Sports Med 2007; 37 (4-5): 448-450CONFERENCE PAPER 0112-1642/07/0004-0448/$44.95/0

© 2007 Adis Data Information BV. All rights reserved.

Marathon Cardiac DeathsThe London Experience

Dan S. Tunstall Pedoe

London Marathon Medical Director 1981–2006, London, UK

Data from the London Marathon, with 650 000 completed runs, show thatAbstractcardiac arrests occur even in the most experienced runners. Although coronaryartery disease was the commonest cause of sudden cardiac arrest (SCA) with fivedeaths and six resuscitations, hypertrophic cardiomyopathy or idiopathic leftventricular hypertrophy (HCM) was diagnosed at autopsy on three occasions.HCM deaths had the same average age as the runners with ischaemic heart diseasewho had SCA or sudden cardiac death. The cardiac arrests were at the finish inless than one-third of cases and the remainder occurred between 6 and 26 miles onthe course. Only one of the eight runners who died had reported symptoms to hisfamily or physician suggestive of cardiac disease. The runner who had reportedpre-race angina pain was investigated with a negative exercise stress test prior tothe marathon and despite this died with a left anterior descending coronary arterystenosis. The cardiac death rate for the London Marathon is 1 in 80 000 finishers.

In the 1976 New York Academy of Sciences ported prevalence of 0.2% in the general popula-Conference on the marathon, Bassler[1] proposed tion.[2]

that marathon runners had immunity to atheroma- Rather than reporting data from a variety of inter-tous coronary artery disease. Unfortunately, this ap- national sources, with very different populations ofpears to be as mythical as Pheidippides running marathon runners and no autopsy data, this articlefrom Marathon to Athens and dying blissfully from will focus on the data from the 26 London Marathona ruptured heart as he announced, “Rejoice! We races, with 650 000 completed runs, showing thatconquer!” cardiac arrests can occur in even the most exper-

ienced runners. The London Marathon was estab-Collapse during and after marathon races is com-lished in 1981 with 7000 runners and now has fieldsmon, and although an uncommon cause, primaryof 35 000 finishers. The age distribution is such thatcardiac arrest from structural heart disease or cardi-the most popular age groups are 35–40 years andac arrhythmia must always be included in the differ-40–45 years. Women entries have gradually reachedential diagnosis. Deaths are most frequently from31% of the field but comprise only about one-sixthsudden cardiac arrest (SCA) associated with struc-of the cumulative total finishers. There have been notural heart disease and failed cardiac resuscitation.female deaths or cardiac arrests.Although deaths in athletes <30 years old are rare,

many causes of sudden cardiac death (SCD) have Data on deaths and major medical problems arebeen reported, with hypertrophic cardiomyopathy collected from the St John Ambulance Brigade and(HCM) being the most frequent.[2] In older athletes, the hospitals receiving casualties. Contact was madecardiac ischaemia (IHD) from atheromatous corona- with the coroner concerning all deaths and the au-ry artery disease is the most frequent cause and topsy was attended by the author for six cardiacHCM is only occasionally reported despite the re- deaths and by a colleague in the other two. Full post

London Marathon Cardiac Deaths 449

mortem reports (including histology) were received the cardiac resuscitations, three of six had run sever-on seven of eight deaths. Data on five of six resusci- al marathons; the most recent arrest was in atations is currently available. 50-year-old who was running his 19th marathon and

10th London Marathon. He had an emergency angi-Although coronary artery disease was the com-oplasty with stenting for a blocked LAD after amonest cause, with 11 SCA resulting in five deathscardiac arrest with successful resuscitation at 20and six resuscitations, HCM or idiopathic left ven-miles.tricular hypertrophy (LVH) was diagnosed at autop-

sy on three occasions. Runners with HCM deaths2. Post Mortem Findingshad the same average age as the runners who arrest-

ed with IHD. SCA occurred at the finish in less thanAll eight hearts from the SCD casualties showedone-third of cases and the remainder occurred be-

obvious cardiac pathology. Four of the IHD deathstween 6 and 26 miles on the course. Only one ofhad double or triple coronary artery vessel diseaseeight runners who died had reported symptoms towith evidence of previous myocardial infarction.his family or physician suggestive of cardiac dis-One heart, from the runner who had pre-race anginaease. The runner who had reported pre-race anginawith a negative exercise stress test, had LAD steno-pain symptoms had a negative exercise stress test 2sis of moderate severity only.months prior to the marathon and despite this died

The three HCM deaths showed severe LVH withwith a left anterior descending coronary arteryheavy hearts. One had debatable myocardial fibre(LAD) stenosis.disarray and could have been labelled idiopathicThere were no deaths in the first 9 years of theLVH or possible HCM. Another had obvious ob-race. Cardiac arrests with successful resuscitationsstructive hypertrophic cardiomyopathy with fibreoccurred in 1983, 1988, 1990, 1997, 1998 and 2006.disarray and the third had no histology performed,The first death occurred in 1990 and at autopsy wasbut the experienced pathologist was certain that theascribed to HCM (but no histology report was sub-diagnosis was HCM.sequently received). There were two additional

