Discussion
The purpose of the current study was to determine whether exercise engagement prior to STEMI is related to lower post-MI peak cardiac biomarker concentrations following STEMI. We found that only 16% of the study population had engaged in any form or type of exercise in the week prior to STEMI onset. Second, those who engaged in exercise in the week prior to STEMI demonstrated significantly lower peak concentrations of cardiac biomarkers following STEMI compared with their physically inactive peers. Although explorative and being underpowered, these effects did not translate to differences in clinical outcomes between STEMI patients engaged in exercise versus their physically inactive peers at 30 days or 6 months. Taken together, these data provide further support for the beneficial effects of regular exercise training, potentially leading to lower post-MI cardiac biomarker peak concentrations.
In our population of STEMI patients, only 16% performed exercise in the week prior MI. Although the number of patients who performed exercise prior MI is small, a low level of physical activity is typical for this population.4 14 However, it is remarkable that those individuals who performed exercise in the week preceding STEMI demonstrated significantly lower STEMI-induced cardiac biomarker peak concentrations of both hs-cTnT and CK. Previous studies demonstrated that hs-cTnT is a cardiac-specific protein, which has shown to increase rapidly following MI, whereas CK is a non-cardiac-specific enzyme but increases following MI as well.15 Research has shown that concentrations of hs-cTnT and CK reasonably relate to infarct size, with smaller infarct sizes being related to lower peak concentrations.15 16 18 Even though the use of cardiac MRI is the golden standard to appropriately determine infarct size, it is not widely available, expensive and contraindicated in a substantial proportion of patients. In addition, the benefits of exercise on STEMI-induced cardiac biomarker concentrations are possibly in line with clinical outcomes from other studies. For example, the Copenhagen study showed that exercise, even in small amounts, is associated with lower risk for mortality following MI,13 and a previous study revealed that lifelong athletes show a better left ventricular systolic function following MI compared with sedentary controls.19 Therefore, our data could provide a first insight into the suggestion that regular exercise is associated with a smaller infarct size. Altogether, the proportion of STEMI patients who engage in exercise is small, but these patients demonstrate lower peak hs-cTnT and CK concentrations than their non-exercising peers.
Comparing both groups, no significant differences between the control group and exercise group were found regarding disease characteristics and patient characteristics. Although statistically not significant, comorbidities were not equally distributed across subgroups, for example, four patients had kidney failure in the control group verus zero patients in the exercise group. Kidney failure relates to higher post-MI cardiac biomarkers, but excluding individuals with kidney failure did not alter the main findings of our study. In addition, even though it was not significant, time from symptom onset to hospital admission was two times higher in the exercise group compared with the control group. This may underestimate the protective effect of exercise on cardiac biomarker release. Indeed, in patients undergoing an STEMI, ischaemia will occur until revascularisation, and cardiac biomarker concentrations will continue to rise.20 Despite the longer ischaemic period, the peak cardiac biomarker concentrations remains lower in the exercise group. Considering the points mentioned above, it seems unlikely that any of these patient-related or MI-related characteristics underlie the differences in cardiac biomarker concentrations post-MI between the exercise and control group. This raises questions about other potential explanations for the difference in cardiac biomarker concentrations between both groups. Exercise may have a direct effect to attenuate myocardial infarct severity.21 22 Studies have demonstrated that exercise can induce a functional or structural change in coronary arteries, potentially via collateral formation. This mechanism is a possible explanation of the protective effect of exercise training on both the occurrence and severity of MI.21 23 24 In addition, studies have demonstrated that exercise also has short-term protective effects. For example, studies in animals revealed that single bouts of exercise prior to planned cardiac ischaemic-reperfusion injury significantly attenuate cardiac injury. Similarly, also in humans, short-term exercise has demonstrated to reduce injury to vascular and cardiac tissue.10 25–27
Although peak concentrations of cardiac biomarkers were significantly lower in the exercise group, we found that these effects did not translate to differences in the clinical course between groups. For example, we found no difference in duration of hospitalisation between the groups. Duration of hospitalisation was typically 3 to 4 days with very little variation, which is mainly depending on the clinical course of the patient. We also found no difference between groups in 6-month MACE, which may relate to the relatively low number of events within this time-frame (n=12), with a majority of these MACE occurring during hospitalisation (n=9). Accordingly, statistical power is too low to provide meaningful insights into the potential impact of exercise engagement prior to STEMI on clinical outcomes. Nonetheless, previous studies show that patients who engage in exercise prior MI show lower long-term MACE incidence compared with inactive peers. Maessen et al compared 18 veteran post-MI athletes with 18 sedentary post-MI controls across 3–10 years. While veteran athletes did not report any secondary events, their sedentary peers reported eight events (six elective PCIs, two recurrent MIs).28 Exercise prior to MI is also associated with lower risk of postinfarct mortality.13 14 On the contrary, physical activity can also act as a trigger for MI, with patients that exercise infrequently particularly at risk.29 30 Given the low number of patients with exercise directly prior MI in our study (n=1 <24 hours prior to MI, n=0 during MI), we do not expect that MIs in our exercise cohort were triggered by exercise. However, a previous study showed that if MI was triggered by exercise, in-hospital and 1-year mortality were significantly lower compared with MIs not triggered by exercise.29 31 At least, our observation of lower cardiac biomarker concentrations, among patients who engaged in exercise prior to STEMI, aligns with observations from other studies, which report that exercise prior to STEMI is associated with lower mortality and morbidity following STEMI.
Limitations
Some limitations must be acknowledged. First, we have used questionnaires to gain insight into the physical activity pattern of the patients. To minimise bias, we used the validated SQUASH questionnaire.17 To further counteract the subjective nature and recall bias, the questionnaire specifically relates to the physical activity pattern in the week preceding STEMI-onset only rather than asking for the typical engagement in exercise training across the past months. It must be considered that although we found a beneficial effect of physical activity on cardiac biomarker concentrations post-MI, it is unclear whether the lower post-MI cardiac biomarker concentrations in the exercise group are causally linked to an acute exercise bout or to higher lifelong exercise levels. Therefore, more longitudinal studies are recommended. Another limitation in this study is that we derived cardiac biomarkers from electronic patient files, leading to a relatively limited number of time points. This minimises the temporal resolution, so we might underestimate the true maximum of hs-cTnT and CK concentrations. Nonetheless, this approach was adopted for both groups and underestimation was probably larger in the control group compared with the exercise group. Additionally, we performed additional analyses exploring mean values and changes from lowest to highest cardiac biomarker levels to see whether taking baseline measurements per individual into account yields the same results (online supplemental appendix 1). These additional analyses confirm our initial observations. Therefore, we believe that this approach cannot explain the difference in effect size between both groups. finally, an important limitation is the relatively small group of patients, with especially a low number of patients in the exercise group. Therefore, we must be careful in drawing conclusions. However, despite the relatively small patient number and the suboptimal measure of infarct size, this study does provide grounds to investigate the effect of exercise on the severity of MI in a larger format with more accurate infarct size assessments (such as MRI) and longer follow-up.