In the present study, we investigated salivary OT changes following the breastfeeding cycle at PD2 and 4 in primiparous mothers who delivered vaginally and who had no abnormal symptoms during the postpartum period. We also utilized both extracted and unextracted methods for OT measurements and examined the correlation between the methods. The results showed that the extraction method showed a significant increase in OT concentration with breastfeeding at PD2, but not at PD4. The unextracted method showed a significant increase at both PD2 and 4. However, when we examined the correlation between the OT concentrations of the extracted and unextracted methods, we found a correlation between the baseline values and the inverse direction, but no correlations were found at the other time points or in the total data. However, the difference between breastfeeding and baseline was significantly positively correlated with each other. Hence, we subsequently dealt with OT data at PD2 measured via the extraction method to investigate whether increased OT associated with breastfeeding may decrease maternal anxiety. The results showed that the higher the OT change indices (ΔOT, AUCi, and AUCg), the lower the state anxiety at PD2 and 4; however, it was not associated with trait anxiety. There was also an association in that those who had higher OT change indices at PD2 were linked to the exclusive breastfeeding rate at the 1-month postpartum follow-up. Therefore, our results suggest that OT release by breastfeeding in the immediate postpartum period may reduce maternal anxiety and increase the rate of subsequent exclusive breastfeeding, as recommended by WHO/UNICEF .
The importance of the extraction method for OT measurement has been raised several times in the past [25–27]. Unlike our previous study  and other reports that used unextracted methods, the OT measurements in the present study were preprocessed using the extraction method. Moreover, we also measured the data using the same unextracted method as in the previous study and performed a novel comparison of the two conditions (the present unextracted data and population comprised a different dataset as that of our previous manuscript ). Thus, it can be observed that the true OT concentration, which is not a false positive, is clearly different from the absolute value of the unextracted method. The correlation of the baseline data alone was significant, but the direction of the correlation did not match. A number of publications [18–20, 28] have examined the relationship between baseline unextracted OT concentrations and various social and psychological metrics, although we believe that these need to be re-validated using extraction methods. In contrast, for ΔOT, there was a significant positive correlation between the extracted and unextracted methods. Even though false positives are included, the amount of change may still capture the overall change associated with the true OT change, even with the unextracted method. The following discussion is based on the results obtained from the extraction method.
Our results in the present study showed that breastfeeding increased salivary OT at PD2 but did not show a significant increase at PD4 or could not detect the changes by the saliva extraction method. Plasma OT, however, has been shown to capture an increase in OT even 2 months postpartum . Therefore, this may be a phenomenon specific to salivary OT measurements. In particular, changes in salivary OT associated with breastfeeding appear to be influenced by the postpartum period. White-traut, et al.  measured salivary OT associated with the breastfeeding cycle within 8 months postpartum and found that baseline levels were the highest and decreased the most during breastfeeding. Carter, et al. (2007)  reported similar results. In addition, Jong, et al. (2015)  failed to observe changes in OT. Our PD4 results did not clearly decline, similar to those found in the studies by Carter and White-Traut; however, the results seemed to reflect a slight increase. We also examined the influences of other demographic and clinical factors that may inhibit OT secretion at PD4. There was an association between longer labor time and the use of induction medication, i.e., prolonged labor and a suppressed increase in OT at PD4. In addition, when examining the relationship between the use of induction medicines and the number of breastfeeding sessions 1–3 days postpartum, the use of induction medicines was associated with fewer breastfeeding sessions. Although positive feedback from OT is essential for parturition to instigate uterine contractions, it was thought that these participants may have been vulnerable to the neurobiological mechanism of OT release. For example, DNA methylation is one of the molecular mechanisms that suppresses gene expression . Hiraoka et al.  reported that the more methylation in the OXT gene promoter region, the higher personal distress, an aspect of affective empathy of the mothers, and that there was brain gray matter volume reduction in the inferior temporal gyrus, which regulates empathy. Epigenetic mechanisms of the OXT gene may regulate OT secretion, leading to suppression of OT secretion during breastfeeding.
In the present study, for the first time, the changes in salivary OT with breastfeeding were captured using the extraction method. We showed that this OT increase was significantly associated with lower state anxiety, but it was not associated with trait anxiety. Unfortunately, we were unable to measure state anxiety as a time course change before and after breastfeeding. Although we only have the results of each one-time assessment at PD2 and 4, we were able to show an association between the increase in OT with breastfeeding and the decrease in state anxiety. However, we cannot deny the possibility that this relationship reflects an increase in OT because the original maternal anxiety was lower. Because there was no association between trait anxiety and OT and because the STAI was taken after the breastfeeding session, the lower maternal anxiety could be highly affected by the change in OT increase associated with breastfeeding. On the other hand, a higher OT increase with breastfeeding at PD2 was also causally associated with exclusive breastfeeding at the 1-month postpartum follow-up. WHO (1998)  encourages mothers to breastfeed for at least 6 months postpartum, recognizing physical and mental benefits to both the mother and infant. The present study results suggest that repeated self-exposure to endogenous OT triggered by breastfeeding in the immediate postpartum period in mothers could reduce their anxiety and lead to successful breastfeeding and ultimately to the health of both mother and child.
The present study had three major limitations. First, this study did not assess maternal anxiety using objective measures, such as behavioral tasks or brain MRI. In general, several studies have shown that OT has anxiolytic effects [32, 33], but whether self-exposure to endogenous OT caused by breastfeeding has anxiolytic effects has not been directly investigated. Although brain MRI would be difficult in the immediate postpartum period, it would have been beneficial if the anxiolytic effect could have been proven through measures, such as behavioral tasks. Second, this study was limited to primiparous women who delivered vaginally. A comprehensive study of the effects of wound pain and other factors in post-cesarean and multiparous women will be a future challenge. Third, the follow-up on breastfeeding rates was limited to 1 month. Although we showed that high OT secretion with immediate postpartum breastfeeding was associated with prolonged exclusive breastfeeding 1-month postpartum, it is desirable to clarify the longer-term effects using a prospective cohort design.