Are there other causes of the health outcome that we might not be looking at?
Sometimes there is a third element, besides the exposure and outcome we are studying, that could interfere with the study identifying the true relationship between the exposure and the health outcome. This situation is called "confounding" And those elements are called "confounders."
One of the most common confounders is smoking. People who smoke become exposed to many toxic contaminants - such as lead, arsenic, formaldehyde, vinyl chloride, and cadmium - by inhaling them directly into their lungs. Smoking has been associated with many diseases including lung cancer, oral cancer, kidney cancer, and cardiovascular disease. If people exposed to contaminants that may cause any of these diseases smoke, it may be not be clear if their disease is due to smoking or exposure to the contaminant in the environment.
Here are other examples of possible confounders:
- Neighborhood Location
Suppose people are concerned that exposure to a chemical coming from a facility in the neighborhood causes asthma. The facility is located in an industrial area of town. There are other industries in the neighborhood near the chemical facility that released different chemicals. These different chemicals are possible confounders because they also may have caused asthma.
- Age demographics
Some illnesses are more common among people during youth and others are more common later in life. Suppose the neighborhood near the facility has a younger population than the neighborhood far away from the facility. Studies have shown that asthma is more common among children than adults. Thus, the higher prevalence in asthma near the chemical facility may be due to the presence of more children in that neighborhood.
Imagine that the neighborhood near the facility is poorer than the neighborhood far away from the facility. Asthma has been linked with poverty and substandard housing conditions. Therefore, higher prevalence in asthma near the chemical facility may be due to substandard housing conditions in the low income neighborhood, and not exposure to the chemical emitted from the facility.
When we don't account for confounding, we may get false results. Luckily, it is often possible to avoid problems of confounding, if you are aware of the confounding factor and gather information on it. For example, if we have information on age of people in our study, we can make comparisons within the same age groups.
The question asked at this stage is:
- What are other possible causes of the health outcome, besides exposure to the contaminant?