|
In
studies of the etiology of chronic disease, the case-control design
has been the most extensively used epidemiologic method over the
last twenty years. This design has also been extensively used
in epidemiologic investigations of TMD.
Case-control
studies are a cost-efficient method of studying the association
of a putative risk factor with disease onset. Because of a large
number of threats to the validity of case-control studies, the
use of this design is sometimes controversial (Feinstein,
1979). However, the successes of the case-control method in
identifying causal processes that were later confirmed by more
rigorous methods have been well documented. Schlesselman
(1982) provides a thorough treatment of the methods of the
case-control study. For a useful synopsis of the methods, see
Greenberg
and Ibrahim (1985).
The elements
of a case-control study include:
A sample
of recent onset cases of the condition of interest;
A sample
of controls selected in a way that does not introduce sampling
bias relative to cases with respect to risk factors; and
Information
gathered on cases and controls using the same methods.
The resulting
data are analyzed through contingency table analysis. In the simplest
case, a two by two table is constructed in which the columns represent
persons with or without the condition under study and the rows
represent persons with or without the risk factor of interest.
Case-control
studies are used to estimate the odds ratio for risk factors of
interest. The odds ratio measures the risk of developing the disease
or condition in a group with a particular factor relative to the
risk of the disease or condition in a group without the risk factor
of interest. Multivariate extensions of contingency table analysis
allow for estimation of odds ratios while controlling for other
variables (Bishop et al., 1978).
While the
case-control design presents difficulties in determining the order
of causation, a well designed and executed case-control study
can provide scientific data that are quite valuable.
Example
|
|---|
| In
an analysis of data from a large health maintenance
organization, LeResche et al. (1997) compared exogenous
hormone use, in the form of oral contraceptives or hormone
replacement therapy, among women referred for treatment
of TMD pain and control women of the same age who were
not referred for treatment. Exposure to oral contraceptives
in young women, and exposure to hormone replacement
therapy with estrogen (but not progestin) in older women
were associated with increased risk of TMD pain, after
controlling for overall use of health services. |
|
|