Research report: Prevalence and factors associated with atrial fibrillation among adults with type 2 diabetes mellitus at Hoima Regional Referral Hospital, Western Uganda: a cross-sectional study

Atrial fibrillation, the most common sustained cardiac arrhythmia worldwide, is far from rare among adults living with type 2 diabetes in western Uganda, according to a new hospital-based study. Researchers screening patients at Hoima Regional Referral Hospital found that 28 of 355 adults with type 2 diabetes carried an electrocardiographic diagnosis of atrial fibrillation, a prevalence of 7.9 percent with a 95 percent confidence interval of 5.4 to 10.7 percent. That figure places the local burden squarely within the range of global estimates and, notably, represents the first such data reported from Uganda, where routine electrocardiographic screening of diabetes patients has rarely been performed.

The study, a cross-sectional survey conducted between April and July 2025 at Hoima Regional Referral Hospital, a public teaching hospital serving the Bunyoro sub-region, enrolled consenting adults aged 18 and older with confirmed type 2 diabetes who attended the hospital’s weekly diabetes clinic. Each participant underwent a structured clinical interview, physical examination with weight, height and blood pressure measurement, laboratory testing of glycated hemoglobin, and a resting 12-lead electrocardiogram. Atrial fibrillation was diagnosed on standard electrocardiographic criteria: an irregularly irregular rhythm with absence of discernible P waves. Patients who were critically ill, taking anti-arrhythmic medications, or already known to have atrial fibrillation were excluded from the analysis.

Participants were recruited by consecutive sampling, a non-probability method the authors acknowledge may introduce selection bias and limit generalizability beyond similar hospital-based populations. The sample size was calculated using Daniel’s formula for prevalence studies, assuming a conservative 30 percent prevalence drawn from a hospital-based study outside sub-Saharan Africa, with a 95 percent confidence level and a 5 percent margin of error. That calculation yielded a minimum of 323 participants, and a 10 percent adjustment for non-response brought the final target to 355. The authors note that the conservative prevalence assumption likely produced a larger sample than strictly required, improving the precision of the estimates.

The demographic profile of the cohort reflects the population attending the referral clinic. Slightly more than half of the participants, 54.6 percent, were male, and 25.6 percent were aged below 45 years. About 35.2 percent had lived with type 2 diabetes for more than a decade, and the majority, 69.0 percent, were being managed with oral antidiabetic medications. Information on hypertension and HIV status was drawn from participants’ clinical records, with hypertension defined by documented diagnosis or current antihypertensive use, and HIV status confirmed by serological results or antiretroviral therapy records.

Clinical presentation among those diagnosed with atrial fibrillation followed the textbook picture of the arrhythmia. Palpitations, dizziness, fatigue, chest pain, and difficulty breathing were all significantly more common in participants with atrial fibrillation, with p-values below 0.05 for each comparison. On examination, tachycardia and an irregular pulse were also more frequently observed in this group. The authors attribute this symptom cluster to the hemodynamic consequences of an irregular ventricular response and impaired cardiac output, and caution that reliance on symptoms alone would miss asymptomatic or paroxysmal cases, underscoring the need for electrocardiographic confirmation.

The heart of the analysis lies in its multivariable logistic regression, which identified four factors independently associated with atrial fibrillation. Hypertension roughly tripled the odds, with an adjusted odds ratio of 3.027 (95 percent CI 1.419 to 6.431; p = 0.020). Poor glycemic control, defined as glycated hemoglobin of 8 percent or higher, carried the largest effect size, with an adjusted odds ratio of 3.813 (95 percent CI 1.762 to 8.265; p = 0.007). Elevated body mass index of 25 kilograms per square meter or above showed an adjusted odds ratio of 3.014 (95 percent CI 1.242 to 7.930; p = 0.029), and HIV infection was associated with an adjusted odds ratio of 2.026 (95 percent CI 1.738 to 6.538; p = 0.048). Variables considered for the model, selected because they showed p-values of 0.20 or less in bivariable analysis, included sex, age category, hypertension, HIV infection, smoking, alcohol intake, diabetic complications, type of diabetes treatment, body mass index and glycated hemoglobin category. The authors report no multicollinearity, with all variance inflation factors below 5 (range 1.012 to 1.036), and a Hosmer–Lemeshow test indicating good calibration (chi-squared 6.046, df = 8, p = 0.642).

