Clinical factors associated with low serum progesterone levels on the day of frozen blastocyst transfer in hormonal replacement therapy cycles

STUDY QUESTION Which factors are associated with low serum progesterone (P) levels on the day of frozen embryo transfer (FET), in HRT cycles? SUMMARY ANSWER BMI, parity and non-European geographic origin are factors associated with low serum P levels on the day of FET in HRT cycles. WHAT IS KNOWN ALREADY The detrimental impact of low serum P concentrations on HRT-FET outcomes is commonly recognized. However, the factors accounting for P level disparities among patients receiving the same luteal phase support treatment remain to be elucidated, to help clinicians predicting which subgroups of patients would benefit from a tailored P supplementation. STUDY DESIGN, SIZE, DURATION Observational cohort study with 915 patients undergoing HRT-FET at a tertiary care university hospital, between January 2019 and March 2020. PARTICIPANTS/MATERIALS, SETTING, METHODS Patients undergoing single autologous blastocyst FET under HRT using exogenous estradiol and vaginal micronized progesterone for endometrial preparation. Women were only included once during the study period. The serum progesterone level was measured in the morning of the FET, in a single laboratory. Independent factors associated with low serum P levels (defined as <= 9.8 ng/ml, according to a previous published study) were analyzed using univariate and multivariate logistic regression models. MAIN RESULTS AND THE ROLE OF CHANCE Two hundred and twenty-six patients (24.7%) had a low serum P level, on the day of the FET. Patients with a serum P level <= 9.8 ng/ml had a lower live birth rate (26.1% vs 33.2%, P = 0.045) and a higher rate of early miscarriage (35.2% vs 21.5%, P = 0.008). Univariate analysis showed that BMI (P < 0.001), parity (P = 0.001), non-European geographic origin (P = 0.001), the duration of infertility (P = 0.018) and the use of oral estradiol for endometrial preparation (P = 0.009) were significantly associated with low serum P levels. Moreover, the proportion of active smokers was significantly lower in the 'low P concentrations' group (P = 0.002). After multivariate analysis, BMI (odds ratio (OR) 1.06 95% CI (1.02-1.11), P = 0.002), parity (OR 1.32 95% CI (1.04-1.66), P = 0.022), non-European geographic origin (OR 1.70 95% CI (1.21-2.39), P = 0.002) and active smoking (OR 0.43 95% CI (0.22-0.87), P = 0.018) remained independent factors associated with serum P levels <= 9.8 ng/ml. LIMITATIONS, REASONS FOR CAUTION The main limitation of this study is its observational design, leading to a risk of selection and confusion bias that cannot be ruled out, although a multivariable analysis was performed to minimize this. WIDER IMPLICATIONS OF THE FINDINGS Extrapolation of our results to other laboratories, or other routes and/or doses of administering progesterone also needs to be validated. There is urgent need for future research on clinical factors affecting P concentrations and the underlying pathophysiological mechanisms, to help clinicians in predicting which subgroups of patients would benefit from individualized luteal phase support. STUDY FUNDING/COMPETING INTEREST(S) No funding/no conflicts of interest.