Article Text

Download PDFPDF

Natural and traumatic sports-related fatalities: a 10-year retrospective study
  1. E E Turk,
  2. A Riedel,
  3. K Püeschel
  1. Institute of Legal Medicine, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
  1. Dr E E Turk, Institute of Legal Medicine, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany; tuerk{at}


Objective: To identify the most common causes of death during sports activity in the population of Hamburg, Germany, and to identify which groups of sportspeople are particularly in danger of dying during sports and in which types of sport most fatalities occur.

Design and setting: We performed a 10-year autopsy-based retrospective study of all 48 335 fatalities in Hamburg and the surrounding areas that were subjected to police investigations between 1997 and 2006 and screened for sports-related deaths. The main outcome measure was cause of death depending on form of sport, age and risk factors.

Results: Most of the fatalities were male. In natural deaths, cardiac causes were the most frequent causes found, with running and football being the most frequent forms of sport in which death had occurred. In some of the cases, sports medical examinations had been performed before death, certifying eligibility for the respective activity. Traumatic deaths were found in all age groups, with younger age groups more likely to have traumatic than natural deaths, and as expected, occurred more commonly in “risky” outdoor activities.

Conclusions: Although exercise can have beneficial effects on health, fatalities related to sports activity occur. Cardiac disease is the main cause of sudden death from natural causes. In patients with pre-existing coronary heart disease, left ventricular hypertrophy constitutes a risk factor for exercise-related sudden death. Traumatic deaths often happen on holiday outside the person’s country of dwelling, and are most commonly attributable to drowning and blunt trauma. Preparticipation medical screening cannot always prevent fatal incidents during sports activity. Postmortem macroscopic and histological examination can clarify the cause of death and legal issues.

Statistics from

It is a widely held view among both professionals and the general public that regular exercise is beneficial for maintaining health, and many studies confirming these beneficial effects have been carried out. In particular, the risk of cardiovascular disease has been shown to decrease with regular exercise, and many studies provide evidence that physical fitness can lower cardiovascular and total mortality.19 However, the positive effects of exercise on cardiovascular health have also repeatedly been discussed critically, and it has been appreciated by many that physical activity can be hazardous as well as beneficial.6 1014 Vigorous athletic activity can trigger cardiac arrest and death in patients with pre-existing heart disease. In young, competitive athletes, exercise can act as a trigger for sudden cardiac death, especially if silent pre-existing heart disease is present.1518 Preparticipation screening protocols have been suggested for both competitive and leisure-time sports activity to prevent sudden cardiac death during exercise.1924

Apart from triggering cardiac events and death or other internal organ damage, sports activity always carries the risk of traumatic injury, sometimes with fatal outcomes, varying with the form of sport.2530 The recent forensic literature mainly deals with sudden cardiac deaths alone.31 32 Larger autopsy-based studies including all deaths related to sports activity are lacking in the recent literature.

We performed a 10-year retrospective study on deaths associated with sports activity, including natural and traumatic deaths, aiming to find the main causes of death and their correlation with the form of sport, epidemiological data and preparticipation screening programmes.


We reviewed the files of all 48 335 fatalities in Hamburg and the surrounding areas, which were subjected to police investigations between 1997 and 2006, and screened these for sports-related deaths. Sports-related deaths were defined as those cases that occurred in causal connection with the exercise, irrespective of the time between sports activity and death. Files and reports of the medicolegal autopsies performed at our institute were analysed for epidemiological data, cause of death and cardiovascular risk factors. The cases included in the study were divided into two major groups: deaths from natural causes and traumatic deaths.

Statistical analysis

Comparison of the distribution of traumatic and natural deaths inside and outside Germany was performed using a 2×2 contingency table and two-sided χ2 test. Age-group comparisons were performed using the two-sided non-parametric Mann–Whitney test.


In total, 176 deaths were identified as sports-related; none of the deceased were professional athletes. Of these, 89 (50.6%) cases had undergone medicolegal autopsy. In a further 59 (33.5%) cases, the cause of death could be established clinicially, as could be concluded from the doctors’ reports and death certificates. In the remaining 28 (15.9%) cases, the cause of death was not determined by autopsy or characteristic clinical symptoms; in these cases, the causes of death were taken from the death certificates.

Of the 176 deaths, 19 occurred on holidays outside Germany. There were 151 (85.8%) male and 25 (14.2%) female fatalities. Mean age was 41 years (range 4 to 88), with 86% being 30–79 years old (60–69: 28%, 40–49: 17%, 50–59: 17%, 70–79: 14%, 30–39: 10%).

