Joint modeling of intercourse behavior and human fecundability using structural equation models.

Area of Science:

• Reproductive biology
• Biostatistics
• Human reproduction

Background:

• Human fecundability, the probability of conception per cycle, is crucial for understanding reproduction.
• Accurate modeling requires integrating biological and behavioral factors, including intercourse timing and frequency.
• Previous models often overlook the complex, day-to-day variations in intercourse behavior within a menstrual cycle.

Purpose of the Study:

• To propose a novel joint statistical model for human fecundability and intercourse behavior.
• To develop a model that accounts for the dynamic and heterogeneous nature of intercourse patterns during the fertile window.
• To enhance the accuracy of conception probability estimation by incorporating detailed behavioral data.

Main Methods:

• A joint model combining a classic conception probability model with a structural equation model (SEM).
• The SEM component allows for flexible modeling of dependencies between daily intercourse behaviors.
• Application to the New York State Angler Cohort Prospective Pregnancy Study for validation.

Main Results:

• The proposed model effectively accommodates diverse intercourse patterns and their dependencies across the menstrual cycle.
• It allows for the inclusion of general covariate effects, providing a more comprehensive analysis.
• Demonstrated the model's utility in a real-world prospective pregnancy study.

Conclusions:

• The joint modeling approach provides a more robust framework for studying human fecundability.
• Accounting for the day-specific nature of intercourse behavior significantly improves reproductive modeling.
• This methodology offers valuable insights into the interplay between behavior and conception probability.