Health Behaviour Unit

The work of the Health Behaviour Unit is designed to advance understanding of behaviours that affect cancer risk, such as smoking, food choice, physical activity, and participation in cancer screening, with a view to developing more effective interventions leading to a reduction in cancer incidence and mortality. The tobacco programme is led by Professor Martin Jarvis.

Dietary composition and energy balance are significant factors in cancer prevention. Our research investigates the factors that determine food choices, within the individual, social and environmental domains. We are studying the development of food choices in a twin cohort in order to establish the contribution of genetic and environmental influences. In relation to the food environment, we study the effects of the energy density and the timing of consumption. We are also investigating the influence of parental feeding style, particularly parental control, where we hypothesis that higher levels of control could be advantageous. We have shown that children;s intake of fruits and vegetables is especially low, and that parents lack knowledge of strategies to improve their children's diets. We have completed a successful intervention to train parents in a strategy to promote children's acceptance of vegetables, and are planning a more extensive trial of comprehensive feeding advice with mothers of young infants.

Recent reviews have linked obesity and physical activity with a number of cancers. We are doing investigations of the aetiology and management of obesity in children, adolescents and adults. We have used a twin design to investigate the behavioural phenotype associated with obesity risk in children. We have also completed two large studies examining the weight trajectory of adolescents and shown that weight control efforts are associated with a healthier lifestyle, but the often concomitant body shape dissatisfaction is associated with poorer mental wellbeing and a less healthy lifestyle. We find that a substantial minority of overweight adults and adolescents make no efforts at weight control, and we are investigating ways to promote healthy weight control without damaging self-confidence and body image. In the clinical domain we have developed a weight management programme, 'Shape-up', for dissemination in self-help groups and primary care. We are also setting up a research clinic for evaluating new therapeutic and preventive programmes for obesity.

Early detection of cancer or pre-cancerous changes is an important factor in reducing cancer mortality, but its effectiveness depends upon high levels of public participation. As part of the UK Flexible Sigmoidoscopy Screening Trial, a trial of bowel cancer prevention in 400,000 adults, we are studying the processes whereby individuals accept or decline screening. Data from the first two centres have been used to develop a strategy to increase screening acceptance. We are also examining the psychological impact of screening and have found no evidence that screening participation is either upsetting or promotes unwarranted complacency about aspects of lifestyle. We are now extending this work to HPV screening, where there are important issues related to the sexually-transmitted nature of the infectious agent.

Public understanding of cancer - its nature, causes, prevention or treatment - is limited and often erroneous. For example, people are unaware that different cancers have different causes and outcomes, are ill-informed about dietary recommendations, don't know that obesity is a risk, and have rarely heard of the possibility that viruses may cause cancer (e.g. 1.6% aware that HPV causes cervical cancer). Attitudes towards treatment and survival are often fatalistic. We have carried out population surveys to assess public understanding of cancer and are now extending this work to include genetic risk and considering the implications for more effective public health education on cancer.

Development and change in health behaviours during the formative years of adolescence period is the focus of the HABITS cohort study, which is following 5000 adolescents in 36 schools across London, from age 11 to 16. Smoking, diet, physical activity and weight are assessed annually over this period of rapid change and development. In addition to assessing health beliefs and attitudes, and family and social influences, we shall also be collecting DNA to permit analysis of possible genetic determinants of behaviour. Associations with emotional wellbeing and life stress are important aspects of the investigation.

In carrying out interviews with potential participants in the cancer screening studies, it has become clear that many people's understanding of cancer - its nature, causes, prevention or treatment - is limited and often erroneous. For example, people are unaware that different cancers have different causes and outcomes, they are ignorant about recommendations to increase fruit and vegetable intake, or the hypothesised role of diet in cancer, and many think that a diagnosis of cancer means a certain and horrible death. From the limited research that is available on public communication about cancer, it was unclear whether people want more information about cancer or whether, as the media would have us believe, publicity about cancer simply increases anxiety. In 1998 we started a programme of work on public understanding of cancer. This has included carrying out two surveys to assess public awareness of warning signs for cancer, knowledge of established cancer risk factors, and belief in causes with little scientific substantiation (food additives, overhead power lines). Our results show that adults have low levels of knowledge, and that those with a poorer understanding of cancer have more fatalistic views about outcomes. We also addressed the concern that publicity about cancer screening would increase anxiety. In fact our results showed quite the opposite, that people welcomed information about how to prevent cancer.

For a list of refereed research papers, see Publications (in navigation on left).