Discussion

Health benefits of exercise

Paffenbarger et al's Harvard alumni study of 1986 35 provides useful data for plotting the relationship between exercise and mortality. It includes a table relating activity levels (as defined by energy expenditure) to risk of all-cause mortality; a simple calculation allows energy expenditure to be expressed as time spent cycling, and a graph of relative risk against cycling time can then be plotted (Figure 1). The reason for the increased mortality above 3500 kcal/week is not clear, and is at odds with other studies which have shown similar gradients for reduction in mortality from increasing activity at much higher levels of energy expenditure.58

Deaths from all causes in Oxfordshire in the years 1994-98 led to a loss of 340,343 life years. Relating this to the health benefits of exercise shown in Paffenbarger's study 35 allows an estimate of the number of life years that could be gained as a result of deaths avoided if cycling activity were to increase, as shown in Table 2. A relative risk of 0.78 was used for the calculations as this is the appropriate relative risk for people exercising at a level equivalent to cycling for 20-40 minutes/day at 10mph. This reflects the potential benefit available even to inexperienced cyclists making short journeys. If 1% of the population took up cycling these people would reduce their risk of all cause mortality by (1.0-0.78) - i.e. by 22%. These calculations are inevitably crude, but they give a rough idea of the order of magnitude of the potential benefits. The calculations are based on an assumption that cycling deaths increase linearly with increasing levels of cycling. It appears, however, that cycling becomes safer as cycling levels increase (see section on Speed and Danger below). I have also assumed that the risk exposure within this sub-population is the same as that within the general population. I have requested more detailed figures on this risk exposure from the DETR, and will update the calculations once I have received them.

Table 2: Life years gained if an extra 1% of the population of Oxfordshire were to cycle regularly

The mortality reduction resulting from regular exercise significantly outweighs the dangers of cycling on British roads. Although the greatest benefits are found in those who expend over 2000 kcal/week, lower levels of exercise also have a beneficial impact on mortality and morbidity. At a population level even relatively meagre increases in activity by large numbers of people could result in significant health gains; individually these gains are likely to be the greatest among those with the highest current risks of death or disease, but the population benefits may be most significant in the larger numbers of people with lower personal risks.

Transport and the environment

Motorised transport imposes significant environmental costs. In order to appreciate the environmental benefits of a switch from motor vehicles to walking and cycling (a so-called modal shift) it is necessary to consider the environmental effects of traffic. Buses and trains are less environmentally damaging than cars and lorries, and may generate cycling and walking at either end of journeys, but detailed consideration of public transport is beyond the scope of this report.

Pollution

Vehicle emissions cause asthma, chronic cough, bronchitis, emphysema, lung fibrosis, and premature death.59 60 Catalytic converters have reduced harmful emissions, but levels remain high, and by 2010 the improvements will be overtaken by traffic growth and the concentrations of pollutants will rise again.61

The recent discovery of water at the north pole 62 is just the latest manifestation of the effects of global warming, which is generally accepted to be attributable to emissions of carbon dioxide and other heat-trapping gases. Motor vehicle emissions are a source of around 20% of these gases, and thus a major contributor to climate change and all its attendant health and social problems.63

Transport is the most pervasive source of noise in the UK daily environment.63 Almost half the population of the EU is adversely affected by road traffic noise at levels that disturb sleep and may have broader effects on health.60 Most traffic noise in urban areas is a result of engine and transmission noise, but at night events such as slamming doors and starting engines can be a major irritant. At high speeds tyre noise becomes the main source of noise; this can be tackled by reducing speeds, erecting acoustic barriers, and constructing special road surfaces.64 65

Speed and danger

Crash, injury and mortality data are extremely poor measures of risk; pedestrian fatalities are rare on both car-free precincts and motorways, but these locations clearly pose very different levels of danger for pedestrians. Much more sophisticated indices are needed accurately to reflect the dangers to which different road users are exposed.3 64 66

The 30mph speed limit is observed by only 31% of drivers.67 The contrast between the high-risk and population approaches to disease prevention applies equally to traffic enforcement: should the police focus on the small numbers of people breaking speed limits by a large margin, or attempt to tackle the lower level infractions by much larger numbers of people? These decisions are complex, and subject to many other factors such as the desirability or otherwise of criminalising a large proportion of drivers and the capacity of the criminal justice system, but they deserve consideration in terms of the potential health benefits of controlling excessive speeds.64 It seems credible that Rose's prevention paradox 68 should apply to traffic danger, and that the majority of harm is a consequence of common, relatively low level, speeding.

