Cardiovascular Risks of Exercise
Avoiding Sudden Death and Myocardial Infarction
Paul D. Thompson, MD
THE PHYSICIAN AND SPORTSMEDICINE - VOL 29 - NO.4 - APRIL 2021
In Brief: Sudden cardiac death and acute myocardial infarction are serious complications of exercise in adults. There is no consensus evidence that physicians can reduce the risk of such events. Coronary risk factor reduction and ensuring that patients can recognize prodromal symptoms are prudent. Preparticipation stress testing has been recommended but often does not reveal any abnormality in patients subsequently suffering an acute cardiac event. Personnel involved with sports should be able to recognize cardiac symptoms and be certified in cardiopulmonary resuscitation.
Exercise can provide adults with multiple cardiovascular benefits, but it also carries risks. Sudden cardiac death and acute myocardial infarction (MI) are the most important cardiovascular complications of exercise based on both frequency and seriousness.
Pathology of Exercise-Related Cardiac Events
Atherosclerotic vascular disease is the primary cause of almost all exercise-related deaths in adults. Ragosta et al (1) examined the causes of death in 80 men and 1 woman who died during or immediately after recreational exercise. Seventy-five (92.6%) of these deaths occurred in individuals older than 29, and all of these 75 deaths were related to atherosclerotic vascular disease. In addition to being the primary cause of exercise-related death, atherosclerotic coronary artery disease (CAD) is the cause of nearly all exercise-related MIs.
Plaque rupture. Sudden coronary death and acute MI among previously healthy adults in the general population usually stem from atherosclerotic plaque rupture with acute coronary thrombosis (2). Before this was widely accepted, Black et al (3) reported evidence of acute plaque rupture in 13 individuals who suffered sudden death or an acute MI during vigorous exertion ("Black's crack in the plaque"). Black believed that the plaque produced coronary spasm with subsequent infarction (Ascher Black, MD, personal communication, December 16, 120215), though recent evidence (2) tends to support plaque-rupture-induced thrombosis.
Others (4,5) have confirmed Black's original observation and have also demonstrated a high prevalence of intracoronary thrombosis in those who have exercise-related coronary events. Giri et al (6) compared coronary angiographic findings in 640 subjects who did (64) or did not (576) suffer an acute MI during or within 1 hour of vigorous exertion. The study demonstrated that coronary clots of more than 2 mm were present in 64% of the exercising patients but in only 35% of the comparison subjects, suggesting that different physiologic mechanisms cause MIs during exercise and at rest.
Burke et al (7) compared the coronary arteries of 25 men who died suddenly during exertion with the coronary arteries of 116 men who died at rest. Atherosclerotic plaque rupture was found in 68% of the exertion-related deaths but in only 23% of the nonexertion group. Coronary plaque rupture is presumably more frequent at "vulnerable" plaques, which are characterized by a thin, fibrous cap, macrophage infiltration of the cap, and a large, lipid-laden, necrotic core (7). The study found that men dying during exertion were not only more likely to die of plaque rupture but were also found to have more vulnerable plaques throughout the coronary tree than those who died at rest (1.6+1.5 vs 0.9+1.2).
Mechanism of injury. How can exercise lead to an acute MI or cardiac death? Exercise could induce arterial injury, worsen an existing injury, or increase the risk of thrombosis in a damaged arterial segment. Black et al (3) postulated that the increased "twisting and bending" of coronary arteries during vigorous exertion increased the frequency of plaque rupture. These motions of the arteries are exacerbated by the exercise-induced increases in heart rate and contractility. Exercise increases the excursion of the epicardial coronary arteries and dilates normal coronary arteries, but in abnormal arteries it can produce vasoconstriction in atherosclerotic segments (8). Such a spasm could induce plaque rupture over a thickened, noncompliant atherosclerotic plaque (9).
Exercise could also facilitate plaque disruption via chemical mechanisms. A 25-minute progressive exercise session to exhaustion in healthy young men has been shown to not only increase platelet aggregability but also to increase platelet-to-leukocyte aggregation and plasma elastase levels (10). Increased aggregation could enhance leukocyte adherence to the arterial wall and, in turn, facilitate leukocyte migration into the vessel wall where they could contribute to disruption of the fibrous cap (10). Elastase that is secreted by activated neutrophils can attack elastic fibers in the extracellular matrix and contribute to plaque disruption.
