On identification and estimation for sufficient cause interaction through a quasi-instrumental variable.

Area of Science:

• Epidemiology
• Biostatistics
• Pharmacology

Background:

• Mechanistic interaction investigates how exposures influence outcomes, with synergism being a key focus in genetic studies and pharmacology.
• Synergism, defined by the sufficient component cause model, is challenging to quantify directly.
• Sufficient Cause Interaction (SCI) is an alternative metric, but existing empirical tests have limitations in power and direct estimation.

Purpose of the Study:

• To propose a novel statistical method for estimating the probability of individual SCI.
• To introduce a quasi-instrumental variable to address limitations in current SCI estimation.
• To develop a more powerful hypothesis test for detecting synergistic interactions.

Main Methods:

• Introduction of a quasi-instrumental variable to model background conditions necessary for SCI.
• Development of a new statistical framework for estimating individual-level SCI.
• Formulation of a novel hypothesis test for synergistic interaction, comparing its power to existing methods.

Main Results:

• The proposed method provides a direct estimation of SCI probability.
• The new hypothesis test demonstrates increased statistical power compared to previous empirical tests.
• The method is applied to investigate synergistic effects of intestinal bacteria in Parkinson's disease etiology.

Conclusions:

• The novel method offers a more powerful approach to estimate and test for synergistic interactions (SCI).
• The quasi-instrumental variable facilitates direct estimation of SCI, overcoming limitations of prior approaches.
• This methodology has significant implications for understanding complex etiological mechanisms in diseases like Parkinson's.