Ascertainment bias in family-based case-control studies

Area of Science:

• Epidemiology
• Genetic Epidemiology
• Biostatistics

Background:

• Family-based case-control studies are common for investigating diseases with adult onset.
• These studies often match controls (siblings/cousins) on age and sex, irrespective of geographic residence.
• Geographic variation in environmental risk factors can complicate interpretation of study results.

Purpose of the Study:

• To evaluate the impact of control selection location on bias in family-based case-control studies.
• To assess the consistency of estimators for gene-environment interaction effects under geographic mismatch.
• To provide guidance on interpreting environmental main effects in such designs.

Main Methods:

• Utilized a family-matched case-control study design.
• Analyzed a population-based sample of cases from a defined geographic region.
• Compared controls selected within and outside the case ascertainment region.

Main Results:

• Estimates of environmental relative risk can be biased if controls are drawn from outside the case region.
• Estimators for gene and gene-environment interaction effects remain consistent, assuming independence of genotype and environment.
• Caution is recommended when interpreting environmental main effects from family-based studies with geographically mismatched controls.

Conclusions:

• Geographic selection of controls is critical for accurate estimation of environmental main effects in family-based case-control studies.
• Gene-environment interaction analyses may be less susceptible to geographic bias under certain assumptions.
• Researchers must carefully consider control selection strategies to avoid biased environmental risk factor assessment.