Identification of Biomarkers for Predicting New-Onset Atrial Fibrillation: A Systematic Review and Meta-Analysis

Aims Atrial fibrillation (AF) is a progressive condition characterized by atrial remodeling and dysfunction. This systematic review explores biomarkers that predict new-onset AF, highlighting their potential to improve early diagnosis and risk stratification in high-risk patients, and prevention of stroke and major adverse cardiovascular events. Methods We conducted a literature search of studies published between January 2014-November 2025 in PubMed, Scopus, Web of Science, and Google Scholar, following PRISMA 2020 guidelines. Studies analysing specific populations and patients with prior or postoperative AF were excluded. Quality was assessed using the Newcastle-Ottawa scale. Effect sizes were expressed as HR with 95% CIs. Results We included 10 cohort studies comprising 472,581patients and 35,271 (7.5%) new-onset AF. Overall, 18 biomarkers were associated with an increased risk of AF, most notably NT-proBNP and sVCAM-1. Conversely, 9 biomarkers were associated with a lower AF incidence, such as ADAMTS13 (HR 0.78, 95%CI 0.70-0.88). A meta-analysis of NT-proBNP demonstrated its association with a higher incidence of AF (HR 1.37, 95%CI 1.19-1.59) with high heterogeneity (I2 = 80%, p<0.01) and Lp(a) was associated with a significant 3% increase in AF incidence per 20 mg/dL increment. Two networks were constructed according to whether biomarkers were associated with a higher or lower incidence of AF, visualising their connection with other biomarkers. Conclusions Well-known biomarkers, such as NT-proBNP, and others not yet incorporated into clinical practice, such as Lp(a) and sVCAM-1, could play a role in the diagnosis and preventive management of AF. Large-scale prospective studies are needed to validate and optimise their diagnostic utility in predicting new-onset AF.