Abstract
Fecundability, the probability of conception in a month or in a menstrual cycle, varies across and within age groups for both women and men. Fertility treatment has become common in a number of countries including Japan, but its impact on the age pattern of fecundability is unknown. By utilizing the previously collected data on time to pregnancy (TTP) of Japanese couples trying to conceive their first child, the present study aimed to estimate fecundability and sterility by women's age and to assess how the estimates may differ by including or excluding assisted conceptions. Duration between discontinuing contraception and conception (including both natural and assisted) resulted in a live birth was called TTP-all, and the duration ending with natural conception was called TTP-natural. TTP-natural was censored when a participant received fertility consultation or treatment. A zero-inflated beta distribution model was used to estimate a proportion of sterile (zero probability of conception) and a distribution of fecundability for each age group. Parameters of the distribution were estimated using the maximum likelihood method. When TTP-all and TTP-natural were used, the sterile proportion of the whole sample was, respectively, 2% and 14%, and the median (interquartile range) of fecundability was, respectively, 0.10 (0.04, 0.19) and 0.11 (0.05, 0.19). The median (interquartile range) of fecundability was 0.18 (0.10, 0.29) for women aged 24 years or younger and 0.05 (0.02, 0.13) for 35-39 years old when TTP-all was used, and the estimates were quite similar with those based on TTP-natural: it was 0.18 (0.10, 0.29) for women aged 24 years or younger and 0.06 (0.00, 0.15) for 35-39 years old. Exclusion of assisted conceptions resulted in larger proportions of sterility, but it had little impact on median or interquartile ranges of fecundability estimates. Fecundability is overall lower at higher ages, while interquartile ranges are overlapping, suggesting that inter-individual variability of fecundability within an age group is as large as the variability across age groups.