PublicationsSystematic reviewsAccidental InjurySummary
Accidental injury, risk-taking behaviour and the social circumstances in which young people (aged 12-24) live: a systematic review. Executive summary


This report presents the findings of a systematic review of the research evidence relating to accidental injury, risk-taking behaviour and the social circumstances in which young people live. It is a broad-ranging review, covering topics as diverse as drugs, alcohol, transport and sport. It contributes a new perspective to the evidence base on risk-taking and injury by assessing explicitly the extent to which risk-taking contributes to accidental injury, and by locating this within the social circumstances in which young people find themselves. As well as examining the evidence for the above, it also contextualises its findings within current Government policy in a range of areas. It concludes that, while there is a large literature on a ‘culture of risk-taking’ among young people, the evidence to support the view that this translates into significant numbers of injuries is limited. Moreover, this review also challenges the idea that ‘risk-taking’ is a helpful umbrella term to describe the motivations underlying a range of activities. While young people undoubtedly undertake actions that result in injury, this review suggests that a move away from individual behavioural explanations towards a focus on structures and material resources is likely to be a much more productive approach to understanding overall patterns of accidental injury.


The background to the review is the establishment by the UK Government of a Task Force to form the basis of cross-departmental activity to reduce accidental injuries in the population. Accidental injuries among young people, which range from sprains in sport to hospitalisation and death due to drugs or transport crashes, among young people have been identified as a particular priority and this review was commissioned to examine the relationships between young people’s risk-taking and the injuries they sustain. This is because in industrialised countries such as England and Wales, accidental injury is the leading cause of death in children aged 0 to 14 years, and a major cause of death in young adults aged 15 to 24. It is also a major cause of ill health and disability in these age groups. Except where otherwise stated, the study population of this report is described as ‘young people’ aged 12-24.  ‘Children’ are below this age range and ‘adults’ above.


Our overarching review question, which was answered in different ways by the different types of evidence in the review, was:

What are the relationships between accidental injury, risk-taking behaviour and the social circumstances in which young people live?

We sought four main sources of evidence to answer this question:

  • research which tells us about the relationships between risk-taking and injury (‘correlational’ studies);
  • UK national statistics which tell us injury rates according to different causes;
  • research which has examined young people’s perceptions of risk, behaviour and accidents and the factors and contexts which influence them (‘views’ studies); and
  • systematic reviews of effectiveness to tell us ‘what works’ in preventing accidental injury internationally.
The inclusion of these different types of research is an important feature of this review. Its conclusions are drawn from international evaluations of injury prevention interventions, from the findings of ‘qualitative’ research examining the views and experiences of young people themselves conducted in the UK, and from gathering together what we know about the number and types of injuries suffered by young people.

We searched fifteen electronic databases, searched ten key journals by hand, scanned reference lists, contacted key informants and organisations, and searched websites for research to include in the review.  After examining the research in detail and assessing it for relevance and quality, the review’s conclusions are based on 84 studies.


The research fell into four main areas: drugs, alcohol, transport and sport. We examined the evidence in each of the areas before mapping the evidence back onto the overall burden of injury suffered by young people


We found that the use of drugs is associated with an increased risk of accidental death among young people and that 12–24 year olds are less at risk than those immediately older. Drugs are not as an important a cause of injury among young people as, for example, transport accidents. However, since the risk of death increases with the length of drug use, this is not to say that preventative interventions should not be targeted at this younger age group. While the use of drugs overall has stabilised over recent years, the use of some drugs – such as heroin, cocaine and ecstasy – has increased. Many more young men than women die from drug overdoses, because more men take drugs, but those women who do use drugs are at higher risk. Certain other groups of young people are more at risk than others. These include young people in deprived areas and men who have recently been released from prison. Ironically, many of the correlational studies adjusted for social and economic status in order to calculate standardised rates, thereby removing the very information which might help inform policy-making to assist marginalised groups.

The included correlational studies did not contain any mention of accidents at the workplace, home or school as a result of drug impairment. This review has identified a need for research in this area focusing more exclusively on young people.

There was a clear disjunction of views between young people who used drugs and those who did not. The young people who did not take drugs regarded them as risky and stated that media images about possible negative consequences dissuaded them from trying them, while those who did tended not to believe ‘official’ messages about possible harms and did not perceive taking drugs as being dangerous. Cannabis in particular was singled out as possibly being good for you, with some young people believing it to be cheaper than alcohol. The recent reclassification of cannabis and the subsequent debate may have helped to reinforce this view.

