Real-time intrauterine dissolved oxygen dynamics across the luteal phase: a potential non-invasive functional biomarker of endometrial status.

Eight healthy women aged 22–35 years were included in the study. All participants reported regular menstrual cycles, with a mean cycle length of approximately 29 days. Body mass index values were within the normal range, and all participants exhibited normal gonadotropin profiles and ovarian reserve parameters, with a mean antral follicle count of 25 and average serum anti-Müllerian hormone (AMH) levels of approximately 3 ng/mL. Regarding reproductive history, two participants had previously delivered one child, and one participant reported a prior spontaneous miscarriage. None of the participants were smokers. Intrauterine oxygen tension (pO₂) was measured serially across the luteal phase, from LH + 0 to LH + 14, with assessments scheduled approximately every 48 h. Individual temporal profiles of pO₂ revealed two distinct patterns of intrauterine oxygen dynamics (Fig.  1 ): a Peak pattern ( n  = 4), characterized by a transient mid-luteal rise in pO₂, and a No-peak pattern ( n  = 4), in which pO₂ remained consistently low throughout the observation period. Fig. 1 Temporal profiles of intrauterine dissolved oxygen concentration (pO₂, Torr) across the luteal phase. The x-axis represents days relative to the luteinizing hormone surge (LH+ days), ranging from LH+0 to LH+14, and the y-axis shows intrauterine oxygen levels (0–50 Torr). Each colored line represents an individual participant, with mean pO₂ values and error bars indicating standard deviations at each time point. pO₂ values represent averages obtained from a 5-minute continuous recording period at each time point.Participants are categorized into two distinct response patterns: A Peak pattern – characterized by a mid-luteal rise in oxygen concentration (participants P002, P003, P004, and P008). B No-peak pattern – characterized by consistently low or stable pO₂ levels throughout the luteal phase (participants P001, P005, P006, and P007) Temporal profiles of intrauterine dissolved oxygen concentration (pO₂, Torr) across the luteal phase. The x-axis represents days relative to the luteinizing hormone surge (LH+ days), ranging from LH+0 to LH+14, and the y-axis shows intrauterine oxygen levels (0–50 Torr). Each colored line represents an individual participant, with mean pO₂ values and error bars indicating standard deviations at each time point. pO₂ values represent averages obtained from a 5-minute continuous recording period at each time point.Participants are categorized into two distinct response patterns: A Peak pattern – characterized by a mid-luteal rise in oxygen concentration (participants P002, P003, P004, and P008). B No-peak pattern – characterized by consistently low or stable pO₂ levels throughout the luteal phase (participants P001, P005, P006, and P007) Among participants exhibiting the Peak pattern (Fig.  1 A), three volunteers (P002, P003, and P004) showed low intrauterine pO₂ values (< 15 Torr) during the early luteal phase (LH + 1 to LH + 2), followed by a marked increase between LH + 4 and LH + 6 ( p  < 0.0001, Tukey’s test). pO₂ values reached peak levels of approximately 40–45 Torr, remained transiently stable between LH + 6 and LH + 7, and subsequently declined by LH + 8 ( p  < 0.0001). One participant (P008) demonstrated an earlier and shorter oxygenation peak, with pO₂ rising between LH + 3 and LH + 5 to comparable peak values (40–45 Torr; p  < 0.0001), followed by a rapid decline after LH + 7 ( p  < 0.0001), suggesting a temporally advanced luteal oxygen profile. In contrast, four participants (P001, P005, P006, and P007) exhibited relatively stable intrauterine oxygen levels throughout the luteal phase (Fig.  1 B). In this subgroup, pO₂ values generally remained below 35 Torr, with no consistent mid-luteal increase, indicating the absence of the transient oxygen peak observed in the Peak pattern subgroup. Potential physiological, pharmacological, and lifestyle-related factors associated with the No-peak pattern are summarized in Table  2 . Table 2 Summary of factors identified in participants exhibiting a No-peak intrauterine oxygen pattern Volunteer Identified factor Contextual description 1 Subclinical thyroid dysfunction with shortened luteal phase Presence of subclinical thyroid dysfunction and a shortened luteal phase (10 days), identified during post hoc review of clinical data 5 Lifestyle-related factors during measurement cycle Reported sleep deprivation during