Improved software to predict a patient’s risk of atrial fibrillation has demonstrated key differences in how the condition presents in men and women, opening the door to more accurate diagnostics.

Health

Atrial fibrillation (AF) is an abnormal heart rhythm characterised by a rapid, irregular heartbeat, and is associated with an increased risk of heart failure, stroke and dementia. Experts predict that due to population ageing in Europe, the prevalence of AF will double by 2050. The incidence and morbidity of AF vary between men and women, and this picture is complicated by the fact that many patients are asymptomatic. A few years ago, cardiologist Renate Schnabel at the University Heart and Vascular Center Hamburg in Germany and her team developed an innovative prediction algorithm to help identify those patients most at risk of AF. Drawing on a combination of sex, age, hypertension, body mass index, prior cardiac accidents and other factors, this algorithm was validated in patient cohorts, but its accuracy was found to be suboptimal. “Our latest project, MMAF, aimed to solve this problem by identifying additional risk predictors,” Schnabel explains. The team benefited from newly available information, including data on the pathogenesis of AF. MMAF, which is funded by the European Research Council (ERC), specifically focused on the atria, the chambers of the heart that receive blood. The team considered electrical and structural differences related to sex and age, and combined all available information in more modern, machine learning algorithms. “We had electrocardiogram raw data available which reflect early electrical changes of the cardiac atria. We also had access to non-invasive imaging information from echocardiography and MRI data to better characterise subclinical changes of the atria. Finally, we used blood and tissue omics covering genetics, gene expression, proteomics, metabolomics of heart tissue and circulating biomarkers to identify new pathways,” says Schnabel.

Varying outcomes

The project has identified key differences in the incidence and impact of AF between men and women. “We could demonstrate that women generally have lower age-adjusted incidence and prevalence of AF compared to men,” adds Schnabel. “However, given the greater longevity of women, the absolute numbers are similar.” Major risk factors, meanwhile, are dependent on sex. Women have higher prevalence of hypertension and valvular heart disease, and lower prevalence of coronary heart disease compared to men. Higher body mass index and obesity carry a higher risk of AF in men, and, when it comes to symptoms, women are more likely to present atypical ones such as weakness and fatigue. Women have a longer duration of symptoms compared to men. Women also report worse quality of life, more frequent depression, and more risk of AF-related stroke, myocardial infarction and mortality. “This is why all risk prediction models needed to incorporate sex as a central variable,” says Schnabel. The project is now wrapped up, but the results are further validated in the EU-funded AFFECT-EU project. A consortium of 26 partners is currently developing a risk-based AF screening strategy using digital applications for rhythm monitoring to reduce the burden of stroke and other AF-related comorbidities in an ageing Europe. Schnabel and her team also plan to submit an ERC proof of concept grant application to assess the implementation and uptake of their new risk prediction algorithm and guide screening efforts. The algorithm, with an accuracy superior to the one previously available, has already been implemented in routine physician software used by general practitioners, internists and neurologists.

Keywords

MMAF, atrial fibrillation, prediction, algorithm, women, heart disease