Types of Studies:EPIDEMIOLOGY - L2
What is epidemiology?
What can epidemiology tell us?
How are epidemiology studies conducted?
What are the different types of epidemiology study designs?
What are the limitations of epidemiology studyies?
How are the results of epidemiology studies interpreted?
What is epidemiology?
Epidemiology is the study of the distribution of diseases in populations and of the factors that affect this distribution, or in other words epidemiology is the study of how often diseases occur in different groups of people and why.
Epidemiology has played a vital role in identifying and quantifying the health risks of cigarette smoking and exposure to agents such as asbestos.
What can epidemiological studies tell us?
Through epidemiology we have been able to learn a great deal about disease incidence and causes of disease. Such studies are important for health risk assessments as they directly study people, however, they must be interpreted with caution.
How are epidemiological studies conducted?
Studies begin with the research question or hypothesis to be answered. For example, do people who drink alcohol have a higher risk of heart disease? The appropriate study populations are then selected and exposure to the alleged disease risk factor is assessed. There a number of different study designs in epidemiology, each with different strengths and weaknesses.
What are the different types of epidemiological study designs?
Some common types are:
These studies follow a group of healthy people with different exposure levels and assess what happens to their health over time. In these studies exposure comes before the disease occurrence which is necessary to establish possible causation. Whilst expensive and time consuming, these studies suffer least from bias as they make fewer assumptions about the study subjects. Cohort studies are most useful for relatively common diseases.
These compare prior exposure to proposed risk factors of individuals with a particular health condition (cases) and those without (controls) to see if people with diseases have higher (or lower) exposure. These studies have a higher potential for bias, but are cheaper and easier. An additional advantage is that such studies allow investigation of rare diseases without having to follow very large populations.
These studies describe patterns or trends on a geographic level and can be used to explore potential associations between community-level exposures and disease. However, ecological studies are the least informative, as they are unable to reliably estimate individual exposures.
What are the Limitations of Epidemiological Studies?
The results of epidemiological studies, whether they show an association or not, will often be affected by limitations of the study design or analysis. Results may be influenced by errors or unidentified bias in the data, the influence of other relevant factors, or by chance variation. In case control studies, bias in exposure assessments may arise in a subject’s ability to recall and report past exposures (recall bias). For example a case control study on mobile phone use will be limited by subject’s ability to remember their past mobile phone use
Another important issue is confounding, whereby an exposure is both a risk factor for the disease and associated with the exposure of interest
For example if a study was examining the association between alcohol and heart disease, cigarette smoking is a confounder because smoking is known to cause heart disease.
The more of these criteria that can be met the greater confidence there is that an agent causes the disease under study. For example, there may be a consistent association found in several epidemiological studies but if this is not backed up by evidence from studies in the laboratory, that is, biological plausibility, then it may not be possible to conclude that the agent caused the disease.
How are the results of epidemiological studies interpreted?
The results of epidemiological studies are given in terms such as relative risk for cohort studies or odds ratios for case control studies. These are statistical terms that give researchers information about the strength of any association detected between disease and exposure to the proposed risk factor.
The five main criteria for establishing a likely association are:
1. the strength of association - the relationship must be clear
2. consistency - repeatable in other study populations
3. temporality - exposure must come before the disease
4. plausibility - it must make sense biologically
5. biological gradient - dose-response relationship, more exposure equals more disease
All these criteria must be met before exposure to an agent can be said to be causing the disease. For example, there may be a consistent association found in several epidemiological studies but if this is not backed up by evidence from studies in the laboratory ie biological plausibility, then the agent cannot be said to be causing the disease