Exercise also has multiple prothrombic effects that could increase the risk of thrombosis over an injured arterial segment. Progressive exercise to maximal exertion on a cycle ergometer has been shown to increase platelet P-selectin expression and platelet-to-platelet aggregation (10), changes that contribute to the development of platelet thrombi. Exercise-induced platelet aggregation is greater in sedentary than in active subjects (11), possibly because exercise training reduces the catecholamine response to any absolute workload, and lower concentrations lessen an induced platelet aggregation.
Exercise might also induce acute events by deepening existing coronary fissures. Physical exertion increases systolic blood pressure, thereby increasing arterial shear forces and possibly increasing coronary fissuring. Plaque rupture without coronary thrombosis is common. Coronary plaque fissuring without thrombosis was found in 17% of people who died of noncoronary atherosclerosis and in 9% of those who died in motor vehicle accidents and from suicide (12). Vigorous exercise may induce coronary thrombosis by worsening previously damaged coronary plaques that were only mildly fissured.
These possible mechanisms for acute cardiac events apply primarily to previously asymptomatic subjects. Patients with known coronary heart disease who die during exertion may, but often do not, demonstrate evidence of an acute coronary lesion or recent myocardial injury (13). Patients without evidence of an acute coronary lesion often manifest evidence of infarction. The absence of any acute lesions in the coronary arteries of these patients suggests that such patients die of ventricular fibrillation originating from areas where myocardial scarring occurred.
Sudden Death During Exercise
The cardiovascular risk of exercise varies by population depending on the prevalence of cardiac abnormalities associated with exercise-related deaths. Since the most common cardiac abnormality and the most common cause of exercise-related complication is atherosclerotic CAD, deaths are more frequent in older populations and in those with known atherosclerosis. Nevertheless, even among adults, exercise complications are rare, and this rarity has limited efforts to determine the incidence of these events.
The Rhode Island and Seattle studies. Two of the most frequently cited estimates were based on only 10 exercise-related deaths in Rhode Island (14) and 9 cardiac arrest deaths in Seattle, but are among the only estimates based on the general population (15).
Thompson et al (14) used data on all deaths during jogging from 1975 through 120210 in Rhode Island. The prevalence of joggers was estimated from a random-digit telephone survey and state population estimates. Data revealed that 1 death occurred per year for every 7,620 male joggers between the ages of 30 and 65. If the patients with known CAD are eliminated from the sample and the assumption made that no other joggers had known CAD, the annual incidence of sudden death becomes 1 death per year for every 15,240 previously healthy joggers. Although this study had methodologic limitations (including survey method, self-reporting bias, and assumptions about CAD patient activity), the incidence for healthy subjects agrees with a subsequent estimate from Seattle (15) of 1 death for every 18,000 physically active men.
The Seattle study (15) also examined the incidence of death based on the individual's habitual activity level. In men who spent less than 20 min/wk doing activities that required more than 6 kcal/min, the relative risk of an exercise-related cardiac arrest was 56 times greater than at rest, but the relative risk was increased only fivefold in men who engaged in such activities for more than 140 min/wk. Thus, data demonstrate that regular exercise both reduces the chance of sudden death during exercise and also transiently increases the risk of cardiac arrest even among habitually active individuals.
This protective effect of habitual exercise may be overestimated when evaluated as the absolute hourly risk for an individual. The lower risk among the active subjects is based on risk per person-hours of exercise. Active men exercise more hours and therefore have a lower risk per hour of exercise than sedentary men. Although the calculated death rates are influenced by a small number of deaths, the annual risk of sudden death during exertion is 1 death per 17,000 in men who expend 1 to 19 min/wk in vigorous activities; 1 death in 23,000 for men who expend 20 to 139 min/wk; and 1 death per 13,000 men who expend more than 140 min/wk. This variation with activity level is considerably less than the 56-fold to fivefold decrease observed when the risk is expressed as the relative risk of exercise and rest.
Although the absolute death rate appears low, the death rate per hour of exercise seen in patients in both the Rhode Island and Seattle studies was increased over the resting rate (14,15). The relative risk of sudden death was seven times higher during jogging than during more sedentary activities in the Rhode Island study (14).
Interestingly, in both the Rhode Island and Seattle studies, the relative risk of an exercise-related sudden death versus other activities was greatest in the youngest adults. In Rhode Island, the relative risk of jogging versus more sedentary activities was 99 times greater than at rest in men 30 to 39 years old, 13 times greater for those 40 to 49 years old, and 5 times greater for those 50 to 59 years old. The number of deaths per hour of jogging varied from 1 per 482,600 hours for the youngest group to 1 per 309,400 hours for the oldest group, suggesting approximately equivalent hourly event rates. In Seattle, two thirds of the cardiac arrests occurred in men younger than 45. The younger men spent more time exercising, however, so that the incidences of cardiac arrest were actually similar to that of the older men (15,16). Thus, in young men, the high relative risk of death during vigorous exercise is partly due to their low risk of sudden death during sedentary activities. This supports the concept that vigorous exercise is indeed a serious provocateur of sudden cardiac death from CAD, especially among young subjects.
