Effect of reactive oxygen species in semen on the pregnancy of infertile couples

We intended retrospectively to investigate whether reactive oxygen species (ROS) levels, detected in whole semen, were correlated with the actual pregnancy rate. A total of 89 patients with data of ROS in semen, attending our male infertility clinics from April 1994 to June 2000, were evaluated. Semen parameters were determined with computer assisted semen analyzer (CASA) and ROS production levels were measured using a computer-driven LKB Wallac 1251 Luminometer after the addition of 40 mu L of 4 mM luminol at the patients' first visits. All of the participants were inquired about their partners' pregnancies after the mean follow-up of 24.0 months (range 1.4 to 74.1). They were divided into two groups (pregnant group: n = 41, non-pregnant group: n = 48) and their characteristics, semen profiles and integrated ROS levels were analyzed. There was no difference between the pregnancy rate of ROS detectable cases and negative cases. However, the mean integrated ROS level in detectable cases of the non-pregnant group was significantly higher than that in detectable cases of the pregnant group (115.61 +/- 74.32 mV/30 min/10(6) sperm versus 7.22 +/- 4.69 mV/30 min/10(6) sperm, P = 0.0033). Then, by calculating the receiver operating characteristics curve with 95% confidence intervals, 4.35 mV/30 min/10(6) sperm was considered as a cut-off value of ROS in semen for pregnancy. These results indicate that (i) highly detectable ROS in whole semen of infertile patients may have implications in their partners' pregnancies and that (ii) detection of ROS in whole semen has a prognostic value for idiopathic male infertility.