Predictive Factors for Live Birth in Fresh In Vitro Fertilization/Intracytoplasmic Sperm Injection Treatment in Poor Ovarian Reserve Patients Classified by the POSEIDON Criteria

The mechanisms underlying poor ovarian response (POR) in assisted reproductive technology remain unclear, there is no consensus on the management of poor responders, the POSEIDON stratification classifies infertility patients into “expected” or “unexpected” groups to provide a more nuanced picture of POR, but few researchers have discussed the independent predictive factors (smoothed plots and the threshold effect) for live birth in POR patients classified by the new criteria. We conducted a retrospective cohort study using clinical data from 6,580 POR patients classified by the POSEIDON criteria in the First Affiliated Hospital of Zhengzhou University, and explored the live birth based on the results before and after the threshold inflection point of each independent influencing factor. Among 6,580 poor ovarian reserve patients classified by the POSEIDON criteria, 1,549 (23.54%) had live births, and 5,031 (76.46%) did not have live births. Multivariate logistic regression analysis showed that female age (OR 0.901; 95% CI 0.887~0.916; P < 0.001), body mass index (OR 0.963; 95% CI 0.951~0.982; P < 0.001), antral follicle counting (OR 1.049; 95% CI 1.009~1.042; P < 0.001) and controlled ovarian hyperstimulation protocol were independent factors predicting live birth in patients with POR. The threshold effect analysis found that the inflection point of female age was 34 years old, and when age was > 34 years old, the probability of live birth in POR patients dropped sharply (OR 0.7; 95% CI 0.7~0.8; P < 0.001). The inflection point of BMI was 23.4 kg/m(2), and BMI had a negative correlation with live birth (OR 0.963; 95% CI 0.951~0.982; P < 0.001). The threshold inflection point of AFC was 8n. Female age, BMI, AFC and COH protocol were independent predictive factors associated with live birth in POR patients classified by the POSEIDON criteria. The smooth curve fit and threshold effect analyses provide clinical management strategies for these patients. In addition, the early-follicular-phase long-acting GnRH-agonist long protocol seems to have a higher live birth rates than other protocols. It is worth highlighting that BMI should be considered as well in the POSEIDON criteria.