The above findings relate not just to two bodies of different types research, but also two different populations of young people. The young people in the views studies did not all take drugs, and those that did were more likely to be taking cannabis than ‘hard’ drugs. In contrast, the correlational studies cover deaths from opiates and other poisons. Our studies, therefore, do not contain the views of those groups of young people whose deaths are covered by the mortality statistics – either from illegal drugs, or due to volatile substance abuse (VSA). We also do not know how many, if any, young people suffer non-fatal injuries as a direct result of taking drugs. We do, however, know that the burden of the more serious injuries – as demonstrated by the mortality statistics – is carried by young people in the lower socio-economic groups.


Anecdotal accounts of accident and emergency departments being filled with alcohol-related casualties on a Friday and Saturday night are common, but at present, it is difficult to quantify the extent of the problem based on national statistics. Some injuries are clearly categorised in hospital episode statistics as having alcohol involvement, but since this is a supplementary code and relies on a blood test or subjective judgement, it is not clear whether these statistics are complete. We also do not know about alcohol-related injuries that do not result in hospital attendance. However, we do know how many young people are admitted to hospital from alcohol poisoning and we also know that almost everyone admitted for this reason is aged between 11 and 17. After a sharp peak among 14 and 15 year olds, hospital admissions for injuries with alcohol involvement decline slowly between the ages of 16 and 30.

Correlational studies have shown that alcohol puts the drinker at an increased risk of injury, that young people are more likely to have injuries than older people, and that young men are more at risk than young women. In the one study that examined ethnicity, minority ethnic status did not increase alcohol-related injuries, and may in fact have had a protective association.

The synthesis which examined young people’s views found that young people find drinking enjoyable with the young people in the studies we found saying that they do not commonly mix alcohol and other drugs. Most young people reported that drinking places them at greater risk of injury, though some do not. The younger teenagers – 14 and 15 year olds – felt most in danger of injury when drinking. Young people felt that they learned to manage their drinking through experience and that unsupervised, outdoor drinking was the most dangerous and was more common among younger teenagers (with injuries being considered less common in licensed venues). Peers encourage both drinking and drunken pranks, but also protect one another when they have become more vulnerable as a result of drink. Young people felt that drinking reduces their perceptions of danger and some stated that injury as a result was inevitable. Most young people were cautious about getting very drunk, though being sick as a result of drinking is common and not regarded as serious. Bad experiences – whether to self or someone else – might change behaviour in the short- but not long-term.

Evidence that interventions are able to reduce alcohol-related injuries among young people is scarce and research on ways of reducing alcohol-related violence was outside the scope of this review. The one study which examined attempts to reduce alcohol-related injuries found some evidence that motivational interviews in A&E departments are more effective than information handouts.


National statistics show that 10,900 young people were killed or seriously injured on the roads in the UK in 2004. Many more young men than women are injured, although the disparity is less among car passengers than for other forms of transport. Those aged under 16 are most at risk of injury on bicycles; at age 16, more are injured on mopeds, and then at 17 and over in cars. Young people aged 17–19 years are the most likely to be involved in drink-drive related accidents. The involvement of drugs in road traffic accidents has yet to be quantified.

Research suggests that younger, as opposed to less inexperienced, motorcyclists are more likely than older motorcyclists to be killed or seriously injured while driving, but also that the factor most clearly related to severity of injury is the engine capacity of the motorcycle involved. Young drivers of cars are also more likely to be injured than older drivers.

Sixteen to twenty-four year olds in the general population in England and Wales in 1996 had a higher number of alcohol-attributable road traffic accident deaths than all other age groups. In addition, within this age group young men had over eight times as many alcohol-attributable road traffic deaths as did young women. Drugs are found in the bloodstream of more young fatal accident victims than older age groups; however, it is difficult to assess whether drugs actually contributed to the accident. A low score based on the father’s occupational classification and low family affluence have both been reported as being predictive of injuries occurring on the roads, but the only independent predictor of child pedestrian injury is ‘playing in the street’.

Young ‘risky’ drivers were more likely to think they would have an accident, but were also less worried about this than ‘safe’ drivers; accidents or dangerous driving were not considered to be embarrassing. Some young people expressed a fatalistic view of the chances of having an accident, and some of those that had been involved in crashes stated that it had not really affected their driving behaviour. ‘Risky’ driving was seen as enjoyable and not considered to be the same as driving unsafely. However, different drivers had different perceptions of what a ‘good driver’ was, with ‘safe’ drivers emphasising safety and ‘unsafe’ drivers emphasising driver skills.