Although these studies (14,15) are not new, there are few more recent studies of exercise-related deaths in the general population. In addition, few, if any, published incidence figures exist in the medical literature for exercise-related sudden cardiac death in adult women. This probably reflects the delayed development of CAD in women, lower rates of vigorous physical activity among older women, and the lower incidence of sudden cardiac death in women in the general population.
Marathon study. If exercise reduces the absolute annual incidence of exercise-related sudden death, then the annual death rate should be considerably lower in middle-aged athletes compared with their inactive peers. Unfortunately, few studies have estimated that incidence among those subjects. Maron et al (17) calculated the frequency of cardiac arrest among 215,413 participants in the Marine Corps and Twin Cities Marathons from 1976 to 1994. During that span, 4 deaths occurred, or 1 death per 50,000 participants. This is about 1 death for every 215,000 hours of competition (assuming a marathon of 4 hours). This figure exceeds the incidence of sudden death among joggers in Rhode Island (1 death per 396,000 exercise hours) (14), and the incidence of cardiac arrest among the most active group in the Seattle study (1 arrest per 4,800,000 exercise hours) (15). Too few marathoner deaths occurred during the study period to make firm scientific conclusions, but the results do suggest that prolonged competitive physical activity may increase the exercise risk of a cardiovascular event.
Myocardial Infarction During Exercise
The risk of MI is also related to the individual's habitual fitness level. Mittleman et al (18) examined the relative risk of suffering MI during or within 1 hour of exercise requiring more than 6 METs. Only 54 (4.4%) of 1,228 MI patients experienced initial MI symptoms during or within 1 hour of exercise. The relative risk of MI during or soon after exercise was at least 5.6 times greater than the risk during less vigorous activity. Common activities associated with MI included lifting or pushing (18%), isotonic activities such as jogging (30%), and yard work such as gardening and chopping wood (52%). In patients who were usually sedentary, the relative risk of MI was 107 times higher during exercise than it was at rest. Among those who regularly exercised at least five times a week, the relative risk was only 2.7 times higher than it was at rest. Patients with diabetes had a relative risk of exercise-related MI that was 18.9 times higher than their risk at rest.
Vigorous vs moderate activity. Willich et al (19) compared the frequency of MIs during or within 1 hour of exercise requiring more than 6 METs with that during less vigorous activity. In a 2-year period, 1,194 patients agreed to participate. Of these, 69 patients (5.8%) had an exercise-associated MI. A control group consisted of individuals who described activities similar to when the patients had their MIs. The patient group was not restricted to patients with first-time MI, and consequently the MI patients had more prior MIs, had more cardiac risk factors, and used more cardiac medications than the controls. Even after adjustment for these differences, the relative risk of a patient having an MI during exercise was 2.1 times higher than at rest.
Angioplasty patients. Giri et al (6) stratified the relative risk of exercise-related MI by the subjects' habitual activity level among 640 patients who underwent primary angioplasty for their MI. Researchers found that 10% (64 of 640) of patients had MIs during or within 1 hour of exertion requiring more than 6 METs. The overall risk of MI during exertion was 10.1 times greater than it was at rest, but the risk varied with the patients' habitual activity level. Among the least active patients, the relative risk of an MI was 30.5 times higher during exercise than at rest, whereas the risk was not significantly higher during exercise than at rest among the most active men (relative risk, 1.2). These results confirm the theory that exercise increases the risk of MI, but suggest that such events occur primarily among habitually inactive individuals who perform unaccustomed physical activity.
Applicability to others. The previous studies (6,18,19) are not readily applicable to athletes or healthy patients because they include those with known heart disease. To obviate this problem, Giri et al (20) calculated the relative risk of MI during exercise versus rest in 547 previously healthy patients whose MI was treated by primary angioplasty. Of these patients, 66 (12%) suffered MIs during or within 1 hour after vigorous exertion. This population included some of the same patients from an earlier study (6). The risk of MI during exertion was 12 times higher than at rest, and the risk was greatest among the least active (relative risk, 55.1), and lowest among the most active men (relative risk, 2.9).
