London Marathon: incidence of injury, illness and death

The London Marathon has been a major participant event since it was first run by 6,500 people, mainly novices, in 1981. It now has 32,000 finishers and is the biggest marathon in the world. This article looks at the extent of illness, injury and death associated with the Marathon and the various factors involved.

Entry to the Marathon is open to anyone over 18. Since 1Marathon runners981, the charity element has expanded significantly, with many participants taking part not because they are traditional runners, but because they have been sponsored by friends and colleagues to raise money for a charity. A survey of runners and charities revealed that more than 75% of participants in the 2002 Marathon were raising money for charity and between them succeeded in raising £32m.

Entrants are sent a medical advice sheet, which gives them responsibility for being fit and well on the day of the race. It suggests they discuss any medical problems with their GP and don’t participate without their agreement. It also suggests they surrender their entry if they cannot run 15 miles comfortably one month before the event. Runners who take advantage of this ‘sick, lame and lazy’ option are guaranteed entry the following year. This medical advice has been widely copied by other races.

First aid services on the day are provided by St John Ambulance, who set out more than 40 first aid posts along the route and at the finish, and two field hospitals at the finish. One of these hospitals has an ‘intensive care unit’ for more serious collapses, but intravenous fluids may be given at other sites, if necessary.

There is a much larger first aid post in the Isle of Dogs, two thirds of the way round the course. There are also cardiac units at the finish and resuscitation facilities along the course and at the finish.

In all, more than 1,000 St John staff volunteer to work on the day, together with other doctors, physiotherapists and podiatrists with an interest in sports medicine, who are recruited to work closely with St John, mainly at the finish. Local ‘receiving hospitals’ are pre-warned about the race and receive written advice from the race Medical Director during the preceding week, with St John liaison officers posted to their accident and emergency departments on the day.

A runner who makes contact with first aiders during the race is logged as a ‘casualty contact’, with diagnosis made by first aid staff, unless the condition requires physiotherapy, medical or podiatric treatment. Each first aid station reports the number of casualties and the primary diagnosis for each.

To minimise lurid newspaper headlines about Marathon casualties, these contacts are divided into categories which clarify the seriousness of the various conditions involved. These include:

  • Social contacts – who stop and ask for such help as a drink, a shoelace or a dressing to treat themselves;
  • Musculoskeletal contacts – with cramps or painful joints, bones or muscles;
  • Topical contacts – with blisters, abrasions, runner’s nipple, skin chafing or subungual haematomas (blood clots under the toenails);
  • Constitutional contacts – who collapse, have chest or abdominal pain, diarrhoea, fits, vomiting etc.

The St John Ambulance reports are supplemented by enquiries to the designated receiving hospitals, which are asked to flag up all Marathon accident and emergency cases.

In 2000, when 32,600 runners completed the race, 4,633 St John Ambulance and 38 hospital contacts were recorded. By comparison, in 1987, when 19,970 runners completed the race, there were 4,984 St John Ambulance contacts and 10 hospital contacts. Totals for the 20 years show a hospital contact rate of 0.13% (one in 787). Hospital admissions are roughly 10% of the hospital contacts, but are increasingly difficult to define, as runners may spend many hours in accident and emergency.

Only those deaths, or collapses leading to deaths, that occur during the Marathon or within the finish area of the race, are considered Marathon deaths. Seven cardiac deaths have been reported in the London Marathon: five from severe coronary heart disease – in 1991, 1994, 1995 1997 and 2003 – and two with hypertrophic cardiomyopathy (HCM, a chronic disorder affecting the heart muscle) – in 1990 and 2001. Five successful cardiac resuscitations have taken place (in 1983, 1988, 1990, 1997 and 1998); all patients had coronary heart disease and were subsequently discharged from hospital. In the millennium race, a young man collapsed at the finish complaining of neck pain and died the following day in hospital following a diagnosis of subarachnoid (brain) haemorrhage.

The overall mortality rate from the 20 years is one in 67,414, or roughly one death for every two million miles run.

The Medical Director of the race is updated on the casualty contact numbers and the numbers taken to hospital by St John Ambulance. The director can check the more serious medical problems at hospitals designated to receive casualties from the race. However, unless specifically notified, he will not be aware of casualties who bypass the race casualty control system and go to other hospitals, or of those who arrive at the designated hospitals later in the day without wearing running clothes and a race number, and thus are not recorded as race casualties.

Graph: finishers and St John contacts 1981-2003

The more attractive, obvious and frequent the aid points are in a marathon, the more likely a tired, cramping, blistered runner will make a ‘pit stop’ and become a marathon medical statistic. Some races offer psychotherapists at the start and massage therapists at the finish, which increases the potential for collecting casualty numbers. Definition of an injury and the numbers are, therefore, contentious.

Most runners suffer from minor injuries such as cramps, blisters, skin chafing and subungual haematomas. The medical staff may be unaware of many of these injuries, especially as the successful runners are euphoric, anxious to go home and usually convinced that they can handle the problems themselves. The staff lose contact with participants immediately after the race as the runners disperse across the UK and to several other countries, taking their non-immediate medical problems to a multiplicity of doctors and physiotherapists. This makes a survey of the impact of the Marathon on medical providers even more difficult than a questionnaire to runners.

