Editors’ introduction: focus on cardiac arrhythmias

While the prevalence of atrial fibrillation (AF) is low in the general population, several risk prediction models have been proposed for selective screening of AF.1 A systematic review of the established models to predict prevalent AF was performed: 14 risk prediction models were identified and tested in 1.2 M individuals with CHA2DS2-VASc ≥2, where 10 464 (0.41%) participants had AF. For discrimination, six prediction models had area under the receiver operating characteristic curves >0.7. The two models with the highest observed prevalence in the highest decile of predicted risk, CHARGE-AF and MHS, showed an observed prevalence of AF of 1.6% with a number needed to screen of 63, thus demonstrating to reliably identify individuals at high risk. Atrial fibrillation commonly occurs in hypertrophic cardiomyopathy (HCM) patients and is a cause of significant morbidity and mortality due to haemodynamic decompensation and thromboembolic phenomena when unidentified and untreated.2 The importance of risk stratification and risk factor control has been recently advocated in this condition.3,4 In an here presented review,5 the data pertaining to significant multimodal risk factors for developing AF in HCM are presented: imaging assessment of left atrial volumetric and functional remodelling as well as electrocardiographic assessment of electrical remodelling using pre-specified cut-offs can be used to predict AF development. Obstructive sleep apnoea is common in HCM patients and has the potential when left untreated to greatly increase the risk for AF as well as overall morbidity and mortality and is potentially treatable: formal sleep evaluation is warranted. Machine learning is an application of artificial intelligence that provides systems with the ability to automatically learn and improve from experience without being explicitly programmed.6 The process of learning begins with observations or data, such as examples, direct experience, or instruction, in order to look for patterns in data and make better decisions in the future based on the examples that we provide. Novel but clinically useful applications of machine learning-based prediction algorithms may have a role in automated screening of a chosen population (e.g. a General Practitioner practice) to narrow the population who could benefit from screening for AF. Here an external validation of the performance of a previously developed AF risk prediction algorithm was performed on patients at highest risk of AF based on routinely collected patient data.7 This is one of the first to demonstrate the validity of a machine learning-based prediction model to assess the risk of AF. However, the performance of the algorithm in the wider real-world clinical setting is unknown. When the relationships of anger, vital exhaustion, anti-depressant use, and poor social ties with incident AF were investigated, vital exhaustion was found to be associated with an increased risk.8 Considering the overall prevalence of depressive disorders in European nations is estimated at 8.5%,9 establishing whether vital exhaustion or anti-depressant use increases AF risk, even if only modestly, could have significant implications. The effect size for the observed association between vital exhaustion and AF in this study was small and clinical relevance cannot be assumed. Further study in larger cohorts with more comprehensive and clinically validated assessments of depression is needed. Coffee is a plant-derived beverage widely consumed in Western countries, but its healthy benefits are still controversial.10,11 Here in two large Mediterranean cohorts’ meta-analysis (one of adults from a general population and another of elderly participants at high cardiovascular risk), light consumption of caffeinated coffee (1–7 cups/week) was associated with a reduction in AF incidence. Perhaps, more interesting, higher levels of caffeinated coffee consumption showed no association with AF. However, heterogeneity of the study population as well the fact that AF was self-reported lower the strength of this information. It is generally assumed that new-onset AF in young people is more often due to genetic predisposition as compared with elderly people. However, the relative contribution of non-genetic risk factors and genetic risk factors in young and old has not yet been determined. Here in a study population of almost 1 million individuals, obesity, male sex, heavy alcohol consumption, smoking, hypertension, diabetes, and chronic kidney disease were associated with an increased risk of new-onset AF among various age groups.12 In conclusion, except for sex, these non-genetic risk factors can be modifiable and efforts to control non-genetic risk factors might have relevance for both the young and old. The overall risk of being admitted to the hospital among patients with AF is ∼44% per year, which is 1.6 times higher than non-AF populations. It is well-established that AF patients have a high prevalence of cardiovascular risk factors and comorbidities, which is one reason why they face an increased risk of cardiovascular and non-cardiovascular adverse events.13 Here a Swiss research group has developed and tested a new risk score for all-cause hospital admissions in patients with AF, based on well-known risk factors and comorbidities, such as age, prior pulmonary vein isolation, hypertension, diabetes, coronary artery disease, prior stroke/transient ischaemic attack, heart failure, peripheral artery disease, cancer, renal failure, and previous falls. Admit-AF score achieved a C-index of 0.64 in the derivation and 0.59 in the validation cohort, confirming that a multi-factorial approach is essential to reduce the admission rate in this high-risk patient population.14,15 The CHA2DS2-VASc score is currently the primary tool for assessing the risk of embolic events in patients with AF.16 However, many risk markers are not limited to those mentioned in CHA2DS2-VASc score and, for example serum uric acid (SUA) levels, left atrial anteroposterior diameter, etc., have also been reported to be associated cardiovascular morbidities17 and left atrial thrombus (LAT). In a single centre, it is here demonstrated that high SUA level or having a history of gout/hyperuricaemia and obesity were risk markers for LAT. Further propensity score analysis found that abnormal uric acid was an independent risk marker for LAT and that the inclusion of this parameter in the CHA2DS2-VASc score increased the accuracy for LAT prediction.18 Although AF occurrence at advance ageing is not trivial, no specific trials have addressed the issue of anticoagulation in patients with AF and active cancer.19 Here a survey promoted by the Council of Cardio-Oncology of the European Society of Cardiology investigated the attitudes towards the prescription of anticoagulants among a small cohort of only 960 physicians (82.4% cardiologists, 75.5% from Europe). Management of anticoagulation in patients with AF and active cancer showed substantial heterogeneity in therapeutic choices, with direct oral anticoagulants having a leading role (62%) but still the use of LMWH remained substantial (24%), while warfarin was chosen by only 7.3%. Limited data exist regarding the incidence and aetiology of life-threatening events such as major cardiac events in general population20 or in athletes.21,22 Here, in the large prospective RACE PARIS registry, which recorded all life-threatening events/fatal events occurring during 46 marathons, and other long-distance races in the Paris area between 2006 and 2016, comprising 1 073 722 runners, the incidence of life-threatening events was 3.35 per 100 000: 1.02 per 100 000 exertional heat stroke and 2.33 per 100 000, major cardiac events, including 18 sudden cardiac arrests (1.67 per 100 000).23 A third of participants with ischaemia-related major cardiac events presented with pre-race clinical symptoms. Major cardiac events were more frequent in the case of a high pollution index. Case fatality was low (0.19 per 100 000). Similarly, a meta-analysis of eight long-distance race registries comprising 16 223 866 runners a low incidence of long-distance race-related sudden cardiac arrest (0.82 per 100 000) and fatality (0.39 per 100 000) is here reported. In conclusion, long-distance race-related life-threatening events remain rare although serious events. Better information for runners on the risk of pre-race clinical symptoms, outside air pollution and temperature may reduce their incidence. Conflict of interest: none declared. The opinions expressed in this article are not necessarily those of the Editors of the European Journal of Preventive Cardiology or of the European Society of Cardiology.