Forms of sport during/related to which the fatalities occurred were swimming (n = 31), cycling (n = 19), running (n = 18), football (n = 17), tennis (n = 15), horse-riding (n = 10), table tennis (n = 7), skittles and mountaineering (n = 5 each), gliding and gymnastics (n = 4 each), handball, skiing and diving (n = 3 each), walking, kayaking, rowing, volleyball, golfing, inline skating, bowling, fitness training, parachuting and fishing (n = 2 each), and badminton, motor racing, Nordic walking, kite surfing, field hockey, wind surfing, ice hockey, power training, basketball, snorkelling, sailing and free climbing (n = 1 each).

The causes of death were divided into traumatic deaths and deaths from natural causes. Mean age was significantly (p<0.0001) lower in the trauma group (45.3 years) than in the natural death group (55.6 years; fig 1). The distribution of traumatic and non-traumatic deaths showed a strong correlation with the form of sport.

Figure 1 Age distribution of natural and traumatic deaths.

Deaths from natural causes

In total, 98 deaths were attributed to natural causes (55.7%). Most of these (n = 87) were a result of cardiac disease (88.8%). Of the non-cardiac natural deaths, recorded causes of death were two strokes, one hypertensive cerebral haemorrhage, one subarachnoid haemorrhage, one pulmonary embolism without prior trauma, one sepsis of unknown origin and one cerebral haemorrhage due to leukaemia. In the latter two cases, sports activity was not seen as the primary cause of the illness, but the physical strain caused by the sports activity was assessed to be a relevant factor contributing to the fatal outcome. In four cases, the cause of death was unknown, but a non-natural death could be excluded.

Of the cardiac deaths, coronary artery disease was by far the most frequent (n = 59; 67.8%) cause. In 30 of these cases, there had been a history of coronary heart disease, and four patients had been checked by a sports medical specialist before participation in the sports activity. The mean age of all coronary heart disease fatalities was 56.7 years (range 34 to 76). Coronary heart disease occurred less frequently in younger than in older patients. Cardiac risk factors were commonly present, but owing to the retrospective nature of the study, could not be elucidated or excluded in all cases. The most common cardiovascular risk factors were arterial hypertension (n = 21), hypercholesterinaemia as diagnosed by postmortem lipid analysis (n = 20), nicotine misuse (n = 5), diabetes (n = 2) and a family history of heart disease (n = 1). Most strikingly, all the people with fatal coronary artery disease found at autopsy (n = 38) had marked left ventricular hypertrophy (100%). Heart weight was determined in all cases and was between 425 and 910 g (left ventricular wall thickness between 1.5 and 3.0 cm). In 30 of these cases (78.9%), the heart weight was greater than the critical mark of 500 g. Four deaths were attributed to hypertrophic obstructive cardiomyopathy; the average age was 23 years (range 19–32). Marked left ventricular hypertrophy was present in three of these cases (heart weight 515–560 g), and histology was positive in three cases. In one case, a young marathon runner, the heart weighed only 370 g and the diagnosis of hypertrophic obstructive cardiomyopathy was made by genetic testing (β-myosin heavy-chain mutation). Cardiac tamponade due to a ruptured aortic dissection was found in four cases (mean age 60 years). Three deaths each could be attributed to myocarditis (all determined by histology, mean age 37 years) and acute decompensated chronic left ventricular failure (mean age 63 years). Dilated cardiomyopathy (mean age 32 years) and left ventricular failure due to aortic valve stenosis (mean age 61 years) were the cause of death in two cases each. In 10 cases, the cause of death was not known but was attributed to cardiac failure because of the history and symptoms at the moment of death.

Although swimming was the form of sport with the largest total number of fatalities, only two deaths were attributed to cardiac failure alone. In contrast, very high numbers of cardiac deaths were found in football (15/17 cases), tennis (14/15), running (13/19), table tennis (7/7) and skittles (4/5). Of these, football was the sport with the lowest mean age (41.5 years, range 11–68), followed by running (mean age 45.2 years, range 19–73). In tennis (63.9 years), table tennis (66 years) and skittles (68.6 years), the mean age was markedly higher. All cardiac deaths in tennis, table tennis and skittles were from coronary heart disease.

Trauma deaths

In total, 67 of all deaths were from traumatic causes (38.1%). Interestingly, of those deaths that occurred in Germany, only 21% were traumatic, whereas 79% of all sports fatalities outside Germany could be attributed to traumatic causes (fig 2). The difference in the distribution of traumatic and natural deaths inside and outside Germany was highly significant (p = 1.9×10−8)

Figure 2 Relative numbers of natural and traumatic deaths within and outside Germany.

