(B-279) The association between childhood asthma and frozen versus fresh embryo transfer in assisted reproductive technology: a nationwide retrospective cohort study using target trial emulation framework

Background: Childhood asthma is rising globally and is reported even more often in children conceived by assisted reproduction. Fresh embryo transfer exposes the embryo to supraphysiologic hormones produced by ovarian stimulation; these hormones can alter endometrial receptivity and epigenetically shift immune development toward type-2 inflammation, potentially raising asthma risk. Frozen embryo transfer avoids this exposure, but evidence comparing the two approaches is limited and inconsistent.

Objectives: To compare the risk of childhood asthma after frozen versus fresh embryo transfer using nationwide registry data.

Methods: We emulated a target trial by linking Taiwan’s national ART registry, birth registry, and National Health Insurance Research Database. All embryo-transfer cycles performed from 2013 to 2020 were included. Propensity scores with inverse-probability-of-treatment weighting balanced baseline characteristics between frozen and fresh transfer groups. The outcome was physician-diagnosed asthma (ICD codes), followed up to 31 December 2022; analyses followed a per-protocol contrast and reported relative risk (RR) per live birth. Sensitivity analyses tested age-specific outcome windows, prevalence definitions, asthma diagnosis with related medication, and differential follow-up. Subgroup analyses were stratified by ovarian response (≥ 10 vs. < 10 oocytes retrieved).

Results: The study included 47,020 frozen and 34,236 fresh cycles (mean maternal ages 36.7 and 37.4 years, respectively; SMD, <0.01 after weighting). In the weighted comparison, childhood asthma risk did not differ between groups (RR = 1.02, 95 % CI 0.97–1.07). All sensitivity analyses supported this null result. In subgroup analyses, frozen transfer was associated with a significantly lower asthma risk among normal- or high-response women (≥ 10 oocytes) (RR = 0.84; 95% CI 0.77-0.92), whereas the opposite appeared in poor responders (RR = 1.13; 95% CI 1.06-1.21).

Conclusions: Overall, we did not observe the significant difference in asthma risk in overall population. However, in women with normal to higher ovarian response, significant lower risk of asthma was observed.