A comprehensive evaluation of pre- and post-processing sperm parameters for predicting successful pregnancy rate following intrauterine insemination with the husband’s sperms

Abstract

Abstract Background To determine the predictive values of sperm parameters pre- and post-processing by density gradient centrifugation for clinical pregnancy rates (CPRs) following artificial insemination by husband (AIH) in infertile Chinese couples. Methods A total of 3,522 AIH cycles from 1,918 couples were retrospectively analyzed. The parameters were compared between the pregnant and non-pregnant groups and further between different etiological groups (Male-factor, Both-male-and-female-factor, and Other-factor). Multivariate logistic regression analysis was performed to create models for predicting the CPRs of each etiological group. Results The overall CPR was 13.3%. There were significant improvements for most sperm parameters after DGC. Multivariate logistic regression analysis indicated that, in overall AIH cases, the top parameters significantly influencing the CPR of AIH were pre-STR (OR = 1.037; P = 0.048) and post-VSL (OR = 1.036; P = 0.011). In the Male-factor Group, the top influencing parameters were pre-VCL (OR = 2.096; P = 0.008), pre-LIN (OR = 1.930; P = 0.002) and post-VSL (OR = 1.316; P = 0.023). In the Both-factor Group, the top influencing parameters were pre-VCL (OR = 1.451; P = 0.008) and post-motility (OR = 1.218; P = 0.049). In the Other-factor Group, the top influencing parameters were pre-VAP (OR = 1.715; P = 0.024), pre-STR (OR = 1.20; P = 0.011) and post-VSL (OR = 1.04; P = 0.017). Moreover, receiver operating characteristic analysis showed that the logistic regression models of the Male- and Both-factor Groups had greater powers for prognostic classification than those of other groups. Conclusions This study demonstrated that some sperm parameters have a collinearity relationship in predicting the CPR following AIH. Moreover, the predictive capacity of a multivariate logistic regression model is better than those of individual parameters, especially for the Male- and Both-factor Groups. In these cases, pre-VCL is the common top influencing factor.