By Liz Banks-Anderson and Katherine Smith
In 1964 Granada Television in the UK followed a group of British children from various walks of life through their lives, reporting on their development, and the societal changes that shape it, every seven years. Directed by Michael Apted, it became known as the Up Series.
Memorably, the first episode of 7-Up shows little Neil, with his strong suburban Liverpool accent, explaining the wetness of an English countryside in winter, and his ambition to drive a motorbus with a loudspeaker to take visitors to the country when he is an adult.
Photo: Getty Images
In 2005 he appeared in 49-Up, having apparently come through the other side of pain – the depression and despair he recounts in 28-Up, a personality disorder he reveals at 35.
While such a long-term filmic study of lives was and is unusual, researchers in health and the social sciences use a similar scientific format as a means of investigating a wide range of health and sociological issues.
The Up Series is non-scientific example of what is known technically as a cohort study – a form of longitudinal research tracking changes observed in individuals over time.
Director of the University’s Centre for Epidemiology and Biostatistics, Professor Mark Jenkins says cohort studies are valuable because they can be used to identify a wide range of social determinants of health in individuals.
“Cohort studies are a type of natural experiment that take advantage of the fact that people differ in terms of characteristics, lifestyle factors, and their environment.
“They can be used to study potential risk factors that would be unethical to ask research participants to use or consume in a clinical trial.
“One of the most successful cohort studies was a British study of 40,000 doctors surveyed on their smoking habits, and followed for two years. This landmark study was one of the first to show that smokers were more likely to die from lung cancer.
The beauty of cohort studies, according to Professor Jenkins is that they can investigate single or multiple risk factors. Among the vast number of variables studied are personal characteristics (weight, age, sex, genes, family history, ethnicity), lifestyle factors (smoking, alcohol consumption, physical activity, diet), occupational or other chemical exposures (radiation, pesticides, electromagnetic fields, passive smoking), environmental factors (sun exposure, pollution, latitude, mosquito infestation), medications (oral contraception, hormone replacement therapy, aspirin, cholesterol lowering drugs), family history (cancer, depression, heart disease), household factors (pets, gas heaters, asbestos) or other risk factors (such as mobile phone use, mosquito nets, condom use).
Professor Jenkins explains that to conduct a cohort study a large group of healthy people (often thousands) are studied for the presence or absence of a range of factors, often via a questionnaire or providing a biological samples or other measurements. Then they are followed over a period of time – often years – to see which people get the disease.
“Factors found to be associated with disease are potential risk factors,” he says.
“This does not mean the risk factor caused the disease (different types of studies are needed to confirm causation) but it does mean the factor is marker for risk. If the association is very strong, the risk factor can be used to predict who is more likely to get the disease.”
While in the past cohort studies have relied largely on samples taken and interviews or questionnaires, genetic advances are now making a huge impact.
“One of the potentially important advances in risk factors is the increasing availability of genetic data,” he says. “Current costs for measuring the DNA of cohort participants are only a small fraction of the costs of even 10 years ago.
“Cohort studies of genetic risk factors for many diseases have the potential to identify means of putting people into high and low risks of disease which can then be used to target prevention to those at most risk. One of the potentially important advances in disease outcomes is the sub-classification of disease based on pathology subtypes. This has enormous potential to identify risk factors that might not be possible if all sub-types of a disease (each with their own risk factors) are naively lumped together.”
Cancer epidemiologist Professor Dallas English from the Melbourne School of Population and Global Health works on two cohort studies – the Melbourne Collaborative Cohort Study with Professor Graham Giles from the Cancer Council of Victoria and Ten to Men (see break out).
“We have examined a range of health outcomes including cancers, cardiovascular disease and type 2 diabetes with respect to alcohol consumption, diet, weight, obesity and genetics,” he says.
Professor English, also Director of the Centre for Molecular, Environmental, Genetic and Analytic Epidemiology and cancer epidemiologist, also points to the British Doctors Study, a cohort study from the early 1950s, which demonstrated clearly for the first time that smokers were more likely to develop lung cancer.
The advantage of a cohort study, Professor English says, is that they can identify factors that increase or decrease the risk of disease and injury, for many different types of disease.
“In a cohort study, one generally begins by identifying a ‘cohort’ of people who are healthy and then following them over time to see which ones develop the disease of interest.
“One great value of cohort studies is that they allow us to study many diseases simultaneously. For example, a single study can show how smoking is related to several types of cancer and to various types of cardiovascular disease,” he says.