One of the criticisms of the Oxford Transport Strategy has been that reducing traffic levels in the city centre has allowed speeds to increase for the remaining traffic, thus increasing the risks for pedestrians and cyclists: within the range of likely urban speeds, every 1mph increase in speed raises mortality by 5%.69 A recent study of congestion and risk supports this, showing that the accident rates for cyclists and pedestrians in urban and peri-urban conditions were very much lower (by between 60 and 85%) in congested than uncongested conditions.70

There is an understandable reluctance on the part of many road safety professionals, and others responsible for transport decisions, to encourage cycling for health because it has a higher casualty rate than many other modes of transport, but this view should be challenged on several grounds. The health benefits of cycling greatly outweigh the risks, and higher rates of cycling, when supported by good infrastructure, are associated with lower casualty levels. This is shown by data from many cities in continental Europe (such as Lüneberg in Germany 71), and the experience of York, where despite a 10% increase in peak-hour cycling over the last 10 years casualties have fallen by 30%.72

Social factors and inequalities

Traffic has many other detrimental consequences for society as a whole, especially groups such as children, the elderly, and the poor;73 these people are also the least likely to have access to cars, and the most likely to rely on buses.74 Physical activity is related to socio-economic status, and is an important factor in health inequalities:75 children in the lowest socio-economic group, for example, are five times more likely to be killed as pedestrians than their counterparts in higher socio-economic groups.73

The development of the modern road network has had a profound effect on the built environment and the way we lead our lives.3 76 The ease of car travel has allowed people to live at greater distances from the places they work and shop, and the synergy between car usage and out-of-town developments has fuelled this disaggregation of communities. Many Oxfordshire villages are now low-density commuter suburbs rather than rural communities.24 It will be difficult to reverse these social and geographical trends, and there is currently little public or political will to do so.

Traffic can also lead to social isolation and community severance. Appleyard and Lintell's study in San Francisco showed that the residents of streets with differing traffic flows (light, medium and heavy) gave different ratings for traffic hazard, stress/noise/pollution, social interaction, privacy and environmental measures, with the busiest streets faring worst.77 Social networks have a positive influence on health; their disruption by traffic adversely affects health.78 Empowering communities to develop their own responses to health problems can help to build social networks at the same time as tackling the problems.79 80

Stakeholder views

One of the most striking features of the interviews was the range of factors that encourage or discourage people from cycling. These seemed to reveal a hierarchy of needs matched to cycling experience. Less experienced cyclists tended to be keener on safety facilities such as segregated routes, while the more experienced ones were more supportive of features, such as advanced stop lines, that give the cyclist priority over traffic. This equates to the hierarchy of needs found in a recent study of regular cycling in previous non-exercisers. At the start of the study 'participants were most concerned about the distance being too far, being unfit, and overcoming practical problems. As they became more used to cycling, distance and lack of fitness became less important, while weather, air pollution and fear of injury became more important.'81

Many people reported feeling threatened by large vehicles, especially buses, which they felt drove too fast and too close. Representatives of the bus companies object to these criticisms as unjustified, stating that buses are not disproportionately involved in crashes with cyclists. But if we are to encourage more people to cycle the concerns arising as a result of perceptions need to be addressed, whether or not the vehicles actually pose a disproportionate threat. The potential public health benefits of increasing levels of cycling are sufficiently large to justify major efforts to encourage anxious cyclists onto the road, and partnership with bus companies and other commercial vehicle operators is an important element in this.

Several interviewees raised objections to poor cycling, citing behaviour such as ignoring traffic signals, and riding on pavements or in pedestrianised areas. The offending behaviour seems to fall into two broad categories: breaking the law, and placing other people in danger. In fact, cyclists pose very little threat to other road users, including pedestrians, and it is noticeable that in countries with higher levels of cycling there is much greater acceptance of pedestrians and cyclists sharing space. While illegal behaviour cannot be condoned, it seems surprising that law-breaking cyclists, who are generally only placing themselves in danger, received greater criticism than speeding drivers who place other road users in danger.

It was notable that the major themes in the interviews were extremely similar to those found in a 1996 survey of attitudes to cycling,82 and suggests that the interviewees were a reasonably representative group. The survey found 'varied and complex' attitudes to cycling, with 'danger from traffic, concerns about personal safety and cycle theft, and poor image cited as major deterrents to cycling. Attitudes of government and institutions [were] seen as favouring the car and giving an inferior status to the bicycle.' Reducing motor vehicle speed, better enforcement of traffic regulations, provision of secure cycle parking facilities, employer/college initiatives, and more information about cycle routes were felt to be the most likely factors to encourage cycle use. Previous studies in Oxford 83 84 have produced similar results.

