A score test for determining sample size in matched case-control studies with categorical exposure

Area of Science:

• Epidemiology
• Biostatistics
• Genetics

Background:

• Matched case-control studies are crucial for investigating disease associations.
• Existing methods for polychotomous (multiple category) exposures in 1:M matched studies are limited.
• Assessing disease risk associated with categorical and ordinal exposures requires robust statistical tools.

Purpose of the Study:

• To develop a score test for assessing disease association with polychotomous categorical exposures in 1:M matched case-control studies.
• To determine the sample size required for detecting specific departures from the null hypothesis of no association.
• To propose a test for detecting trends in disease risk with ordinal exposure variables.

Main Methods:

• Development of a score test statistic for hypothesis testing.
• Utilizing the test's power function to calculate the required number of matched sets for sample size determination.
• Application of methods to both categorical and ordinal exposure variables.

Main Results:

• A novel score test is proposed for analyzing polychotomous exposure-disease associations in 1:M matched case-control studies.
• The study provides a method to calculate the necessary sample size for detecting associations.
• A specific test for detecting trends in disease risk with ordinal exposures is presented.

Conclusions:

• The proposed score test offers a general solution for analyzing polychotomous exposures in 1:M matched case-control studies.
• The sample size calculation method aids in efficient study design.
• The methods are validated using real and simulated data, including disease-gene association studies.