Assessment of HHARMS2-AF for risk stratification: a 10-year follow-up study (2014-2024) in a northwestern Colombian population

Abstract

Atrial fibrillation (AF) is the most prevalent arrhythmia worldwide and represents a significant public health burden. Several modifiable risk factors (RFs)—such as hypertension, obesity, smoking, alcohol consumption, sleep apnea, and physical inactivity—contribute to its development. However, Latin American populations have not thoroughly assessed their combined effect.

Objective: To develop and validate the HARMS2-AF model to predict AF risk based on modifiable RFs, assessing its applicability in primary care settings.

Materials and Methods: A retrospective cohort study was conducted with 2,482 participants from Northwestern Colombia, followed over 10 years. RFs evaluated included hypertension, obesity (BMI ≥30 kg/m²), smoking, alcohol consumption (≥15 standard drinks/week), sleep apnea, physical inactivity, diabetes mellitus, age, and sex. Cox proportional hazards regression models (univariate and multivariate) were used to generate HARMS2-AF scores based on beta coefficients. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC) and calibration tests.

Results: Hypertension was the most prevalent RF (79.1%) and the strongest predictor of AF (HR 4.0; 95% CI: 3.2–4.8). Obesity, high alcohol intake, smoking, and sleep apnea were also significantly associated with AF. Physical inactivity and diabetes mellitus did not reach statistical significance in multivariate analysis. The HARMS2-AF model showed a two-fold increase in AF risk per 1-point increment (HR 2.38; 95% CI: 2.14–2.62), with an AUROC of 0.81.

Conclusion: The HARMS2-AF model is a simple and effective tool for predicting AF risk based on modifiable RFs. Its implementation in primary care could enhance early identification and targeted prevention strategies in Latin American populations.