Each of these associations fits with established mechanistic understanding of how cardiometabolic disease reshapes atrial tissue. Chronic pressure overload from hypertension can drive left atrial enlargement, fibrosis and electrical instability, creating the substrate for disorganized atrial electrical activity. Chronic hyperglycemia promotes oxidative stress, inflammation and structural remodeling of atrial tissue, pathways consistent with the observed link between poor glycemic control and arrhythmia risk. Obesity contributes to left atrial enlargement, systemic inflammation and neurohormonal activation, all of which facilitate atrial fibrillation. The HIV association, plausible through chronic immune activation, systemic inflammation and endothelial dysfunction, carries particular weight in a region where antiretroviral therapy has extended survival and the non-communicable disease burden among people living with HIV is expected to grow.

Comparative context situates the Ugandan figure within a global spectrum. The observed prevalence of 7.9 percent is slightly higher than the 6.1 percent reported in a hospital-based Tanzanian cohort screened by electrocardiography, and higher than a Japanese study documenting atrial fibrillation in 4.4 percent of adults with type 2 diabetes. It is lower than the 11.7 percent seen among older adults with type 2 diabetes in a Pakistani cross-sectional study, and slightly above the approximately 7.6 percent pooled prevalence reported in a recent systematic review of multiple type 2 diabetes cohorts. The authors attribute such variation to differences in age distribution, study settings, cardiovascular comorbidity burden and detection strategies, particularly the use of systematic electrocardiographic screening.

The study’s conclusions carry caveats that the authors confront directly. The cross-sectional design precludes causal inference, and reverse causality cannot be excluded: poor glycemic control may promote atrial fibrillation, but atrial fibrillation itself, through stress-related catecholamine release and hemodynamic effects, could worsen glycemic control. The single-center, hospital-based recruitment may overestimate associations, since tertiary-care patients tend to carry more comorbidities than the general diabetes population. Survivorship bias may work in the opposite direction, as patients with the most severe disease may not survive to be counted in outpatient settings. Because diagnosis rested on a single resting 12-lead electrocardiogram, paroxysmal episodes could have been missed, meaning the reported figure is best understood as point prevalence rather than cumulative burden, and the true prevalence may be higher. Echocardiographic evaluation was not performed, leaving structural contributors such as left atrial enlargement and valvular disease unassessed.

For clinical practice in resource-limited settings, the authors propose a pragmatic path: rather than universal electrocardiographic screening, which may be constrained by cost, infrastructure and workforce, clinics could target high-risk patients, particularly those with hypertension, poor glycemic control, overweight or obesity, or HIV infection. The authors frame their findings as a foundation for future multicenter and prospective research incorporating echocardiographic assessment and long-term rhythm monitoring.

Subject of Research: A cross-sectional study of atrial fibrillation prevalence and associated factors among adults with type 2 diabetes in western Uganda.

Article Title: Prevalence and factors associated with atrial fibrillation among adults with type 2 diabetes mellitus at Hoima Regional Referral Hospital, Western Uganda: a cross-sectional study

Article References: Hersi, A. G., Sandeyl, A. A., Abdi, F. D., Jayte, M., Mswelo, V. E., Hussein, Z. A., Asad, H., Jama, H. A., Yusuf, A. A., Karshe, A. H., Agwu, E., Abdallah, M. A., Amandua, J., & Hirsi, A. M. (2026). Prevalence and factors associated with atrial fibrillation among adults with type 2 diabetes mellitus at Hoima Regional Referral Hospital, Western Uganda: a cross-sectional study. Cardiovascular Diabetology – Endocrinology Reports, 12(1), Article 55. https://doi.org/10.1186/s40842-026-00327-y

Image Credits: AI Generated

DOI: 10.1186/s40842-026-00327-y

Keywords: atrial fibrillation, type 2 diabetes, Uganda, cross-sectional study, ECG screening, hypertension, HIV, glycemic control, body mass index, HbA1c, sub-Saharan Africa, cardiovascular risk