A multistate model of fecundability and sterility

Area of Science:

• Demography
• Reproductive Epidemiology
• Biostatistics

Background:

• Estimating fecundability and sterility is crucial for understanding natural fertility patterns.
• Existing models often lack the granularity to distinguish different sterility states or model fecundability heterogeneity effectively.

Purpose of the Study:

• To develop and apply a multistate hazards model for estimating fecundability and sterility.
• To analyze the impact of preexisting versus acquired sterility on fertility.
• To model the distribution of fecundability among nonsterile couples.

Main Methods:

• Development of a multistate hazards model incorporating distinct sterile and nonsterile states.
• Distinguishing between sterility present before conception attempts and sterility developing during exposure.
• Modeling fecundability as log-normally distributed within the nonsterile population.
• Application to data on first birth intervals from diverse populations (Taiwan, Sri Lanka, Amish).

Main Results:

• Heterogeneity in fecundability was statistically significant across most age groups.
• Both preexisting and newly acquired sterility were found to be unimportant determinants of fertility before age 40.
• The model effectively estimates fecundability and identifies distinct sterility states.

Conclusions:

• Sterility's role as a determinant of natural fertility appears to increase significantly only at later reproductive ages.
• The developed model provides a more nuanced understanding of factors influencing conception probabilities.
• Fertility variation is substantial, influenced by age-related changes in fecundability and sterility onset.