Allen J Wilcox1, Quaker Harmon1, Kevin Doody2, Don P Wolf3, Eli Y Adashi4. 1. Epidemiology Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Durham, NC, USA. 2. Center for Assisted Reproduction, Bedford, TX, USA. 3. Dept. Obstetrics and Gynecology, Oregon Health and Science University, Portland, OR, USA. 4. Dept. Obstetrics and Gynecology, Brown University, Providence, RI, USA.
Abstract
STUDY QUESTION: What proportion of fertilized human ova are lost before implantation? SUMMARY ANSWER: An estimated 40 to 50% of fertilized ova fail to implant. WHAT IS KNOWN ALREADY: Preimplantation loss is not detectable with current technology. Published estimates of preimplantation loss range from 10 to 70%. STUDY DESIGN, SIZE, DURATION: We combine data from epidemiologic, demographic, laboratory and in vitro fertilization studies to construct an empirical framework for the estimation of preimplantation loss. This framework is summarized in a user-friendly Excel file included in supplement. PARTICIPANTS/MATERIALS, SETTING, METHODS: We draw from multiple sources to generate plausible estimates of fecundability, sterility, transient anovulation, intercourse patterns and the proportion of ova fertilized in the presence of sperm. We combine these estimates to generate a summary estimate of preimplantation loss. This estimate can be considered an average for couples in their prime reproductive years. MAIN RESULTS AND THE ROLE OF CHANCE: Under a plausible range of assumptions, we estimate that 40 to 50% of fertilized ova fail to implant. LIMITATIONS, REASONS FOR CAUTION: A crucial factor in estimating preimplantation loss is the probability that an ovum will be fertilized when exposed to sperm. Human data are available only from in vitro fertilization (IVF), which may not accurately represent events in vivo. We therefore assume a range of in vivo fertilization rates, from 64% (human IVF data) to 90% (mouse data). WIDER IMPLICATIONS OF THE FINDINGS: Our estimate of preimplantation loss takes into account the biological processes relevant to fertilization and loss. Using this empirical basis for estimation, we find support for the usual assumption that risk of loss is highest in the earliest days following fertilization. Furthermore, this framework can provide improved estimates as better reproductive data become available. To the extent that our estimates are accurate, more fertilized ova are apparently lost in vitro than in vivo, suggesting that further improvements in IVF success rates may be possible. STUDY FUNDING/COMPETING INTEREST(S): This study was supported by the Intramural Program of the National Institute of Environmental Health Sciences, NIH. Professor Adashi serves as Co-Chair of the Safety Advisory Board of Ohana Biosciences, Inc. The other authors have no competing interests. TRIAL REGISTRATION NUMBER: N/A.
STUDY QUESTION: What proportion of fertilized human ova are lost before implantation? SUMMARY ANSWER: An estimated 40 to 50% of fertilized ova fail to implant. WHAT IS KNOWN ALREADY: Preimplantation loss is not detectable with current technology. Published estimates of preimplantation loss range from 10 to 70%. STUDY DESIGN, SIZE, DURATION: We combine data from epidemiologic, demographic, laboratory and in vitro fertilization studies to construct an empirical framework for the estimation of preimplantation loss. This framework is summarized in a user-friendly Excel file included in supplement. PARTICIPANTS/MATERIALS, SETTING, METHODS: We draw from multiple sources to generate plausible estimates of fecundability, sterility, transient anovulation, intercourse patterns and the proportion of ova fertilized in the presence of sperm. We combine these estimates to generate a summary estimate of preimplantation loss. This estimate can be considered an average for couples in their prime reproductive years. MAIN RESULTS AND THE ROLE OF CHANCE: Under a plausible range of assumptions, we estimate that 40 to 50% of fertilized ova fail to implant. LIMITATIONS, REASONS FOR CAUTION: A crucial factor in estimating preimplantation loss is the probability that an ovum will be fertilized when exposed to sperm. Human data are available only from in vitro fertilization (IVF), which may not accurately represent events in vivo. We therefore assume a range of in vivo fertilization rates, from 64% (human IVF data) to 90% (mouse data). WIDER IMPLICATIONS OF THE FINDINGS: Our estimate of preimplantation loss takes into account the biological processes relevant to fertilization and loss. Using this empirical basis for estimation, we find support for the usual assumption that risk of loss is highest in the earliest days following fertilization. Furthermore, this framework can provide improved estimates as better reproductive data become available. To the extent that our estimates are accurate, more fertilized ova are apparently lost in vitro than in vivo, suggesting that further improvements in IVF success rates may be possible. STUDY FUNDING/COMPETING INTEREST(S): This study was supported by the Intramural Program of the National Institute of Environmental Health Sciences, NIH. Professor Adashi serves as Co-Chair of the Safety Advisory Board of Ohana Biosciences, Inc. The other authors have no competing interests. TRIAL REGISTRATION NUMBER: N/A.