The sports medicine definition of an injury as something that prevents training for a defined number of days is impossible to apply when severe muscle stiffness is almost universal and full training may not be part of the runner’s post-race agenda. Aches and pains and severe delayed onset muscle stiffness are common after a marathon and may only be appreciated as significant injuries if they fail to subside in the following two weeks. Some runners may experience severe pain for days after a marathon race when walking up or down stairs.

Deaths occurring during or shortly after a marathon are naturally blamed on the event, particularly by the media, but may, in fact, be random and possibly unrelated. For example, a known epileptic ran the Marathon, went home, suffered a fit in his bath while nobody was in the house, and drowned. If the fit was an unlikely event, precipitated by running the Marathon, it could legitimately be blamed on the race; however, without knowing the frequency of the fits, whether or not the man had taken his medication and other factors, the culpability of the Marathon is indirect.

Another runner died in his sleep 36 hours after completing the Marathon. He told his wife how well he had felt during and after it. He went swimming the next day, but his wife was awakened by him having a terminal anoxic fit (a fit caused by lack of circulating oxygen) that night. He was found to have HCM at the post mortem. A claim was made in the press that the Marathon caused his death, and it is conceivable that a lingering biochemical or endocrine effect of prolonged exertion precipitated a fatal cardiac arrhythmia.

This raises many questions, such as, for how long after a marathon can the run itself be blamed for death, when in the presence of a lethal condition that can kill at any time? Deaths caused by HCM can occur at any time and an infrequent or unusual event may be blamed as the cause. Epileptic fits may occur in close proximity to running a marathon, and a statistical analysis of fit frequency and the total number of epileptics running the marathon would be needed to draw sound conclusions. If HCM has an incidence of one in 500, and people with this condition are not inhibited from running, it can be calculated that about 1,000 runners with this condition have run the Marathon and only two have died during the race.

Breakdown of diagnoses
Social (‘vaseline’) 97 2%
Constitutional 197 4%
Topical (‘blisters’) 731 15%
Muskoskeletal 3,963 79%
Total 4,988 100%


Breakdown of ‘constitutional’ casualties
Chest pain/breathing 3 2%
Headache 7 4%
Dehydration 14 7%
Collapse 17 9%
Feeling cold 25 13%
Nausea/vomiting 45 23%
Tired 86 44%
Total 197 100%

Figures from the 1987 London Marathon

Finishers and hospital contacts
Year Finishers Seen in hospital
1981 6,418 11
1982 15,758 34
1983 15,776 19
1984 15,649 15
1985 15,841 6
1986 18,031 5
1987 19,970 10
1988 21,100 38
1989 22,651 19
1990 24,871 20
1991 23,080 24
1992 23,657 15
1993 24,369 20
1994 25,000 40
1995 25,272 40
1996 26,000 90
1997 29,000 27
1998 30,000 59
1999 30,700 35
2000 32,600 35
2001 30,071 25
2002 32,200 41
2003 32,300 58

Questionnaires have been used to assess marathon morbidity in locally based marathons, but cannot be applied to major international races. They have a notoriously poor return. For example, a small survey of British doctors running the London Marathon in 1996 showed that less than 20% returned a questionnaire after the race, making the finding of a low percentage reporting upper respiratory tract infections in the week after the marathon invalid. The anticlimax and fatigue following completion of a marathon appeared to militate against completing and returning a questionnaire.

The totals of St John Ambulance casualty contacts are the numbers declared the day after the Marathon, once St John Ambulance has had returns from all station crews who disperse to much of southern England after the race. This number is sometimes subsequently corrected and discrepancies may occur when spectators are included in some returns and not in others. The total number of runners who make contact with the medical first aid posts may, if they outnumber the first aid provision, be under-reported as treatment may take priority over reporting, if a first aid post becomes swamped.

Accurate reporting of race casualties also becomes a problem where the fallen runner may have more than one diagnosis, eg exercise-associated collapse, plus blisters, plus subungual haematoma, plus groin chafing, but is only reported under the presenting complaint of collapse. A further complication is that the same runner may make contact with more than one aid station, making pit stops for ‘repairs’ at several and being counted as a fresh casualty or contact at each point. However, the multiple reporting error was assessed in one marathon and found to be a minor source or error in the grand total.

Casualties are assessed rapidly by first aiders and only very few are seen by trained diagnosticians. The diagnoses are, therefore, anatomical rather than accurate, where pain is the prime complaint. A painful shin may be a fatigue fracture, but there is usually no easy follow-up. Exercise-associated collapse may be registered under a variety of names, eg hypothermia, collapse and severe fatigue, even in hospital cases.

In summary

Based on 23 years’ experience, the approximate overall risks of running the Marathon are:

  • contact with St John: 1 in 6;
  • contact with a hospital accident and emergency department: 1 in 800;
  • hospital admission: 1 in 10,000;
  • death: 1 in 67,414 – a risk which is comparable to many daily activities.

Dan Tunstall Pedoe

Adapted and updated from ‘Morbidity and Mortality in the London Marathon’, a paper in Marathon Medicine, RSM Press Ltd, £19.95, 2001.

Share this

Follow us