P-443 An association between olfaction and unexplained recurrent pregnancy loss

Abstract

Abstract Study question Is there a male body-odor associated with unexplained recurrent pregnancy loss (uRPL)? Summary answer The body-odor of men from uRPL couples may have different chemical composition than control men. What is known already Recurrent pregnancy loss is defined as 2-3 consecutive pregnancy losses, affecting about 1% of all couples, and in about half of them, the reason for the pregnancy losses in unexplained (uRPL). In our previously published study, we found that women who experienced uRPL have altered perceptual and brain responses to men’s body-odors in comparison to control women. This suggests a possible link between uRPL and the olfactory system, resonating with the Bruce effect in mice. In the Bruce effect, the particularities of male body-odor are also a factor in the effect. Study design, size, duration We set out to ask whether the body-odor (BO) of men in uRPL is different from the BO of control men. To test this, we used two main approaches: have women smell BOs from uRPL-men and control-men and rate the perceived odor; and sample the BOs using a machine (electronic nose) and test if the composition of the BOs is objectively different. Participants/materials, setting, methods We collected BOs from spouses of uRPL and control women. We recruited women with uRPL, control women (with no history of pregnancy losses) and nulliparous women to sniff the BOs and rate them for perceived intensity, pleasantness, sexual attraction, and fertility. We then used an electronic nose (eNose, Airsense) to sample the BOs and used a Linear SVM classifier on this data, to try and classify the BOs to the two groups. Main results and the role of chance Twenty-one women with uRPL and 24 control women sniffed and rated 34 BO jars (17 uRPL-men). We found that uRPL-men BO was rated as more pleasant, sexually attractive and fertile than control men (rmANOVAs for each parameter but intensity revealed a significant effect of male group (all F(1,43)>14.24, all p < 0.001), no women-group effect and no interactions). We then recruited 51 nulliparous women, not related to the original study, to do the same task, and the results replicated: nulliparous women rated uRPL-men’s BO significantly more pleasant, sexually attractive and fertile than control men’s BO (all t(50)>2.88, all p < =0.006), while the intensity was not different (t(50)=0.87, p = 0.38). We then used a PEN3 eNose to sample 37 male BOs (18 uRPL). The PEN3 eNose contains 10 different metal oxide sensors, and each sensor is coated in a unique material that makes it sensitive to different sets of chemical compounds. Based on the data from the eNose, a Linear SVM classifier successfully classified the odors to uRPL-men or control-men at 69.9% accuracy in a five-fold cross-validation test (p < 0.001 estimated by repeating the process 1000 times and shuffling the labels). Limitations, reasons for caution This study shows an association between the body-odor of men and uRPL, however it shows a correlative connection (not causative). Therefore, the interpretation for these current results are limited. Wider implications of the findings These initial results suggest that uRPL-men’s and control-men’s BO have different chemical composition. This implies a possible contribution of men to the effect we observed in our previously published study, and combined, further strengthens a link between uRPL and the olfactory system that we continue to investigate. Trial registration number not applicable