Screening for cardiac and respiratory problems in elite sport – compare and contrast

J. Hull,J. Rawlins
DOI: https://doi.org/10.1080/17476348.2016.1193437
2016-06-08
Expert Review of Respiratory Medicine
Abstract:The sudden death of a young athletic individual while participating in a sporting activity is a devastating event. Such occurrences have an immediate and profound impact on all involved, including family, team members, and coaching staff. This has driven a number of European countries [1], professional sporting organizations [2,3], and international cardiac societies [4–6] to recommend integration of screening for conditions that may predispose to sudden cardiac death (SCD), in pre-participation medical assessment. Indeed, although not universally accepted, this approach is promoted by a number of international sporting federations, across sporting disciplines [2,3,7] with a persuasive emphasis placed on the theory that detecting anomalies mitigates risk. The success of any screening policy is dependent upon rigorous application of a series of widely accepted criteria [8]. For a screening program to be effective, it must identify a condition/disease state accurately and in a timely fashion, to allow initiation of a cost-effective intervention that has a beneficial impact on prognosis. In this respect, the process of screening athletes for cardiac anomalies presents a number of challenges. First, a wide number of rare conditions have been associated with exerciseassociated SCD [9]. For example, the estimated prevalence of hypertrophic cardiomyopathy, recognized as the commonest cause of SCD during exercise under 35 years of age, is just 1 in 500 [6,10]. Second, there is currently no single test, sufficient to identify all relevant conditions. Finally, intervention varies with diagnosis, but the relative risk and impact on prognosis is often uncertain. In general, any treatment begins with a recommendation to cease sporting activity. In contrast, we now know that the prevalence of respiratory disease (specifically airway dysfunction) in elite athletes is extremely high [11], when compared to the general population. Indeed, asthma ± exercise-induced bronchoconstriction (EIB) is the most common chronic medical condition encountered in high-level athletes [12] and has often been identified in up to half of the members of large elite endurance teams [13]. Often considered a benign condition, asthma in young athletes is in fact associated with a recognized mortality. Specifically, Becker et al. found a high proportion of asthmarelated deaths occur in elite or competitive athletes [14]. In addition, untreated airways disease in athletes has a potential detrimental impact on both their health (i.e. risk of exacerbations) and performance (i.e. limiting ability to both train and compete). It is incorrect to assume that athletes with respiratory disease will always have symptoms and thus ‘self-present’. Intriguingly, the relationship between the presence of classical airway-centric symptoms (e.g. cough, wheeze, and dyspnea) and objective evidence of EIB in elite-athletic populations is actually weak [15,16]. Moreover, when athletic cohorts are ‘screened’, with surrogates of exercise (e.g. with the indirect bronchoprovocation test, eucapnic voluntary hyperpnoea [EVH]), there is almost always a ‘hidden’ prevalence signal; i.e. individuals with evidence of airway hyperreactivity but no or relatively few symptoms. Indeed in a study evaluating EIB screening in the Great British Olympic team, the prevalence of EIB was approximately 20% with seven individuals identified having no prior diagnosis of asthma [13]. Hence, when screening for occult respiratory disease is compared to the approach adopted, and often accepted, for the detection of cardiac anomalies in athletes [2–5], it seems intuitive to argue that consideration should also be given to commensurate respiratory screening. A key consideration in any screening program concerns the test or diagnostic criteria that are employed. Any screening methodology should be simple, reproducible, and effective at detecting the target abnormality [8,17]. In cardiac screening, due to the wide number of potential cardiac diagnoses possible, no single test is sufficient. Current European Society of Cardiology recommendations rely upon the identification of subtle nonspecific abnormalities on clinical history, examination, and 12-lead ECG, initiating further more complex diagnostic investigations [2]. This approach differs from the current recommended by the AHA that relies purely on a clinical history and examination [18]. In addition, certain sporting organizations also mandate regular transthoracic echocardiography [5]. Interpretation can, therefore, be challenging, and should be undertaken by individuals with particular experience and expertise in this area.
What problem does this paper attempt to address?