The presence or absence of other people influenced driving behaviour. Some passengers (e.g. parents) tend to reduce risky driving, whereas others (e.g. peers) might encourage more risky driving. Young men were more likely to take risks than young women. Young people stated that they judge the degree of acceptable risk depending on the situation. Young people said they were more likely to drive riskily when driving alone or late at night when the roads are quieter than during the day or when they were responsible for others in the car. Some young people felt they ‘grew out’ of risky driving as they got older with more expensive cars and family responsibilities. They also said that the social expectation that they would drive riskily made it more likely that they would do so.

Young people expressed a difference in attitudes to drink-driving versus driving on cannabis. Drink-driving was generally considered dangerous and not socially acceptable, whereas driving on cannabis was more acceptable and not thought to be dangerous. Some young people stated that a lack of public transport (or alternatives, such as taxis) made it more likely that they would drink and drive.

There is a large evidence base of systematic reviews that look at interventions designed to reduce traffic injuries. Legislation on the wearing of motorcycle helmets has been found to be effective, as has legislation and enforcement on reducing drink-driving. Raising the minimum driving age and introducing graduated licensing schemes for young people have also been shown to be effective in reducing accidents. Curfew laws, which restrict the times during which new drivers can drive, also reduce accidents. Seat belt campaigns have led to behaviour change and consequential reductions in injuries.

Interventions based on models of behaviour change to reduce drink-driving are ineffective or have a negative effect. Combining different approaches has more effect than using a single approach. Education or skills training has either negative or no effects on driver behaviour and subsequent accidents, possibly because these approaches lead to over-confidence or early licensing. The effectiveness of pedestrian education is not proven. Environmental modification and the enforcement of speed limits may be more effective at pedestrian accident prevention.

Evidence relating to community-based interventions is mixed. Some studies found that the mobilisation of social support was effective in reducing drink-driving. However, while evidence of their effectiveness is increasing, community-based interventions have not yet demonstrated a sustained impact in reducing injury.

Bicycle helmets

Despite the debate on whether or not cycle helmets should be compulsory, the balance of evidence suggests that wearing helmets reduces injury in the event of an accident.

Young people in the ‘views’ studies did not think cycling was very risky and did not think that accidents would happen to them. Cycling conditions (e.g. at night or in bad weather) affected their perceptions of risk and some young people believed that helmets were only needed for certain types of journey. The fact that risk perceptions differed for different types of journey did not predict the wearing of cycle helmets, only the intention to wear one, but young people who ‘could never forgive themselves’ if they had an injury which a helmet could have prevented were more likely to wear one.

Young people differed in their perceptions about the protection offered by helmets, with some feeling that the head was not the part of the body most likely to be injured in an accident. Beliefs about the safety offered by helmets do not necessarily predict whether a young person is likely to wear one.

‘Critical incidents’ – whether to the young person or to someone known to them – were reported to change behaviour, but only in the short-term. Teenagers in one study were more likely to wear a helmet if they were engaged in more risky cycling, whereas another study came to the opposite conclusion. Most young people did not feel that wearing a helmet affected the way they cycled, though some expressed concern about the possibility that motorists would think they were less vulnerable and so drive more dangerously around them.

Peers tended to be a negative influence on the wearing of cycle helmets due to concerns about being teased, particularly when first starting to wear one. Parents were often a positive influence, and some young people felt legislation would also make them more likely to wear helmets. Young people also mentioned that cheaper and more attractive helmets would be more inviting to wear.

Education interventions through a single medium were not effective at increasing the wearing of cycle helmets, but multifaceted interventions were effective. Education interventions targeting small groups had mixed effects, with some studies reporting positive results and others finding no effect.

Acknowledging that the price of helmets can be a disincentive, particularly for disadvantaged children, some interventions which provided subsidised or free helmets did increase helmet use. There was conflicting evidence about the relative effectiveness of providing free and subsidised helmets. There is some evidence to suggest that promotional campaigns have more effect in high income areas, and also that younger children were more likely than teenagers to change their behaviour (especially girls). There is also some evidence to suggest that community-based interventions are more effective than interventions based in the school.

Most primary research on helmet wearing focused on younger children, rather than young people up to the age of 24. Strategies targeted at the older age group may therefore need to be developed and tested.

The balance of evidence suggests that a combination of legislation and enforcement is the most effective means of reducing cycling injury. However, the introduction of compulsory helmet wearing can also lead to a reduction in cycling. In order to avoid this, six reviews recommend a multi-faceted approach in which legislation is preceded by education and promotional campaigns.


The greatest numbers of sports-related injuries occur to young people playing football (27%) and rugby (10%) – because these are the most widely played sports. When activity rates are taken into account, rugby is by a long way the most dangerous mass-participation sport. Stick-based sports, such as hockey, also have high injury rates, and eye injuries are more common in racquet sports. One study found that half of all injuries to young people in an accident and emergency department were sports related, and most studies which examined differential injury between sexes found that young men suffered more injuries than young women. The Health Survey for England found that sport/exercise accident rates peaked in young people between the ages of 13 and 15 years old.