Exercise Risks in Patients With CAD
The best estimates on the risk of cardiac arrest or death for patients with CAD are from studies of cardiac rehabilitation programs. Unfortunately, the two primary studies were published in the mid-1970s (21) and 120210s (22) and might not be directly applicable to current patients, who are more likely to receive aggressive medical and interventional therapy. Haskell (21) queried 30 rehabilitation centers and reported 1 cardiac arrest for every 33,000 patient-hours of participation, 1 MI for every 223,000 patient-hours, and 1 death for every 116,000 patient-hours.
Van Camp and Peterson (22) surveyed 167 rehabilitation programs and reported 1 cardiac arrest for every 112,000 hours, 1 MI for every 294,000 hours, and 1 death for 784,000 hours of participation. The lower exercise event rates seen in this cardiac rehabilitation study may derive from improved patient selection and supervision and better treatment of underlying CAD. Both studies suggest that cardiac arrest is more frequent than MI in this patient group and also demonstrate that prompt treatment of patients who experience cardiac arrest during physical activity contributes to the overall low mortality rate of cardiac rehabilitation programs.
Reducing Adverse Exercise Events
Preventing cardiac-related events in adults ultimately requires preventing atherosclerotic cardiovascular disease from developing. Clearly, those who have the most risk factors have the greatest risk of developing cardiac problems in the general population, and this applies to preventing exercise-related cardiac events. Exercise-related MIs are shown to be more frequent in patients who smoke (6) or are hyperlipidemic (6), obese (6), diabetic (18), and least physically active (6,18).
Prodromal symptoms. Exercising adults must recognize cardiac prodromal symptoms such as chest discomfort and unexpected dyspnea so they can seek appropriate attention and avoid exertion if such symptoms appear. Among 13 individuals who died during or immediately after exercise from CAD, 6 had prodromal symptoms before the event (23). Noakes (24) summarized the clinical history of 36 marathoners who suffered sudden cardiac death or acute MI and found prodromal symptoms in 71% of the 28 cases in which information was available. In contrast, Ciampricotti et al (4) noted prodromal symptoms in 47% of sedentary men who developed acute coronary syndromes at rest, but in only 8% of athletes during exercise. Consequently, although prodromal symptoms can help identify individuals at risk for exercise-related events, symptoms are variable and may be less frequent among athletes, possibly because of the rapid progression of previously noncritical coronary lesions in active patients.
Stress testing. The issue of preprogram exercise stress testing is also controversial. The American College of Sports Medicine recommends that high-risk individuals undergo exercise stress testing before starting vigorous exercise (25) (see, "Risks of Exercise Stress Testing," below). The high-risk classification includes men older than 40 and women older than 50, individuals with more than one CAD risk factor, and those with known CAD. The American College of Cardiology and the American Heart Association considered the evidence for using screening exercise tests in formulating their Guidelines for Exercise Testing (26). A minority of the writing committee (including me) favored designating preprogram exercise testing as not useful. The majority classified such testing as not well-established by evidence or opinion but did not call such testing unjustified.
The committee's reluctance to endorse exercise testing in asymptomatic adults is that exercise testing is a poor predictor of the major cardiac complications (MI and sudden cardiac death) during exercise. A true positive exercise test requires the presence of a hemodynamically significant coronary obstruction, whereas acute coronary events often involve plaque rupture and thrombosis at the site of previously unobstructive atherosclerotic plaque (27).
Assessing stress tests. Studies evaluating the test's utility in ostensibly healthy populations support this hypothesis. McHenry et al (28) studied 916 Indiana state troopers who underwent maximal exercise tests, had repeat testing at intervals of 1 to 5 years, and were followed for a mean of 12.7 years. Only 61 men (6.7%) ever demonstrated a positive electrocardiographic response to exercise. Of these, 21 (34.4%) developed clinical signs of coronary disease (angina pectoris, 18; acute MI, 1; sudden cardiac death, 1). One additional asymptomatic man underwent coronary bypass surgery. Only 44 (5%) of the men with normal electrocardiographic responses subsequently developed clinical CAD (MI, 25; angina pectoris, 12; sudden death, 7). These data demonstrate that ostensibly healthy individuals with positive exercise tests have a sixfold higher chance of developing CAD (34% vs 5%), but that the predominant presentation is angina and not sudden cardiac death or MI. This is probably because positive tests identify individuals with hemodynamically significant coronary lesions that are often tolerated until the appearance of angina.
