Moving beyond risk ratios in sibling analysis: estimating clinically useful measures from family-based analysis

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Family-based studies often rely solely on relative effect measures (odds or hazard ratios).
• This limits the interpretability of findings, especially for clinical and policy applications.
• Shared familial confounding can bias results in traditional analyses.

Purpose of the Study:

• To introduce and demonstrate the marginalized between-within (MBW) framework for family-based analyses.
• To enable estimation of absolute risks and clinically relevant metrics.
• To account for shared familial confounding more effectively.

Main Methods:

• Overview of sibling comparison methods and the MBW framework.
• Application of the MBW model to Swedish registry data on maternal smoking and infant mortality.
• Decomposition of effects into within- and between-family components with a global baseline.

Main Results:

• The MBW framework enables estimation of absolute quantities (cumulative risks, average treatment effects, attributable fractions, numbers needed to harm).
• Absolute measures derived from the MBW model provided more interpretable and policy-relevant insights than relative estimates alone.
• MBW model yields similar relative estimates to conditional logistic or stratified Cox regression but allows for absolute measures.

Conclusions:

• The MBW framework strengthens the interpretability of family-based analyses.
• It enables absolute and policy-relevant estimates for binary and time-to-event outcomes.
• This approach moves beyond the limitations of solely relying on relative effect measures.