[Roaming through methodology. XXXVII. Unintentional population stratification in genetic case-control studies].
Hottenga JJ., Houwing-Duistermaat JJ., van Duijn CM.
In the field of genetic epidemiology, the appropriate selection of patients and controls within a case-control study is essential to obtain accurate results. Within the genetic case-control study design, unintentional population stratification can cause confounding of results. The probability of this type of confounding arising is determined by the selection of cases and controls from different populations and by the differences in their genetic background. In follow-up studies, confounding by population stratification may also occur. Recent developments in the fields of statistics and genetics now make it reasonably simple to test for population stratification in case-control studies, and to prevent confounding by this stratification.