Screening only high-risk individuals produces similar results. In the Lipid Research Clinics Primary Prevention Trial (29), the predictive value of a positive exercise test for an acute exercise event was only 4%. Of the participants in the study, 62 men developed an exercise-related event (54 acute MIs and 8 sudden deaths). Only 11 of the 62 events occurred in men who had a positive exercise test. The authors concluded that routine exercise testing is not effective in preventing exercise-related acute cardiac events, even in high-risk patient populations. This conclusion, however, ignores the high rate of false-positive tests in asymptomatic active populations. Other techniques such as radionuclide or echocardiographic imaging would not greatly alter this conclusion but should reduce the incidence of false-positive ECG responses. (See "Avoiding Cardiac Events: What to Recommend?" below.)
The Current State of Affairs
Although cardiovascular complications of exercise are often ignored by public health officials, between 4% and 10% of MIs are associated with exercise (6,18). Thus, the cardiac risks of exertion remain a major concern for practitioners who often fear legal liability for not detecting cardiac abnormalities. More research is needed to determine the at-risk patients, groups for whom exercise risks outweigh the benefits, and the quantity and intensity of exercise required to reduce CAD. Although risk likely rises with exercise intensity, the medical literature suggests that CAD prevention benefits may accrue with moderate levels and provide benefits to more patients with less risk.
Risks of Exercise Stress Testing
The risks of exercise stress testing, like those of exercise itself, vary with the prevalence of disease in the population being tested. Exercise testing among healthy athletes has an extremely low risk, with no complications reported for exercise tests performed on 353,638 sportsmen (1). In contrast, 24 (9.1%) of 263 patients with known ventricular arrhythmias who underwent 1,377 maximal exercise tests experienced ventricular fibrillation, ventricular tachycardia, or bradycardia requiring cardioversion, intravenous medication, or closed-chest cardiac compression (2).
The complication rates from studies (3-9) suggest that there are approximately 1 myocardial infarction, 2 ventricular fibrillations or major arrhythmia, 0.3 deaths, and fewer than 3 hospitalizations for every 10,000 tests, or approximately 6 major complications per 10,000 tests. This low rate has prompted some clinicians to use nonphysician staff to administer the exercise stress tests (9), but as of January 1, 192021, physicians are required to be physically present at the exercise testing site (10).
Avoiding Cardiac Events: What to Recommend?
Reducing the risk of exercise-related cardiac events is extremely difficult because the events are rare and the predictive value of most screening procedures is poor. There are, however, prudent measures that should be followed even if they have unproven effectiveness.
Risks and warning signs. Adults should be cautioned to reduce coronary artery disease (CAD) risk factors and not to regard exercise as a panacea for CAD prevention. Active adults should also know the nature of cardiac prodromal symptoms (eg, chest discomfort, unexpected dyspnea, unusual fatigue) and their need for prompt medical attention. Active people or athletes who think they may have cardiac symptoms during exercise should undergo a careful cardiac evaluation before they return to training or competition.
The risk of exercise is underestimated in some patients with known disease. The public health emphasis on the benefits of exercise encourages some CAD patients to believe that they can cure their disease by vigorous exercise. These patients often ignore the aggressive management of other risk factors and may disregard the risk of vigorous exercise, sometimes with fatal consequences. All CAD patients should pursue an exercise training regimen, but most patients should be prohibited from extremely vigorous exertion and competition.
Cautions for physicians. It should be emphasized that cardiac discomfort is often not perceived as "pain," but as discomfort, tightness, or heartburn. Indeed, "gastrointestinal" symptoms are such a common precedent of cardiac disease that all gastrointestinal evaluations should include a stress test. Exercise stress testing is especially useful in evaluating nonspecific discomfort in active people. Excluding important cardiac disease in symptomatic athletes is probably the most efficient way to prevent exercise-related complications, since many athletes have premonitory symptoms before their final event.
Suggestions for coaches. Sporting event officials should learn and update their cardiopulmonary resuscitation skills. Coaches are most frequently present when athletes collapse, and, if properly trained, staff may be able to prevent exercise-related deaths. Competency in cardiac resuscitation should be a prerequisite to obtaining a coaching certificate.
This article was adapted from the recently published book: Thompson PD (ed): Exercise and Sports Cardiology, New York City, McGraw-Hill Medical Publishing, 2021 (to order: 1-800-262-4729 [ISBN:0-07-134773-9]).
Dr Thompson is the director of preventive cardiology in the Division of Cardiology at the Hartford Hospital in Hartford, Connecticut. Address correspondence to Paul D. Thompson, MD, 80 Seymour St, Hartford, CT 06102; e-mail to [email protected].