Invited Commentary: Two-Phase, Generalized Case-Control Designs for Quantitative Longitudinal Outcomes and Evolution...

Area of Science:

• Epidemiology
• Biostatistics
• Public Health Research

Background:

• The case-control study design has evolved over 50 years, moving beyond the rare disease assumption for risk ratio estimation.
• Modifications allow for rate ratio estimation and using individuals as their own controls (e.g., case-cohort, case-crossover studies).
• Efficiency gains are now recognized through outcome-dependent sampling for dichotomous and continuous outcomes in longitudinal studies.

Purpose of the Study:

• To summarize recent advances in case-control study design and analysis.
• To highlight methods enabling more general and efficient use of case-control studies.
• To comment on the application of generalized case-control methods to causal inference.

Main Methods:

• Review and commentary on evolved case-control study designs and analysis techniques.
• Discussion of outcome-dependent sampling for dichotomous and continuous outcomes.
• Exploration of generalized case-control methods for longitudinal data and causal inference.

Main Results:

• Case-control studies can be highly efficient, even without the rare disease assumption.
• Outcome-dependent sampling increases efficiency, particularly for estimating associations with trajectories and patterns of change over time.
• Recent advances allow for broader and more efficient application of case-control designs.

Conclusions:

• Generalized case-control methods offer significant efficiency improvements in epidemiological research.
• These advanced designs and analyses are crucial for understanding exposure-outcome relationships, especially in longitudinal studies.
• The application of these methods extends to strengthening causal inference.