

13 Dec 2021 19:22

Last Modified:

05 Dec 2022 15:56

Publisher DOI:


Related URLs:

PubMed ID:


Additional Information:

Fadi Alijla and Chepkoech Buttia contributed equally to this work.




Actions (login required)

Edit item Edit item
Provide Feedback