Recent policy initiatives from central government emphasise a broad perspective on health, embracing agencies beyond those engaged in formal health care. Co-operation between agencies is essential for tackling health and transport problems, and successful joint working requires transparency, trust, and a recognition of the differing viewpoints of different groups. This can be especially difficult when involving the public in decision making, given the inevitable power imbalances and multiple viewpoints inherent in the process,85 and the professional norms of the experts involved.86 But complex problems such as transport cannot be solved by independent agencies each pursuing their own agendas, and real progress will not be possible without addressing and tackling these issues.

Many different groups and individuals feed their interests into the policy-making process. These interests often conflict, and the way in which this is resolved depends upon the people doing the resolving, just as the ways in which a balance is struck between the desires of drivers and the wishes of cyclists depend upon who is doing the balancing: these are not objective processes. An understanding of the alternative rationalities involved, and a consideration of different cultural constructions of risk, can help to overcome apparently irreconcilable differences between groups or individuals with opposing viewpoints.87

Sedentary middle-aged people are among the groups most likely to benefit from reducing their car use and increasing their walking and cycling. They may have a bicycle sitting in their shed, but prefer to leave it there and drive instead. It was not feasible to conduct original research for this report, and although both local and national surveys were studied this is no substitute for ascertaining local people's views. Almost all stakeholders interviewed for this report cycle at least reasonably frequently, and thus may not be best placed to identify strategies to remove the barriers perceived by non-cyclists.

High quality public consultation is difficult and expensive to perform, but if local transport policies are successfully to address the cycling needs of current non-cyclists the public must be fully involved in the process. The recent blockades of oil refineries in protest at current levels of fuel taxation demonstrate the difficulty of reasoned argument about the complex issues involved in transport, health, and the environment. Nevertheless, an intelligent debate between experts and the public needs to be conducted, and imaginative approaches to public consultation are required. This may include initiatives such as local radio phone-ins, local parliaments and similar community fora, and other mechanisms by which views, and potential solutions, can be identified.

Limitations of this report

The use of mortality data allows basic comparisons at the level of life years lost and gained, but is a crude measure of health benefit, and depends upon many assumptions. Road death statistics provide a very distorted view of true danger as a result of multiple social and behavioural adaptations leading to risk avoidance and risk compensation.87 88 Physical exertion can trigger myocardial infarction in unfit individuals,89 but regular activity provides a strong protective effect against this.90 Significant long term morbidity is not likely to be a significant public health problem in terms of cycle casualties,20 91 92 but is a major issue in coronary heart disease; this has not been factored into the calculations, which therefore underestimate the potential health benefits of modal shift.

The calculations in this report were based on a relative risk obtained from a study that considered men aged 35 to 74, but because of the small numbers of casualties in Oxfordshire they were performed on data relating to men and women of all ages. Since men have far higher cycling casualty rates than women this may have overstated the health benefits, but the Oxfordshire results are consistent with calculations on national data for men aged 35-74; these showed a ratio of life years gained to life years lost of 269, and 245 for men and women of all ages.

One of the problems with the evidence on the health benefits of exercise is the variability of the indices used. Studies have investigated workplace activity,2 58 leisure activity,93 94 commuting,95 and overall activity,96 and the measurements made include fitness,32 97 and self-reported activity.34 Most of the major observational studies on physical activity involve a degree of self-selection in terms of the exercise level for the participants, so may tend to overstate the benefits; the calculations in this report have thus been based on the most conservative figures available in appropriate studies.

The aim of this report is to investigate the public health benefits of reducing car use and increasing walking and cycling, and make policy recommendations as a result. It does not attempt to determine the most appropriate strategies for promoting cycling or other exercise, although studies have looked at this,98 99 100 101 and have also considered its cost effectiveness.102 103 104 105 Detailed financial calculations have not been made, for a number of reasons. The cost implications of the recommendations to the Health Authority are small, and the recommendations to the local authorities could be met with relatively minor reallocation of part of their transport budgets from car facilities to cycle facilities; cycle parking spaces are, for example, much cheaper than car parking spaces. Beyond this the financial implications of potential health benefits are too complex to be dealt with in a report of this type.

Although the interviewees were chosen to be broadly representative of local stakeholders, it was impossible to include representatives of every viewpoint. The Travel for Work Project Officer in Cambridge (whose post is funded by Cambridgeshire Health Authority) was consulted about local schemes she is involved in and provided an example of their literature,106 but there was no broader comparison with other cities or districts.

The health and safety of children, and wider issues about travel to and from school, are important and complex issues. Childhood obesity is increasing, and exercise rates in childhood are declining. Lifetime exercise patterns are established in childhood, and are also linked with other factors relevant for health such as self-esteem and confidence. Issues such as these have been studied in detail elsewhere 1 101 and it was beyond the scope of this report to consider them in isolation.

Top | Home | Previous | Next