Results
During the 3-year study period, resuscitation was attempted in 3303 patients suffering an OHCA in the study area (Figure 1). After exclusion of 227 patients with an EMS witnessed arrest (per the Utstein recommendations), 410 patients with a non-cardiac arrest, 79 patients aged younger than 10 or older than 90 years, 11 patients whose age was unknown, and 52 patients without information regarding physical activity, a study population of 2524 OHCA cases remained. Of these, 143 (5.7%) were exercise related. Patient and resuscitation characteristics are shown in Table 1. Patients suffering an exercise-related OHCA were younger and more likely to be male. Exercise-related OHCAs occurred more frequently in public places, were more frequently witnessed, had higher rates of bystander CPR and AED use, and were more likely to have a shockable initial rhythm than non-exercise-related OHCAs. Ventricular fibrillation, rather than rapid ventricular tachycardia, was the predominant shockable initial rhythm (in 98% of exercise-related OHCAs and 96% of non-exercise-related OHCAs, respectively). Exercise-related OHCA occurred during or shortly after cycling (n = 49, 34.3%), tennis (n = 22, 15.4%), workouts at the gym (n = 16, 11.2%), swimming (n = 13, 9.1%), and miscellaneous other activities (n = 43, 30.1%).
The incidence of exercise-related OHCA was 2.1 per 100 000 person-years and was more than 10-fold higher in men (4.0) than in women (0.3) (Table 2). In persons aged 35 years or younger, seven exercise-related events (six men, one woman) occurred, resulting in an incidence of 0.3 per 100 000 person-years. In persons older than 35 years, the incidence of exercise-related arrest was 3.0 per 100 000 person-years. The incidence of non-exercise-related OHCA was 35.5 per 100 000 person-years and was almost three times higher in men (52.1) than in women (19.4) (Table 2). In persons aged 35 years or younger, the incidence was 2.8 per 100 000 person-years, again higher in men (3.9) than in women (1.7).
Almost half (46.2%) of the 143 persons suffering an exercise-related OHCA survived, compared with 17.2% in non-exercise-related OHCA (Table 3). Exercise-related OHCA survival was three times higher in persons >35 years compared with those aged ≤35 years (47.8 and 14.3%, respectively; P = 0.09). Survival after non-exercise-related OHCA was similar for persons aged ≤35 years and those aged >35 years (17.7 and 17.2%, respectively; P = 0.97). All survivors of exercise-related OHCA were neurologically intact, which was not the case for those surviving a non-exercise-related OHCA (P = 0.01).
Exercise-related OHCA, when compared with non-exercise-related OHCA, had higher survival rate (OR, 4.12; 95%CI, 2.92–5.82; P < 0.001).Given the interaction between age group and exercise (age 35 years or younger P = 0.43, age 36–50 years P = 0.045, age 51–65 years P = 0.156), an interaction term was included in the final multivariable model (Table 4 and Table 5). The association between exercise-related OHCA and survival to discharge remained statistically significant after adjustment for age, gender, location, bystander witness, initiation of bystander CPR, AED use, initial rhythm, and EMS response time (adjusted OR, 2.63; 95%CI 1.23–5.54; P = 0.01). Analyses using neurological intact survival as an endpoint yielded similar results (univariable analysis: OR 4.56, 95%CI 3.22–6.44; P < 0.001; multivariable analysis: OR 2.89, 95%CI 1.37–6.13; P = 0.005).
The results of stratified analyses for those with a shockable initial rhythm (to diminish confounding by time interval from collapse to emergency call) and according to age groups are presented in Table 5. The adjusted odds ratio for survival of those suffering an exercise-related OHCA was 3.13 (95%CI, 1.42–6.87; P < 0.01) among patients with a shockable initial rhythm (n = 1218 of whom 114 were exercise related) and 2.98 (95%CI, 1.29–6.91; P = 0.01) among patients with a shockable initial rhythm at public places (n = 507, of whom 113 were exercise related). There was no relation between exercise-related OHCA and non-shockable initial rhythm (n = 1240 of whom 22 were exercise related): 0 vs. 2.7% neurologically intact survival in exercise and non-exercise-related OHCA respectively.
There was no relation between exercise-related OHCA and survival in the small group of persons ≤35 years (n = 69 of whom seven were exercise related; crude OR 0.77, 95%CI 0.08–7.08; P = 0.82). Events in the seven exercise-related OHCA victims aged 35 years or younger occurred during: cycling, horse riding, martial arts, tennis, soccer, and swimming. A diagnosis was available (hypertrophic cardiomyopathy, HCM) in only one of those seven persons. Odds ratios for survival in exercise-related OHCA increased with higher age subgroups.
The majority of OHCA survivors underwent coronary revascularisation, with no difference in the rate of percutaneous coronary interventions (PCI) between those surviving an exercise-related OHCA and those whose OHCA was not exercise related (48.5 against 41.3%; P = 0.35). Coronary artery bypass grafting (CABG) was performed more often in exercise-related events (19.7 against 9.7%; P = 0.01). Implantation rates of cardioverter defibrillators (ICD) did not differ between survivors of exercise-related and non-exercise-related events (34.8 against 28.3%; P = 0.34).