Related Experiment Videos
Log-linear models for the analysis of matched cohort studies
T R Holford1, M B Bracken, B Eskenazi
1Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, CT 06510.
American Journal of Epidemiology
|December 1, 1989
Summary
Conditional logistic regression is a standard for matched case-control studies. This study extends its application to cohort designs, adapting linear logistic and log-linear models for prospective data analysis.
Area of Science:
- Epidemiology
- Biostatistics
Background:
- Conditional logistic regression is widely used for matched case-control studies.
- Existing software for linear logistic and log-linear models can be adapted for these analyses.
Purpose of the Study:
- To describe the application of conditional logistic regression to cohort designs.
- To develop an approach for adapting linear logistic and log-linear models for prospective data analysis.
Main Methods:
- The study extends conditional logistic regression to cohort designs.
- Methods are developed to adapt linear logistic and log-linear models for prospectively collected data.
- Specific matching scenarios including matched pairs and 2:1 matching are discussed.
Main Results:
- The paper demonstrates the utility of conditional logistic regression in cohort studies.
- Adaptations of linear and log-linear models are presented for prospective data.
- Numerical examples illustrate the application of these methods.
Conclusions:
- Conditional logistic regression can be effectively applied to cohort designs.
- The developed methods provide a flexible approach for analyzing prospectively collected matched data.
- The techniques are applicable to various matching configurations in epidemiological studies.