Statistical Modeling of Sensitive Period Effects Using the Structured Life Course Modeling Approach (SLCMA)

Area of Science:

• Developmental biology
• Biostatistics
• Epidemiology

Background:

• Sensitive periods are critical developmental windows where experiences significantly impact life outcomes.
• Current research methods often overlook sensitive periods, favoring simpler exposure hypotheses.
• A lack of sophisticated statistical tools hinders the study of time-dependent exposure-outcome relationships.

Purpose of the Study:

• To introduce and provide a guide for the structured life course modeling approach (SLCMA).
• To enable researchers to model complex life course hypotheses, specifically sensitive periods.
• To determine which hypothesis best explains variation in repeated exposure-outcome relationships.

Main Methods:

• The SLCMA utilizes the least angle regression (LARS) variable selection technique.
• LARS is a type of least absolute shrinkage and selection operator (LASSO) estimation.
• The approach incorporates selective inference for unbiased effect estimates and confidence intervals.

Main Results:

• The SLCMA facilitates the creation of parsimonious models for exposure-outcome relationships.
• It allows for the identification and modeling of sensitive periods in developmental research.
• Provides a framework for robust statistical inference on selected models.

Conclusions:

• The SLCMA is a powerful tool for investigating sensitive periods in developmental research.
• It addresses the limitations of existing methods for studying time-dependent exposures.
• Implementation guidance and R code are provided to facilitate its use.