Trans-generational stress regulation: Mother-infant cortisol and maternal mental health across the perinatal period

Highlights

• Antenatal depression and cortisol may be important for developing infant stress regulation.

• Hair cortisol concentrations are useful for measuring cortisol in pregnancy and the postpartum.

• Maternal hair cortisol concentration across pregnancy was not associated with maternal depression.

• Antenatal maternal depressive symptoms were both associated with lower infant cortisol at 12 months of age.

• Lower infant cortisol reactivity was associated with higher infant externalizing symptoms.

Abstract

Understanding maternal mental health and cortisol regulation across pregnancy and the relationship to the development of the offspring’s stress regulation is critical to a range of health outcomes. The aim of this study was to investigate infant and maternal cortisol in women with depression. Data were obtained from 241 pregnant women within the Mercy Pregnancy and Emotional Wellbeing Study (MPEWS), a selected pregnancy cohort study. Depression was measured using the Structured Clinical Interview for DSM-IV (SCID-IV) and repeat Edinburgh Postnatal Depression Scale (EPDS). Repeated measures of antidepressant use, stressful events, anxiety symptoms and maternal hair cortisol concentrations (HCC) and infant cortisol at 12 months postpartum in saliva and hair. Socio-emotional outcomes were measured at 12 months by maternal report on the Brief Infant and Toddler Socio-emotional Assessment (BITSEA). This study found that maternal depression was not associated with maternal HCC. Anxiety, stress and antidepressant use were not associated with maternal HCC. Independently, higher maternal 3rd trimester maternal depressive and anxiety symptoms were associated with lower infant cortisol response at 12 months of age. A higher number of postpartum stressful events was associated with lower infant cortisol response. Lower infant stress reactivity was associated with higher externalizing symptoms at 12 months of age. Future studies are required to understand implications for later mental health.

Introduction

Research continues to show that maternal perinatal depression predicts less than optimal child outcomes, including a range of child mental health problems (Gentile, 2017; Lewis et al., 2014). However, there is less certainty as to the potential mechanism for this transmission (Lewis et al., 2014). The relationship between the experience of stress and the vulnerability to depression is well established and theoretically, it is suggested that the earlier the exposure to stress the stronger the relationship (De Kloet et al., 2005; Lupien et al., 2009). Clearly, such findings point to a possible role of the hypothalamic pituitary adrenal (HPA) axis where maternal mental health and cortisol regulation may play an early role in fetal neurodevelopment and subsequent vulnerability for child socio-emotional and developmental outcomes (Gunnar and Quevedo, 2007). Understanding the interaction between maternal depression and cortisol in pregnancy and how this might influence the development of infant stress regulation is the first step in then understanding a potential fetal programming pathway across early life that may influence vulnerability to later psychopathology (Glover, 2015; Glover et al., 2010).

A recent review of cortisol and perinatal depression identified 47 relevant studies but one of the key findings was the considerable variation in measurement of cortisol and the frequent reliance on screening self-report measures for depressive symptoms (Seth et al., 2016). Only four studies identified in this review were rated as high quality through meeting minimum criteria for measurement of cortisol. This review found that hypercortisolemia was associated with an increase in short-term depressive symptoms in the early postpartum and hypocortisolemia was associated with chronic depressive symptoms. A major limitation identified within this review was a paucity of studies that utilized diagnostic measures for mental health and use of community samples with low rates of clinically significant symptoms.

Notably, none of the studies identified in the Seth et al review utilized hair cortisol measurement. There are a number of significant advantages of hair for examination of cortisol. Hair can be more consistently collected than saliva or urine, since it will not be as affected by short-term environmental influences and provides average cortisol levels across time (Short et al., 2016). A systematic review has examined the emerging research examining hair cortisol and prenatal distress (Mustonen et al., 2018). This included perceived stress, symptoms of anxiety or depression or pregnancy related stress. Of the six studies identified in the review, none used diagnostic measures of mental disorders, all drew samples from a community with low levels of symptoms of mental illness and only one study used an additional measure of cortisol in conjunction with hair cortisol (Kramer et al., 2009; Mustonen et al., 2018). Of note, the two studies that found an association between hair cortisol and distress utilized hair samples collected in pregnancy. Furthermore, separately, both maternal emotional health in pregnancy and pregnancy cortisol have been associated with later child cortisol functioning (O’Connor et al., 2013; Pearson et al., 2015).

A review of research that has examined the link between adverse early experiences and child psychopathology has highlighted the importance of understanding cortisol regulation in infants as a potential pathway from early exposures to later vulnerability to psychopathology (Koss and Gunnar, 2018). While findings have been mixed, there have across several studies been trends to suggest increased levels may be associated with vulnerability to internalizing disorders and a clearer association with reduced cortisol associated with externalizing disorders (Kao et al., 2018; Koss and Gunnar, 2018; Ruttle et al., 2011). The HPA system and cortisol response change substantially over the first year of life with some stability and lower responsiveness to everyday stressors only being achieved around 12 months of age, supporting measurement of infant cortisol at 12 months or older (Gunnar and Quevedo, 2007; Lupien et al., 2009).

While it is important to understand the potential influence of maternal psychopathology on maternal and infant cortisol over the perinatal period, it is equally important to understand the impact of treatments. Yet none of these recent reviews of perinatal distress and depression and cortisol have reported on perinatal antidepressant use and cortisol despite the widespread use of antidepressants to treat depression and anxiety in pregnancy (Jimenez-Solem et al., 2013). Antidepressant treatment has been found to alter HPA functioning and impact cortisol levels (Kieviet et al., 2016; Manthey et al., 2011).

In our study, we firstly aim to examine the relationship between maternal hair cortisol concentrations (HCC) and depression across pregnancy and 12 months postpartum. In addition, we will examine if antenatal HCC is associated with antidepressant use in pregnancy. Secondly, we aim to test our hypothesis that maternal HCC across the perinatal period will predict infant HCC and salivary cortisol response and reactivity. Thirdly, we test the hypothesis that maternal mental health in pregnancy will be indirectly associated with infant cortisol (both HCC and cortisol reactivity at 12 months postpartum) through maternal cortisol. Finally, we examine if infant cortisol measures are associated with externalizing or internalizing symptoms at 12 months of age.