Most traumatic deaths occurred in swimming (n = 21; 67.7% of all swimming fatalities), horse-riding (n = 10; 100%), cycling (n = 10; 52.6%), mountaineering (n = 5; 100%), gliding (n = 4; 100%) and skiing (n = 3; 100%). The most common causes of death were drowning (n = 29), four of which were under the influence of alcohol, blunt head trauma19 and polytrauma.11 In three cases, death was due to sepsis after traumatic injury, two patients developed pulmonary thromboembolism after leg fractures and three patients died of haemorrhagic shock after a fall from a horse. One patient died from brainstem infarction after basilar artery dissection due to cervical spine fracture following a head-first dive into shallow water.

The youngest traumatic fatalities occurred in swimming (age range 5–84 years, four of the fatalities aged < 12 years) and horse-riding (mean age 30.6 years, age range 4–64; four died aged ⩾17 years). Three young children went under water suddenly during their swimming lessons. Helmets were only worn in three of the horseback-riding fatalities. The other seven deaths from falls from horses were a result of blunt head trauma.

Combination of trauma and natural death

Ten deaths were attributed to a combination of natural disease and trauma. All these deaths occurred in the water, where the mechanism was internal organ failure with subsequent drowning (swimming n = 8, diving and fishing n = 1 each). In nine cases, the underlying natural disease was of a cardiac nature. In the tenth case, the patient had epilepsy; toxicology revealed subtherapeutic concentrations of antiepileptic drugs, pointing to a lack of treatment compliance as a possible mechanism.

A summary of all natural, traumatic and combined deaths is shown in table 1.

Table 1 Summary of all natural, traumatic and combined deaths


In all trauma deaths and in 43 of the natural deaths that were examined by medicolegal autopsy, toxicology screening for a large number of prescribed and illicit drugs and for alcohol was performed. Blood was positive for alcohol in foun drowning cases (0.7–1.3%). Diuretics and sympathomimetic drugs, which are sometimes used to increase performance, were not found in any of the cases. Screening for steroids or erythropoietin was not performed in any of the cases, owing to a lack of suspicion. Based on the clinical histories and autopsy results, none of the cases analysed in the present study was classified as drug-related.


Deaths due to natural causes

In line with previous studies, we found that most of the exercise-related natural deaths involved men and that cardiac illness was the main cause of death, especially in adult athletes.3335 In people aged ⩽35 years, we found that the leading cardiac causes of death were congenital abnormalities, a finding that also been shown previously.15 36 37 Congenital atrioventricular conduction anomalies were not found in any of the cases investigated. However, these anomalies generally escape detection at autopsy,38 39 and thorough histological examination of the cardiac conduction system was performed only in cases with suspected myocarditis.

Right ventricular cardiomyopathy has previously been found to be a common cause of sudden death in competitive athletes.16 In the present study, none of the deaths was attributed to this disease. However, the morphological diagnosis of right ventricular cardiomyopathy is difficult, especially if no clinical data are available with which to correlate the findings. Furthermore, there might be a regional difference in the incidence of right ventricular cardiomyopathy.34

The forms of sport that most often resulted in cardiac death in the present investigation were found to be football, running, tennis, table tennis and skittles. This might be explained by three different phenomena. Firstly, in tennis, table tennis and skittles, the deceased were older and had a high rate of pre-existing cardiac disease, which made them more likely to have an acute cardiac event in the first place. Secondly, vigorous exercise has been shown to be especially risky for cardiac death.6 40 Thirdly, the death rate has bseen shown to increase with the duration of exercise.41 42 Football demands both long and intense exertion from the athlete, which might account for the particularly high number of cardiac deaths, and running (or jogging) is often associated with particularly long exercise duration.