the study cycle 6 Pharmacological treatment (mesalazine) Ongoing mesalazine therapy for ulcerative colitis during the study period 7 No identifiable clinical or systemic factors No detectable systemic, endocrine, or gynecological pathology identified during screening or follow-up Summary of factors identified in participants exhibiting a No-peak intrauterine oxygen pattern Box-and-whisker plots summarizing intrauterine oxygen tension across luteal days are shown in Fig.  2 . Fig. 2 Box-and-whisker plots of intrauterine oxygen concentration (pO₂, Torr) across the luteal phase. The x-axis represents days relative to the luteinizing hormone surge (LH+ days), and the y-axis shows intrauterine oxygen tension (0–60 Torr). Each data point represents the average intrauterine pO₂ obtained from a 5-minute continuous recording period for each participant at each time point. Boxes indicate the interquartile range (25th–75th percentiles), with the median shown as a horizontal line; whiskers extend to the 5th–95th percentiles. Individual data points are displayed, and mean values are indicated by crosses. A Peak pattern. B Early and shorter oxygen peak. C No-peak pattern Box-and-whisker plots of intrauterine oxygen concentration (pO₂, Torr) across the luteal phase. The x-axis represents days relative to the luteinizing hormone surge (LH+ days), and the y-axis shows intrauterine oxygen tension (0–60 Torr). Each data point represents the average intrauterine pO₂ obtained from a 5-minute continuous recording period for each participant at each time point. Boxes indicate the interquartile range (25th–75th percentiles), with the median shown as a horizontal line; whiskers extend to the 5th–95th percentiles. Individual data points are displayed, and mean values are indicated by crosses. A Peak pattern. B Early and shorter oxygen peak. C No-peak pattern Within the Peak pattern subgroup, three participants displayed a progressive rise in pO₂ between LH + 4 and LH + 6, reaching median values of approximately 40–45 Torr, followed by stabilization at LH + 6–7 and a significant decline by LH + 8 (Fig.  2 A; **** p  < 0.0001, Tukey’s test). A distinct early-peak profile was observed in one participant, with pO₂ increasing between LH + 3 and LH + 5 and declining rapidly thereafter (Fig.  2 B; **** p  < 0.0001). By contrast, participants in the No-peak subgroup maintained consistently low pO₂ levels across the luteal phase, with median values ranging from approximately 5 to 25 Torr and no discernible mid-luteal elevation (Fig.  2 C). The intrauterine oxygen measurement procedure was well tolerated. Reported visual analog scale (VAS) scores ranged from 0 to 2, indicating minimal discomfort. No procedure-related adverse events were observed.

This was a prospective, observational, single-center feasibility study conducted at Vall d’Hebron University Hospital (Barcelona, Spain). Eight healthy women aged 18–35 years with regular menstrual cycles were recruited between January and June 2022. Eight healthy female volunteers from the general population (e.g., university students, hospital staff, egg donors) aged 18–35 years were enrolled. Eligibility required normogonadotrophic status, regular menstrual cycles, and body mass index (BMI) < 30. All participants underwent baseline clinical and laboratory screening to identify major endocrine disorders, chronic inflammatory or infectious diseases, and anatomical abnormalities. Transvaginal ultrasound was performed to evaluate pelvic anatomy and ovarian morphology. Only participants without clinically significant conditions fulfilling exclusion criteria, summarized in Table  1 , were enrolled. Table 1 Summary of exclusion criteria applied during participant selection Exclusion Criteria Description Medical pathology • Insulin-dependent diabetes mellitus • Cushing’s syndrome • Uncontrolled thyroid dysfunction • Hepatic and/or renal insufficiency • Any condition contraindicating ovarian stimulation and/or pregnancy • Antiphospholipid syndrome Uterine pathology • Endometriosis • Uterine cancer • Congenital malformations • Endometrial polyps • Uterine fibroids Smoker • Current smoker Recent hormonal treatment • Use of oral contraceptives or intrauterine device within the last 3 months Language proficiency • Inadequate understanding (oral and written) of the Spanish language Summary of exclusion criteria applied during participant selection • Insulin-dependent diabetes mellitus • Cushing’s syndrome • Uncontrolled thyroid dysfunction • Hepatic and/or renal insufficiency • Any condition contraindicating ovarian stimulation and/or pregnancy • Antiphospholipid syndrome • Endometriosis • Uterine cancer • Congenital malformations • Endometrial polyps • Uterine fibroids Eligible women received verbal and written information about the study from consultant gynecologists and provided written informed consent. Participation was voluntary, and participants could withdraw at any time. The trial protocol was approved by the Vall d’Hebron University Hospital Ethics Committee (Comitè d’Ètica de la Investigació amb Medicaments, CEIm) on 15 October 2021 (approval number: PR(AMI)336/2021) under the title “La concentración de oxígeno disuelto intrauterino (y en canal cervical) como marcador de la ventana de implantación: Validación clínica.” All procedures were conducted in accordance with the 1964 Declaration of Helsinki and its later amendments. Between cycle days 3 and 5, participants attended the Reproductive Medicine Unit at Vall d’Hebron Hospital for baseline assessment. A high-resolution transvaginal ultrasound was performed to evaluate uterine anatomy and rule out abnormalities. Participants performed daily home urinary LH testing (Easy@home, Premom) from cycle days 10 to 17 to identify the LH surge (defined as LH + 0). After surge detection, participants attended serial study visits across the luteal phase, scheduled approximately every 48 h from LH + 0/1 to LH + 13/14, depending on individual timing and logistical constraints. At each visit, intrauterine oxygen measurements were performed. Ovulation was corroborated by ultrasound evidence of follicular rupture and/or corpus luteum development. Dissolved oxygen at the uterine fundus was quantified using a modified version of the fiber-optic microsensor technique described by Ottosen et al. [ 45 ]. Briefly, rather than withdrawing the catheter and leaving the sensor resting against the endometrium, we used a rounded-tip embryo-transfer catheter (Labotect Guidance Catheter No. 13365) and secured the microsensor (PreSens IMP-PSt7) with a rotating Luer-lock adapter. This configuration kept the sensor tip exposed while preventing contact with the endometrial surface. Under abdominal ultrasound guidance, the Labotect catheter was advanced to approximately 1 cm short of the fundal endometrium. The Luer-lock was then rotated, and intrauterine oxygen tension was recorded every second for 5 min using an OXY-1-ST oximeter (PreSens). Intrauterine pO₂ was recorded continuously for 5 min at each time point, with measurements acquired every second. Reported values correspond to averaged pO₂ over the recording period to reduce the impact of short-term fluctuations. Each participant was assessed once per time point within a single menstrual cycle, and no repeated measurements were performed within the same day or across multiple cycles. Participant discomfort during each intrauterine measurement was assessed using a visual analog scale (VAS), where 0 indicated no pain and 10 represented maximum pain. The primary outcome was intrauterine oxygen tension (pO₂) at the uterine fundus across the luteal phase, reported in Torr. Intrauterine oxygen tension (pO₂) is reported in Torr (1 Torr = 1 mmHg). For comparison with international units, values can be converted to pascals (1 Torr = 133.322 Pa) or expressed as a percentage of atmospheric oxygen at sea level using O₂ (%) = (pO₂ [Torr] / 760) × 100. LH + 0 was defined as the day of the urinary luteinizing hormone (LH) surge, which precedes ovulation by approximately 24–36 h. LH + 1 corresponds to the first post-ovulatory day. Ovulation was identified by detection of a urinary LH surge and corroborated by transvaginal ultrasound demonstrating follicular rupture and/or corpus luteum formation. The luteal phase was defined based on ultrasonographic evidence of ovulation, including visualization of the corpus luteum and a secretory endometrial pattern. Given the exploratory nature of this pilot feasibility study, no formal sample size calculation was performed. Intrauterine oxygen tension (pO₂) is presented as box-and-whisker plots (Torr), with boxes indicating the interquartile range (25th–75th percentiles), whiskers spanning the 5th–95th percentiles, the median shown as a horizontal line, the mean marked with a “+,” and individual points plotted for values outside the whiskers. Outliers were identified and excluded using the Robust Regression and Outlier Removal (ROUT) method. Temporal changes in pO₂ across luteal days were assessed by one-way ANOVA with Tukey’s post hoc comparisons in GraphPad Prism v9.3.1.

This study provides an initial in vivo characterization of intrauterine oxygen dynamics across the luteal phase in healthy women. Serial intrauterine pO₂ measurements revealed two distinct temporal oxygenation patterns across the luteal phase. Previous studies have demonstrated that following ovulation, the endometrium undergoes marked vascular remodeling, including spiral artery growth and coiling, maturation of the subepithelial capillary plexus, and increased vascular branching. These processes coincide with progesterone-driven endometrial differentiation and angiogenic signaling [ 46 – 49 ]. Such vascular adaptations are thought to support local perfusion and metabolic demands within the endometrium, including increased oxygen availability within the uterine cavity. Consistent with this, animal models have reported cyclical increases in intrauterine oxygen tension around the peri-implantation period [ 36 – 38 ] and earlier human studies have observed higher intrauterine pO₂ during the luteal phase [ 40 ]. In the present study, participants displaying the Peak pattern showed a reproducible temporal sequence characterized by low intrauterine oxygen tension during the early luteal phase, followed by a sharp mid-luteal rise, a brief plateau, and a subsequent decline. This profile temporally coincides with the expected period of enhanced endometrial vascularization and may be consistent with coordinated angiogenic and perfusion changes associated with endometrial functional readiness. One participant exhibited an earlier and shorter oxygen peak, suggesting interindividual variability in the timing of endometrial physiological changes. In contrast, participants exhibiting the No-peak pattern did not demonstrate a mid-luteal increase in intrauterine pO₂, with values remaining relatively low throughout the observation period. Although this pilot study was not designed to establish causality, a structured post hoc review of baseline screening information (including the initial clinical and laboratory assessment performed prior to study initiation) together with follow-up discussion with participants identified plausible physiological, pharmacological and lifestyle-related factors that could disrupt endometrial vascular maturation and thereby influence intrauterine oxygen dynamics. These included subclinical thyroid dysfunction with a shortened luteal phase, ongoing mesalazine treatment, and acute lifestyle factors during the measurement cycle, such as sleep deprivation (Table  2 ). While exploratory and hypothesis-generating, these observations support the notion that intrauterine pO₂ profiling may be sensitive to systemic and environmental modulators of endometrial function, warranting confirmation in larger, adequately powered studies. The integration of intrauterine oxygen profiling with molecular characterization of the endometrial environment represents an important area for future research. Angiogenic factors such as vascular endothelial growth factor (VEGF) play a central role in endometrial vascular remodeling and implantation [ 50 ], and can be assessed in uterine fluid using multiplex immunoassays [ 51 ]. Correlating intrauterine pO₂ dynamics with local molecular profiles could provide a more comprehensive understanding of endometrial physiology and help support the biological relevance of oxygen-based measurements. The factors identified in the No-peak subgroup may be interpreted in light of existing literature. Thyroid dysfunction, prostaglandin inhibition, and lifestyle factors such as sleep deprivation have been associated with altered vascular and hormonal regulation (including changes in angiogenic mediators such as VEGF) [ 52 – 56 ]. Although these observations do not establish causality, they suggest that intrauterine oxygen dynamics may be sensitive to systemic and environmental influences. Hypoxia has been implicated in endometrial physiology through the regulation of angiogenesis, inflammation, and tissue remodeling, processes that are closely linked to the establishment of endometrial receptivity [ 57 – 60 ]. In this context, the temporal changes in intrauterine oxygen observed in the present study may reflect underlying vascular and metabolic dynamics within the endometrium. Embryonic metabolic requirements also support a role for dynamic intrauterine oxygen regulation. Early embryos rely on predominantly anaerobic metabolism, transitioning toward oxygen-dependent metabolism at the blastocyst stage [ 41 – 44 ]. Accordingly, a low-oxygen environment followed by a transient increase may be physiologically relevant. Several limitations of this study should be acknowledged. The small sample size and single-center design limit the generalizability of the findings and preclude definitive conclusions regarding clinical outcomes. In addition, the study population consisted of a small and selected group of healthy volunteers, including hospital staff, medical students, and egg donors, which may further limit the generalizability of the findings to broader infertility populations. Each participant was evaluated during a single menstrual cycle, preventing assessment of intra-individual reproducibility across cycles. In addition, intrauterine oxygen measurements required catheter-guided sensor placement, which may introduce variability related to uterine anatomy or probe positioning. Although major confounding factors were screened for, unmeasured physiological, pharmacological, or lifestyle-related influences may have affected endometrial vascularization and intrauterine oxygen dynamics. In addition, molecular characterization of uterine fluid (e.g., angiogenic factors such as VEGF or cytokines) was not performed, which could have strengthened the biological interpretation of intrauterine oxygen dynamics. Despite these limitations, this pilot feasibility study suggests that real-time intrauterine pO₂ profiling is technically feasible, well tolerated, and capable of identifying distinct oxygenation patterns across the luteal phase. As a feasibility study, this work was conducted in healthy volunteers to characterize baseline intrauterine oxygen dynamics under physiological conditions. By providing a direct, functional assessment of intrauterine conditions within the same cycle, intrauterine oxygen profiling may complement existing methods for evaluating endometrial receptivity. From a clinical perspective, if validated in IVF populations, this approach could provide a novel real-time strategy to assess endometrial functional status and help identify a physiologically optimal window for embryo transfer based on local endometrial conditions rather than fixed temporal criteria. Larger, prospective studies are warranted to determine whether intrauterine oxygen dynamics can predict implantation success, clinical pregnancy, and live birth outcomes.

This feasibility study provides the first in vivo description of dynamic intrauterine oxygen fluctuations across the luteal phase in healthy women. These preliminary temporal patterns suggest that intrauterine oxygen tension may reflect physiologically relevant changes in endometrial status. Although no clinical extrapolations can be made from this small pilot cohort, real-time intrauterine oxygen assessment may represent a promising approach for future evaluation of endometrial readiness. Further studies in IVF populations are required to determine its relationship with implantation, clinical pregnancy, and live birth outcomes.

Infertility affects nearly 190 million people worldwide. Despite major advances in assisted reproductive technologies, live-birth rates per embryo transfer remain suboptimal. Large European registry data report overall live-birth rates ranging between 40% and 50% per transfer, although outcomes vary substantially depending on maternal age, with rates exceeding 60% in women under 35 years and declining to below 10% in women over 42 years [ 1 , 2 ]. While maternal age remains one of the main determinants of reproductive success, the introduction of preimplantation genetic testing for aneuploidy (PGT-A) has significantly improved embryo selection by enabling the transfer of euploid embryos. However, even under these optimized embryonic conditions, implantation failure persists, highlighting the contribution of non-embryonic factors, particularly endometrial receptivity and embryo–endometrium synchrony [ 3 – 7 ]. The period during which the endometrium is generally considered more receptive (often referred to as the window of implantation, WOI) is typically estimated to occur during the luteal phase, around days 19 to 21 of a standard menstrual cycle [ 8 – 10 ]. In hormone replacement therapy cycles, the WOI is conventionally assumed to occur after five days of progesterone exposure (P4 + 5) [ 11 ]. However, accumulating evidence indicates that both the timing and duration of the WOI may vary between individuals and even between cycles within the same woman, with reported displacement in up to 30% of IVF patients and higher rates in those with repeated implantation failure [ 12 , 13 ]. Hormonal, anatomical, and molecular approaches have been extensively investigated to assess endometrial receptivity [ 14 , 15 ]. Although estradiol and progesterone play essential roles in regulating endometrial proliferation and differentiation [ 16 – 18 ], circulating hormone levels alone do not adequately reflect the dynamic and functional state of the endometrium. Similarly, imaging techniques, histological evaluation, and Doppler-based vascular assessments have shown limited predictive value for implantation outcomes [ 15 , 19 ]. More recently, advanced five-dimensional Power Doppler ultrasound has been proposed as a promising tool to exclude