Athletics injuries are reduced if a coach is present, and there is some evidence to suggest that adult/guard supervision can reduce injuries at swimming pools and beaches. Systematic reviews failed to find strong evidence for interventions to reduce sports injuries, and no studies in any of the reviews tackled the issue of injury in relation to social deprivation. Much of the research is focused on younger children and the question of whether or not these interventions are appropriate to a teenage audience needs to be considered.

With regard to specific sports, the use of custom-fitted mouthguards was found to reduce oral injuries in rugby and changes to the rules reduced very serious injuries in rugby and ice hockey. Similarly, the use of protective equipment reduced injuries to players of American Football, the use of belts benefited weightlifters, and eye/face protectors reduced squash injuries. For those with previous ankle sprains, ankle supports and taping were beneficial in reducing future sprains (in soccer and other sports, such as basketball). Specially designed baseball bases, which come away from the ground easily when players slide into them, were effective in lowering injury rates. Despite considerable research, there is little strong evidence to help runners avoid injury. Neither stretching nor warm up/down regimes appear to be associated with reduced rates of injury.

The environment in which sport takes place can affect injury rates. Smaller rinks are associated with more injuries in ice hockey, and the design of swimming pools can also change injury patterns. Removing monkey bars and increasing the depth of protective bark in public playgrounds can reduce injuries significantly. Reducing potential hazards in school playgrounds has also limited rates of injury.


Each chapter contains specific implications and recommendations relating to the specific topic areas. Presented here are the overarching recommendations resulting from looking across the different causes of injury.

While the above topic areas cover most of the causes of accidental injury among young people, there is little evidence that consistently links individual risk-taking with accidental injury. This suggests that a move away from individual behavioural explanations towards a focus on structures and material resources is likely to be a much more productive approach to understanding patterns of accidental injury among young people.

Very few correlational studies examine possible structural causes of injury, and those that do so use different and/or unclear definitions. We need not only clearer definitions of social and economic status (SES), but better quality information on how different factors within these definitions interrelate and how they impact on accidental injury rates among young people.

There are pronounced differences between young men and young women both in rates and types of injury and in their exposure to different injury risks. Studies which examine young people’s views are often eloquent on the subject of the different cultures of masculinity and femininity which contribute to shaping their behaviour, particularly with respect to transport and alcohol-related accidents where there is almost an expectation that taking risks is a normal feature of young men’s lives. This is an essential context for understanding, for example, why in the 16–24 year age group young men are eight times more likely than young women to have an alcohol-related road accident. A focus on gender differences reveals some notable research gaps, for example the higher drug-related mortality risk among young women, and the greater vulnerability of young women to road accidents when travelling in groups.

Bearing in mind stated Government aims of increasing physical activity in the population as a whole, we also recommend that attention be given to ensuring that young people are not exposed to additional risk of injury as a result. In particular, if cycling is to be promoted as a healthy and environmentally friendly form of transport, we need to address the significant risks faced by cyclists sharing busy roads with much faster and heavier vehicles. The significant variations in injury rates between different sports should also be examined and the place of dangerous sport in the school curriculum should particularly be considered.

Risk-taking behaviour as an umbrella concept cannot be regarded as a useful model to explain accidental injury among young people. However, since some behaviours clearly lead to injury, we recommend a more fine-grained approach which is more sensitive to the underlying motivations, history and reasons for injury. Rather than focusing on ‘risk-taking’, behavioural interventions need to take account of people’s different situations and recognise a striking finding across the systematic reviews – that education/information interventions alone have limited or no impact. This may not be because young people are unreceptive to the messages of such interventions, but because the influences of other factors, including the environment and social context, have an overriding importance.

The contribution of this review

This review is the first of its kind to look across risk topics and include such a wide range of research designs. It presents information not only on ‘what works’, but on the causes and extent of injury from a large range of sources, and it synthesises the views of the people who are the focus of interest and locates this information within current UK Government policy. It is therefore able to draw conclusions with some authority, since its broad perspective enables it to look across research activity in so many areas.

Examining research from a ‘behavioural’ viewpoint has enabled the review to take a critical look at the evidence base which underpins many current health promotion interventions. The conclusions about general approaches to research and the difficulty of explaining accidental injury through behaviourist approaches alone are strengthened by the broad scope of the review.

Copyright 2019 Social Science Research Unit, UCL Institute of Education :: Privacy Statement :: Terms Of Use :: Site Map :: Login
Home::About::Projects::Training::Research Use::Resources::Databases::Blog::Publications