Maternal prenatal hair cortisol is associated with prenatal depressive symptom trajectories

Abstract

Maternal prenatal cortisol levels have been inconsistently associated with self-reports of prenatal psychological distress (PD). Previous research has linked hair cortisol concentration (HCC) evaluating cumulatively the previous months with cross-sectional PD measures that usually cover the past week(s), which may lead to misleading conclusions on their relations. We aimed to investigate how maternal HCC relates to cumulative PD measures across pregnancy. Maternal prenatal cortisol levels have been inconsistently associated with self-reports of prenatal psychological distress (PD). Previous research has linked hair cortisol concentration (HCC) evaluating cumulatively the previous months with cross-sectional PD measures that usually cover the past week(s), which may lead to misleading conclusions on their relations. We aimed to investigate how maternal HCC relates to cumulative PD measures across pregnancy.

Methods Subjects (N = 595) were drawn from the FinnBrain Birth Cohort Study. Maternal HCC was measured from hair samples collected at gestational week (gwk) 24 (HCC1, n = 467) and at delivery (HCC2, n = 222). As HCC1 and HCC2 comprised mostly of different subjects, they were considered as independent populations. Maternal PD assessments at gwks 14, 24, and 34 were the Edinburgh Postnatal Depression Scale (EPDS), the anxiety subscale of the Symptom Checklist (SCL-90), the Pregnancy-Related Anxiety Questionnaire -Revised2 (PRAQ-R2), and a daily hassles scale. Cumulative PD comprised of the mean scores of two consecutive assessments (mean1 = gwks 14 and 24; mean2 = gwks 24 and 34). In addition, EPDS and SCL scores were modelled by using growth mixture modelling to identify symptom trajectory categories. Regression models were adjusted for age, body mass index, education and use of selective serotonin/serotonin-norepinephrine reuptake inhibitor medication.

Results In the adjusted regression model, higher HCC2 was related to the “consistently elevated” prenatal depressive symptoms trajectory in comparison to “consistently low” (β =.71, p =.021) and “low and increasing” (β =.82, p = .011) symptom trajectories. Additionally, the cumulative mean (mean 1) of daily hassles in relationships was associated with HCC1 (β = 0.25, p = .004). General or pregnancy-related anxiety symptoms were unrelated to HCC after adjustment for the covariates.

Conclusions The assessment of cumulative or trajectory measures of PD can reveal important associations with maternal prenatal HCC, even though the associations are generally weak. Of the different dimensions of PD, prenatal trajectories of depressive symptoms were most consistently linked with end-pregnancy HCC levels.