HCM deaths in 2001 and 2005. The death in 20053. Discussionwas initially attributed to idiopathic LVH and was

subsequently listed as HCM (i.e. severe, unex- Cardiac-related marathon deaths have been cal-plained LVH) after one cardiac pathologist de- culated to occur between 1 in 50–100 000 runs.[3] Inscribed fibre disarray, but this view was not shared the London Marathon, apart from the cardiac deaths,by another. The remaining deaths in 1993, 1995, there have been two fatal cases of subarachnoid1996, 1997 and 2003 were from IHD following haemorrhages whose initial symptoms occurred dur-cardiac arrest with unsuccessful resuscitation. Al- ing or shortly after the run. The overall death rate forthough the numbers are small, there was little differ- the London Marathon is therefore 1 in 65 000, with aence in the average age of those who died from rate of 1 in 80 000 for cardiac deaths. These deathHCM (47 years old), those who died from IHD (49 risks have been calculated on a time-of-exposureyears old) or those who were resuscitated (48 years basis to be less than riding a motorcycle and aboutold), nor in the average distance run before collapse 3–4 times greater than riding a bicycle (in Europe)by the HCM deaths (16 miles) and the IHD deaths for the same length of time.[4,5] The individual risk is(15.6 miles). therefore very low and comparable to a large num-

ber of daily activities.1. Cardiac Events inDeaths associated with marathons have been re-Experienced Runners

ported from all over the world sometimes occurringOne death from IHD occurred in a 47-year-old in clusters as in the 2006 Los Angeles Marathon

runner who was running his 38th marathon and (two deaths) and in the 2006 Great North Run halfanother in a 52-year-old who was running his 11th marathon in Newcastle, England, (four deaths). Thismarathon, was an ultra marathoner and was a mem- is a feature of a Poisson distribution, where an eventber of the 100km association ultra running club. Of that occurs rarely but has many opportunities to do

© 2007 Adis Data Information BV. All rights reserved. Sports Med 2007; 37 (4-5)

450 Tunstall Pedoe

so, will occasionally occur in a small cluster after any sort of entry filter, which is bound to be conten-many years with no occurrences. tious, circumvented, inefficient and may render the

marathon liable if the screening failed to prevent a3.1 Practical Procedures case of SCD.

Protocols for cardiac arrest should be agreed3.3 Should We Accept Marathon Deathsupon by the volunteers and staff before the race.as Inevitable?Although it is against the road racing rules, runners

occasionally pass their race entry number to anotherA marathon is not a health event, but training for

runner, so the race number is not a reliable means ofthe marathon promotes a healthy life style. Increas-

identification in all cases. The name and details ofing numbers of high risk runners are entering mara-

the dead runner should not be released until therethon races and the question is whether, despite the

has been formal identification by a relative or closehigher risk of deaths while running a marathon, the

friend, and the family has been informed and givenoverall trend is beneficial. For every marathon

permission to release the details. A runner whodeath, there may be thousands of runners who are

collapses is highly visible and the press may chasepostponing or preventing their heart disease. Com-

the story or be informed by people in the crowd,pleting a marathon is frequently the ultimate goal of

ambulance personnel, police or hospital staff. Thea jogging-running programme. Prohibiting selected

runner who has a cardiac arrest may well not berunners from competing and achieving this goal may

certified dead until he reaches hospital. The Britishprevent deaths during the marathon, but branding a

Press may use mobile phone scanners to listen intogenerally safe activity more dangerous than it really

conversations and pretend to be a relative makingis may be doing more harm than good.[5]

legitimate hospital enquiries about the status of therunner. We block access to the ‘downed’ runner’s

Acknowledgementsdetails on the marathon participant database andhave no discussion of the event on mobile phones or

The author has indicated that he has no affiliation orradios, but rather use land lines or personal contact. financial interest in any organisation (other than the LondonFacilities should be available for the race medical Marathon) that may have a direct interest in the subject matter

of this article.director to meet with the relatives of seriously ill ordeceased runners to break the news to them in aquiet and confidential environment. References

1. Bassler TJ. Marathon running and immunity to atherosclerosis.In: Milvy P, editor. The marathon: physiological, medical,3.2 Open Entry or Filtered Entry Throughepidemiological, and psychological studies. Ann N Y Acad SciMedical Screening? 1977; 301: 579-92

2. Maron BJ. Sudden death in young athletes. N Engl J Med 2003;The philosophy of the London Marathon from its 349 (11): 1064-75

3. Maron BJ, Poliac LC, Roberts WO. Risk for sudden cardiacinception has been open entry with the runner underdeath associated with marathon running. J Am Coll Cardiol

guidance from their own medical advisors and tak- 1996; 28 (2): 428-314. Tunstall Pedoe DS. Sudden cardiac death in sport: spectre oring full responsibility for being fit and well on race

preventable risk? Br J Sports Med 2000; 34 (2): 137-40day. We send all entrants medical advice, which 5. Tunstall Pedoe DS. Sudden death risk in older athletes: increas-outlines the risks and precautions.[6] To date, we ing the denominator. Br J Sports Med 2004; 38 (6): 671-2

6. Tunstall Pedoe DS. Morbidity and mortality in the Londonhave not taken the responsibility for excluding peo-Marathon. In: Tunstall Pedoe DS, editor. Marathon medicine.

ple on medical grounds. Many complete the mara- London: RSM Press, 2000: 197-2077. Tunstall Pedoe DS. Marathon myths and marathon medicine. In:thon with what physicians would regard as serious

Tunstall Pedoe DS, editor. Marathon medicine. London: RSMmedical problems.[7] Mass screenings of runners and Press, 2000: 3-14verifying medical certificates of fitness to run, in35 000 runners from all over the world is not practi- Correspondence: Dr Dan S. Tunstall Pedoe, 29 Meynell Cres-cal.[3,4] The onus is on each runner to take responsi- cent, London, E9 7AS, UK.bility for themselves and not for the race to initiate E-mail: danTPMARDR@aol.com

© 2007 Adis Data Information BV. All rights reserved. Sports Med 2007; 37 (4-5)