Epidemiology is the study of the patterns and causes of disease in populations. Epidemiologists collect data on who gets a disease, where and when, and look for factors that are linked with it. For CVD, epidemiological studies identified the main risk factors, such as smoking, high blood pressure and high blood cholesterol.Types of study.• Cohort study. A large group of people (a cohort) is followed over a long time, recording their lifestyle and health, to see who develops the disease. The Framingham Heart Study in the USA, started in 1948, followed thousands of people and their children in this way.
• Case-control study. People who already have the disease (cases) are compared with similar people who do not (controls), looking back at differences in their lifestyle or history.Correlation and causation. Epidemiology usually shows a correlation: a link between a factor and a disease. A correlation does not prove causation. Other evidence, such as a biological mechanism explaining the link or experimental data, is needed to show that the factor actually causes the disease.A confounding variable is another factor that affects the outcome and can make a link look stronger or weaker than it really is. For example, people who drink a lot of alcohol may also be more likely to smoke.Judging a study. A good study uses a large, representative sample, which gives more reliable results, controls for confounding variables, uses valid and accurate measurements, and is published and peer reviewed so other scientists can check it.
• Case-control study. People who already have the disease (cases) are compared with similar people who do not (controls), looking back at differences in their lifestyle or history.Correlation and causation. Epidemiology usually shows a correlation: a link between a factor and a disease. A correlation does not prove causation. Other evidence, such as a biological mechanism explaining the link or experimental data, is needed to show that the factor actually causes the disease.A confounding variable is another factor that affects the outcome and can make a link look stronger or weaker than it really is. For example, people who drink a lot of alcohol may also be more likely to smoke.Judging a study. A good study uses a large, representative sample, which gives more reliable results, controls for confounding variables, uses valid and accurate measurements, and is published and peer reviewed so other scientists can check it.