A very important aspect of sudden cardiac death in athletes is the legal issue for the doctor who has performed a preparticipation medical screening and certified eligibility for sports activity.46 In the present study, owing to its retrospective design, we had only very poor information on the patients’ previous medical history in most cases. In only four cases did we know that there had been medical consultation before the exercise. All the doctors had to face investigation for causing the athletes’ death by culpable negligence. In the case of a teenage marathon runner who collapsed due to hypertrophic cardiomyopathy 200 m before the finish line, the general practitioner had performed only a routine physical examination without auscultation of the heart, and a resting ECG had shown no abnormalities. Another young marathon runner who collapsed due to myocarditis after 41 km had performed very well when given a stress ECG, which showed no pathological changes at 310 W. Two patients with known coronary heart disease had undergone a pre-exercise sports medical examination including exercise stress testing, which showed ST segment depressions interpreted as “non-specific” in one case and normal in the other. Both died of myocardial infarction during the exercise. Exercise stress testing did not lead to prohibition from sports activity in any of the cases, suggesting that, as previously reported, it seems to have a relatively low sensitivity even in patients with pre-existing coronary pathology, although in hindsight the ST segment changes in one case could have been seen as a prohibition criterion. Although exercise stress testing is suggested as a routine screening method for certain groups of athletes by some authors,24 43 there is controversy about its use.39 44 45

Comprehensive recommendations exist for preparticipation screening programmes, especially for competitive athletes but also for recreational sports activity, which cannot be discussed here.21 22 4752 It has to be stressed at this point that preparticipation medical screening is an important, invaluable means of preventing sports-related cardiac fatalities and it is not the aim of the present investigation to question this. We found only four cases with fatal outcome despite preparticipation screening, and there are probably countless athletes every day who do not die during sports activity because they have received sufficient medical screening.

We would, however, like to emphasise one particuar point, as preparticipation screening programmes should constantly be improved. As we found marked left ventricular hypertrophy in all the people with pre-existing coronary artery disease (suggesting that left ventricular hypertrophy is a risk factor for cardiac events during exercise, particularly in patients with pre-existing coronary conditions), echocardiography should be performed in all patients with coronary artery disease who want to participate in sports activity, not only to screen for left ventricular function and wall motion abnormalities as recommended previously,52 but also for ventricular wall thickness, and patients with marked left ventricular hypertrophy should be disqualified from vigorous exercise and cardiac stress. Furthermore, in asymptomatic younger individuals, it might be reasonable to perform lipid diagnostics to exclude the possibility of silent coronary pathology, as this is a cheap and simple method, and if a pathological lipid pattern is found, further diagnostic tests could be used. Larger, prospective studies will have to be performed to clarify these issues.

Traumatic deaths

Strikingly, the relative amount of traumatic deaths was significantly higher in fatalities that occurred outside Germany (79%) than in those which occurred inside Germany (21%). This phenomenon might be explained by a greater readiness to take risks when on holiday. Another aspect might be that certain “risky” forms of sport can simply not be performed in the Hamburg area (such as mountaineering, kite surfing, skiing and swimming under dangerous circumstances).

Three swimming deaths were young children who went under water during their swimming lessons. This indicates that in many cases, supervision was insufficient and that with better supervision, such deaths may be prevented in the future.

The message from cycling and horse-riding accidents with the leading cause of death being blunt head trauma (15/20 cases = 0.75%) is that wearing a helmet might protect from fatal incidents in these activities. Currently, German law does not decree the wearing of helmets when cycling or riding. Given our observations, such a law might be beneficial to prevent fatal trauma, at least in children.

What is already known on this topic

  • Exercise undoubtedly has beneficial effects on health, but sports activity can also have adverse effects on the athlete’s health and might even result in death from natural or traumatic causes.

  • Most sudden, sports-related deaths are the result of cardiac disease.

  • Preparticipation medical screening is an invaluable means of preventing sports-related fatalities.

What this study adds

  • Left ventricular hypertrophy can be regarded as a risk factor for fatal cardiac events during sports activity in patients with pre-existing coronary heart disease.

  • Non-natural, traumatic sports-related deaths are more common on holidays outside the country, with drowning and blunt trauma being the most common causes of death.

  • Although preparticipation medical screening is important in the prevention of sports-related fatalities, it cannot always prevent fatal incidents during sports activity and the doctors involved might face legal issues.


Although exercise can have beneficial effects on health, fatalities occur related to sports activity. Cardiac disease is the major cause of sudden death from natural causes. In patients with pre-existing coronary heart disease, left ventricular hypertrophy constitutes a risk factor for exercise-related sudden death. Traumatic deaths often happen on holidays outside the country and are most commonly attributable to drowning and blunt trauma. Preparticipation medical screening cannot always prevent fatal incidents during sports activity. Postmortem macroscopic and histological examination can clarify the cause of death and legal issues.


We thank Professor Dr H P Beck-Bornholdt for statistical analysis.



  • Funding: No funding was taken up for this publication. Ethics committee approval was not necessary as it was a retrospective study. All data have been made anonymous so individuals cannot be recognised.

  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles