Causation and association

Causation and association are two terms that are often used interchangeably. However, causation and association are two different concepts, and it is often difficult to distinguish one from the other. Epidemiology and risk assessment are two fields where causation and association are used. Epidemiology is the study of disease frequency and duration in populations, and deals with the incidence and control of disease. Risk assessment involves the process of identifying and estimating risk.
Epidemiological studies are conducted in both human and animal populations and try to study the nature of populations. This can be accomplished through the usage of observational and experimental studies, however, most often, observational studies are most commonly used. (A clinical trial is an example of an experimental study.) It can also be defined as "the relation between two events that holds when, given that one occurs, it produces, or brings forth, or determines, or necessitates the second". In order for there to be a causal relationship between two events, the second event must immediately follow the first.
Association
Association is defined as two events occurring together more or less often than expected by chance. Specifically, in epidemiology, the association is the strength of the relationship between a disease and its possible cause, and this is measured through either an odds ratio, risk ratio, rate ratio, or p-value.
Usage
Epidemiology
Epidemiology can’t prove causation, it can only demonstrate association or tendencies. David Hume, an 18th century Scottish philosopher and essayist stated that causation is a binary relationship between experiences, and when we continually see conjunction between two events, we conclude that there is a cause and effect relationship between the two events. In addition, epidemiologists often will call one factor of a study the “cause” when in actuality it is the main factor that they want to examine. There may be other factors that contribute to the cause, however; they are temporarily ignored, as it is outside scope of their study. Also, the presence of bias (modeling, selection, and information) and confounding also make it difficult to determine a true causal relationship.
Risk assessment
In risk assessments, determining causation of exposure(dose)-response relationships from data can be challenging. Although risk assessments are founded on dose-response models, it requires integrating information from diverse sources and disciplines including epidemiology, toxicology, and molecular biology. . In order to form a causal relationship, a risk assessor must draw sound inferences about causal relations from one or more observational studies, address and resolve biases that can affect a single multivariate empirical dose-response study, and apply the results to human health risks and benefits. Also, there may be ethical issues that may arise, such as infecting a healthy animal or individual with a potentially deadly disease.
Example 2
Causation can be used is in an individual person where the cause can be isolated with the effect. For example, if an individual falls on the ice and breaks their ankle, we can say that falling on the ice was the cause, and the fracture was the effect.
Another example would be if an example were to burn their hand by accidentally touching a hot stove burner. Again, we can say that the cause was touching the burner, and the effect was the burn.
Association
Example 1
Lamictal (lamotrigine) is a drug used in the treatment of epilepsy and bipolar disorder. While many side effects can occur from taking the drug, including a severe rash, dizziness, drowsiness, or blurred vision, it can only be shown that these side effects are associated with taking the drug, as not all patients will develop these side effects, and not all people who develop a severe rash, dizziness, drowsiness, or blurred vision, will be due to taking Lamictal, as there are other factors that could cause these symptoms.
A recent study conducted by the Food and Drug Administration concluded that taking Lamictal may cause aseptic meningitis, or an inflammation of the brain or spinal cord not caused by bacteria. However, since 1994, only 40 cases have occurred in those taking Lamictal. This study can only show an association between taking Lamictal and aseptic meningitis.
Example 2
A common example where causation is used inappropriately is in the relationship between smoking and cancer. Since not all smokers develop lung cancer, and some non-smokers develop lung cancer, we can only say that smoking is associated with cancer, even if the association is strong. There are other factors in addition to smoking that could influence a person's likelihood to develop lung cancer.

Main Menu

See Also

Main Menu