Designs combining instrumental variables with case-control: estimating principal strata causal effects

Area of Science:

• Epidemiology
• Biostatistics
• Genetic Epidemiology

Background:

• Instrumental variables (IV) are widely used for causal inference in cohort studies.
• Combining IV with efficient designs like case-control sampling presents methodological challenges.
• Increasing use of Mendelian randomization and high data costs necessitate cost-effective sampling strategies.

Purpose of the Study:

• To evaluate the performance of standard instrumental variables analysis in combined case-control designs.
• To propose a novel statistical method for accurate causal effect estimation in such integrated designs.
• To address the need for efficient analysis of genetic data from cost-effective sampling.

Main Methods:

• Demonstrated the inadequacy of standard instrumental variables methods for case-control samples.
• Developed and presented a new methodological approach for causal effect estimation.
• Applied the proposed method to a relevant case study in oncology.

Main Results:

• Standard instrumental variables analysis yields biased causal effect estimates when combined with case-control sampling.
• The proposed method provides appropriate estimation of causal effects in these combined designs.
• The method's utility is confirmed through an illustrative oncology study.

Conclusions:

• The standard instrumental variables approach is not suitable for case-control designs.
• A new method is necessary and effective for estimating causal effects in combined IV and case-control studies.
• This advancement supports more efficient and cost-effective causal inference in genetic epidemiology.