EUVIMEDEUVIMED entdecken
Crossref

L26/P-540 Cross-population validation of expected live-birth probabilities in recurrent pregnancy loss (RPL) cohorts from Japan and Denmark

H. Yoshihara, R. Bato, D. Westergaard, A. Kolte, M. Sugiura-Ogasawara, H. Nielsen, J. Brosens

Human Reproduction · 01.07.2026 · 41 (Supplement_1)

DOI: 10.1093/humrep/deag083.873 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von H. Yoshihara und 6 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Study question Are expected live-birth probabilities derived from Danish nationwide registry data transferable to RPL cohorts in Japan and Denmark? Summary answer Registry-derived expected live-birth probabilities closely matched observed outcomes in Japanese and Danish RPL cohorts, supporting cross-population transferability. What is known already Risk of pregnancy loss increases with advancing maternal age and the number of previous pregnancy losses. Age-related risk is primarily driven by embryonic aneuploidy. In contrast, the stepwise increase in age-independent recurrence risk after each pregnancy loss has recently been linked to progressive erosion of the endometrial decidual reaction, which is essential for stable maternal–fetal interface formation. These findings raise questions about whether pregnancy loss risk is biologically hardwired and conserved across populations, or significantly modified by environmental, lifestyle, and comorbidity-related factors. Study design, size, duration This retrospective study comprised two components. First, demographic and clinical characteristics were compared between two tertiary prospective RPL cohorts from Japan (n = 1,465) and Denmark (n = 1,821). Second, observed live-birth rates were compared with expected live-birth probabilities based solely on participant-level age and number of prior pregnancy losses. Participants/materials, setting, methods RPL patients attending tertiary referral centres in Japan and Denmark were included using de-identified secondary-use datasets. Subject-level expected live-birth probabilities were calculated from a Danish nationwide registry-based cohort comprising 2,722,441 pregnancy outcomes in 1,285,230 women (1977–2017). Observed live-birth rates in the subsequent pregnancy were compared with expected values. Calibration analyses and summary performance metrics were generated, with robustness assessed using bootstrap resampling and repeated random subsampling. Main results and the role of chance The Danish (n = 1,821) and Japanese (n = 1,465) cohorts differed in several baseline characteristics. The Japanese cohort enrolled earlier (median year 2012 vs. 2014; p < 0.001), had a higher proportion of primary RPL (78% vs. 60%; p < 0.001), lower maternal BMI (21.2 vs. 24.7; p < 0.001), and fewer previous pregnancy losses at referral (2.67 vs. 3.89; p < 0.001), reflecting differences in national definitions. Maternal age at referral was similar (33.5 years vs 33.5 years; p = 0.5). Despite these differences, expected live-birth probabilities closely matched observed outcomes. In the Japanese cohort, the expected live-birth rate was 72.0%, compared with an observed rate of 71.3% (median difference: 0.7 percentage points; 95% CI: −0.96 to 1.44; p = 0.48). In the Danish cohort, expected and observed rates were 60.8% and 60.6%, respectively (95% CI: −1.57 to 0.23; p = 0.739). In the combined cohort, expected and observed rates were 65.2% and 65.3% (95% CI: −0.74 to − 0.04; p = 0.022). Calibration was strongest within the mid-to-high probability range (0.6–0.9). Bootstrap and subsampling analyses demonstrated decreasing variability with increasing sample size and stabilisation at n ≥ 400, indicating that the observed agreement is unlikely explained by chance alone. Limitations, reasons for caution Secondary analyses required harmonisation across datasets, and some clinical variables were defined differently between cohorts. These findings apply at the population level, at least between two countries with broadly comparable high socioeconomic status, and do not provide individual-level risk estimates for pregnancy loss in subsequent pregnancies. Wider implications of the findings Cross-population agreement between expected and observed live-birth rates suggests that registry-derived pregnancy outcome probabilities, stratified by maternal age and prior losses, can benchmark outcomes across diverse RPL cohorts, facilitating international comparisons and study planning. The findings further suggest that intrinsic biological processes determine pregnancy loss risk at a population level. Trial registration number No

Zitieren

APA

Yoshihara, H., Bato, R., Westergaard, D., Kolte, A., Sugiura-Ogasawara, M., Nielsen, H., & Brosens, J. (2026). L26/P-540 Cross-population validation of expected live-birth probabilities in recurrent pregnancy loss (RPL) cohorts from Japan and Denmark. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.873

Vancouver

Yoshihara H, Bato R, Westergaard D, Kolte A, Sugiura-Ogasawara M, Nielsen H, et al. L26/P-540 Cross-population validation of expected live-birth probabilities in recurrent pregnancy loss (RPL) cohorts from Japan and Denmark. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.873

BibTeX
@article{Yoshihara2026,
  author = {Yoshihara, H. and Bato, R. and Westergaard, D. and Kolte, A. and Sugiura-Ogasawara, M. and Nielsen, H. and Brosens, J.},
  title = {L26/P-540 Cross-population validation of expected live-birth probabilities in recurrent pregnancy loss (RPL) cohorts from Japan and Denmark},
  journal = {Human Reproduction},
  year = {2026},
  volume = {41},
  number = {Supplement_1},
  doi = {10.1093/humrep/deag083.873}
}
Kostenlos starten →