patients with low implantation potential; however, these findings require further validation in larger studies [ 20 ]. Advances in “omics” technologies have enabled large-scale profiling of gene expression, proteins, and microRNAs associated with endometrial receptivity [ 21 ]. However, most of these approaches rely on endometrial biopsy, require delayed analysis, and have not consistently demonstrated improvements in live birth outcomes when used to guide personalized embryo transfer [ 22 – 29 ]. More recently, molecular profiling of endometrial fluid has been explored through the analysis of free and extracellular vesicle–associated microRNAs [ 30 – 33 ]. In parallel, transcriptomic analyses of RNA expelled into uterine fluid have suggested that specific gene expression signatures may help distinguish between receptive and non-receptive endometria [ 34 ]. Although promising, these approaches still require further validation and standardization before routine clinical implementation. Taken together, current approaches to assess endometrial receptivity remain limited by invasiveness, delayed readouts, or an inability to capture the dynamic and localized nature of endometrial function within a given cycle. Endometrial receptivity is a temporally regulated and functional state that depends on coordinated vascular, metabolic, and cellular processes, which may not be adequately reflected by static or indirect measurements. In this context, intrauterine biophysical parameters, such as oxygen tension, carbon dioxide, pH, and temperature, represent attractive candidates for real-time assessment of endometrial function. These parameters reflect fundamental aspects of tissue metabolism, vascularization, and cellular homeostasis, and must be synchronized with the metabolic and developmental requirements of the embryo during early development [ 35 ]. Animal studies have reported cyclical increases in intrauterine oxygen tension around the peri-implantation period [ 36 – 38 ]. In addition, earlier studies in women have described intrauterine oxygen variations across the menstrual cycle [ 39 , 40 ], although these measurements were limited by the available technology at the time. Embryos undergo a metabolic transition from predominantly anaerobic metabolism during cleavage to increased oxygen-dependent metabolism at the blastocyst stage [ 41 – 44 ]. Accordingly, intrauterine oxygen levels are expected to remain low during early cleavage stages, minimizing oxidative stress, and to increase around the time of implantation to accommodate the rising metabolic demands of the blastocyst. Given that dissolved oxygen levels in endometrial fluid may reflect local vascular development and tissue integrity, we hypothesized that profiling intrauterine pO₂ during the luteal phase could provide a direct, functional assessment of endometrial status and embryo–endometrium synchrony. To explore this concept, we applied a fiber-optic oxygen microsensor technique adapted from Ottosen et al. [ 45 ] to characterize intrauterine pO₂ dynamics across the luteal phase in healthy women.

Additional file 1: Supplementary Table S1. Individual descriptive statistics of intrauterine pO₂ measurements (Torr) for participants exhibiting a Peak oxygenation pattern (P2, P3, P4, P8), including number of values ( n ), median (interquartile range), and mean ± standard deviation by luteal day. Additional file 1: Supplementary Table S1. Individual descriptive statistics of intrauterine pO₂ measurements (Torr) for participants exhibiting a Peak oxygenation pattern (P2, P3, P4, P8), including number of values ( n ), median (interquartile range), and mean ± standard deviation by luteal day. Additional file 2: Supplementary Table S2. Individual descriptive statistics of intrauterine pO₂ measurements (Torr) for participants exhibiting a No-peak oxygenation pattern (P1, P5, P6, P7), including number of values ( n ), median (interquartile range), and mean ± standard deviation by luteal day. Additional file 2: Supplementary Table S2. Individual descriptive statistics of intrauterine pO₂ measurements (Torr) for participants exhibiting a No-peak oxygenation pattern (P1, P5, P6, P7), including number of values ( n ), median (interquartile range), and mean ± standard deviation by luteal day. Additional file 3: Supplementary Table S3. Group-level descriptive summary of intrauterine pO2 (Torr). Values represent mean of participant medians (range of medians across participants). Raw per-visit oxygen measurements are available from the corresponding author upon reasonable request. Additional file 3: Supplementary Table S3. Group-level descriptive summary of intrauterine pO2 (Torr). Values represent mean of participant medians (range of medians across participants). Raw per-visit oxygen measurements are available from the corresponding author upon reasonable request.