3.2.3 Cohort studies
These focus on groups of people who show certain attributes or characteristics (e.g. with respect to their health behaviour). The groups are then observed over a period of time in order to discover what happens to their individual members and to check whether there are any associations between behaviour and the development of disease. In this way the famous epidemiologist, Sir Richard Doll, and colleagues investigated doctors’ smoking habits and prospectively followed the sample over forty years, by which time two-thirds of the sample had died. The finding obtained during the first twenty years was that doctors who were heavy cigarette smokers were thirty-two times more likely to die of lung cancer than doctors who were non-smokers (Doll and Peto, 1976). Mortality associated with smoking doubled during the second half of the study (Doll et al., 1994). Examples of cohort studies include the National Study of Health and Growth of School Children, started in 1974 (Stamatakis et al., 2005), and the Millennium baby study (Millennium Cohort Study, 2006).
Longitudinal studies are a form of cohort studies, which study groups of people over time; they can be retrospective as well as prospective. Longitudinal studies can be very costly in time and money and require the following up of subjects, which may prove difficult. However, they have the advantage of being able to accumulate very useful information for the determination of the long-term effects of biological, environmental and social factors on health. For example, comparing cohorts born every twelve years – say, in 1946, 1958 and 1970 – could throw light on the effects of changes in social and health policy and education on beliefs about diet, smoking, exercise and other issues of interest to those engaged in promoting public health.