Anxiety and depressive symptoms, and stress biomarkers in pregnant women after in vitro fertilization: a prospective cohort study

Abstract

Does in vitro fertilization (IVF) affect the course of anxiety and depressive symptoms as well as physiological stress from pregnancy to postpartum period? IVF mothers have more anxiety symptoms and higher stress biomarker levels but fewer depression symptoms than natural conception mothers at the third trimester of pregnancy, but these differences are negligible during postpartum period. Cross-sectional studies have found an association between IVF and high stress levels during the prenatal period. There is, however, no follow-up study about the IVF effect on the mental health status from pregnancy to postpartum, adopting simultaneous measurement of self-reported symptoms and stress biomarkers. This is a prospective cohort study. A total of 243 eligible women were recruited during the third trimester of pregnancy (60 women after successful IVF and 183 who conceived naturally). The recruitment was performed during a 12-month period, and the follow-up was carried out until 3 months after delivery. This study was performed in the Division of Obstetrics in a regional referral center. The State scale of the State-Trait Anxiety Inventory (STAI-S) and the Beck Depression Inventory-Sort Form (BDI/SF) were used as anxiety and depression indicators, respectively; salivary cortisol and α-amylase levels as stress biomarkers. Anxiety, depression and stress biomarkers were measured at the third trimester of pregnancy (T1), at 48 h after birth (T2) and at 3 months after birth (T3). Associations with IVF were assessed using ordinal mixed models for anxiety and depressive symptoms and linear quantile models for stress biomarkers. Relative to natural conception mothers, IVF mothers had higher STAI-S scores at T1 (P = 0.016, odds ratio (OR) = 2.46), and this difference remained steady from T1 to T2 (P = 0.37, OR = 0.70) and from T2 to T3 (P = 0.36, OR = 0.69). In the case of depressive symptoms, the IVF group obtained lower BDI/SF scores at T1 (P < 0.001, OR = 0.192). This difference was apparently reduced from T1 to T2 (P = 0.072, OR = 2.21) and remained constant from T2 to T3 (P = 0.107, OR = 2.09). It is important to note that whereas the mean BDI/SF score was not clinically significant for any group (it was lower than the cut-off 4), the mean STAI-S score of the IVF group at T1 was so (it was higher than the cut-off 19). As for stress biomarkers, IVF mothers had higher cortisol levels at T1 (P = 0.043, Δlog(cortisol) = 0.88) compared to natural conceptions. From T1 to T2 cortisol levels of both groups increased at the relatively same rate (P = 0.81, Δlog(cortisol) = -0.16). However, the progressions tended to be different from T2 to T3, with IVF mothers exhibiting a sharp decrease in cortisol levels (P = 0.059, Δlog(cortisol) = -0.94), while natural conceptions value remained steady. In the case of α-amylase, there were no statistically significant differences between both groups at T1 (P = 0.7, Δlog(α-amylase) = -0.095). On the contrary, while IVF mothers showed sustained α-amylase levels across the time, the progression was different in the natural conception group, who showed a decrease in α-amylase levels from T1 to T2 (P = 0.049, Δlog(α-amylase) = 0.596) and a non-significant increase from T2 to T3 (P = 0.53, Δlog(α-amylase) = -0.283). Since this follow-up study has been carried out from the third trimester of pregnancy, the findings cannot be generalized to extremely preterm births. IVF women may have lower depressive symptoms for being pregnant. However, due to the potential pregnancy complications associated with IVF, they may have higher physiological stress and clinically significant anxiety at the third trimester of pregnancy but not during postpartum. Taking into account that both prenatal high maternal cortisol levels and prenatal clinically significant anxiety increase the risk of disturbance in the fetal neurodevelopment, psychological therapy should be extended during pregnancy in IVF women. MV funded by FIS PI17/0131 grant from the Instituto de Salud Carlos III (ISCIII) and RETICS funded by the PN 2018-2011, and the European Regional Development Fund, reference RD16/0022/0001; AG-B funded by a 'Juan Rodés' Grant (JR17/00003) from the ISCIII. CC-P funded by a 'Miguel Servet I' Grant (CP16/00082) from the ISCIII. Authors declare no competing interests.