Hormonal transitions such as across pregnancy and postpartum may trigger depressive episodes in some women. It is not known why, but estrogen sensitivity may play a critical role. A preclinical human risk model showed that depressive symptoms induced by pharmacological sex-hormone manipulation is linked to increases in serotonin transporter (SERT) brain binding, which lowers serotonergic brain tone. It is currently unknown if these findings translates to women across pre- to postpartum transitions.

This longitudinal project studies a group of women who will deliver by planned caesarian, thus permitting the collection of cerebrospinal fluid (csf) containing central markers of serotonergic signaling, at the latest point in pregnancy. The women are followed across late pregnancy, delivery and 6 months postpartum to illuminate relations between sex-hormones, stress-regulation, estradiol sensitivity, csf markers of neurotransmission, serotonin transporter genotype variance, and potential development of subclinical or manifest depressive symptoms. Further, markers of relevance for the infant brain development and stress-regulation will be obtained from placenta tissue and umbilical cord blood. A subgroup of 70 women will participate in a brain imaging program early postpartum (week 3-5), which includes an evaluation of brain activity and structure and in vivo molecular brain imaging serotonergic markers. Thus, serotonergic markers in csf can be combined with postpartum molecular brain imaging of key features of serotonin signaling. Women in the imaging program are selected based on variation in their level of mental distress immediately postpartum (day 2-5).

The study's main hypothesis is that women with high-expressing SERT genotypes are more sensitive to peripartum hormonal transition in terms of changes in serotonergic tone and emergence of depressive symptoms and that such an association will be stronger in the presence of candidate gene transcript biomarkers of oestrogen sensitivity. A further hypothesis is that in vivo molecular brain imaging and csf based serotonergic markers will be associated with depressive symptoms both early and later postpartum.

Ideally, this project will provide a rationale for future targeted prevention and/or treatment of perinatal depression in women at high risk, which holds grand potential to protect not only mother but also infant brain health long-term.

Motivation:

Major depressive disorder (MDD) affects twice as many women as men and women are at an increased risk during hormonal transition phases such as pregnancy and birth. A highly relevant subpopulation within the mixed MDD diagnostic category comprises women who develop perinatal depression (PND). PND is defined as a depressive episode with onset during pregnancy or up to 4 weeks postpartum, however epidemiological studies show that the risk of developing depression is heightened for 6 months postpartum. PND affects 10-15% of mothers postpartum. Why certain women are at high risk of developing perinatal depression (PND) remains unclear but recent studies suggest that these women might be particularly sensitive to the transition from high levels of placenta-produced sex-steroids in pregnancy to the hormone withdrawal phase postpartum. Further, pharmacologically induced changes in ovarian sex-hormones can produce depressive symptoms in a subgroup of otherwise healthy women and that the emergence of depressive symptoms is linked to both estrogen fluctuations and increases in serotonin transporter (SERT) brain binding (which putatively lowers serotonergic brain tone). Intriguingly, common gene variants that index SERT expression levels show "gene BY environment" associations with risk for depression, such that high-expressing SERT genotypes render women more vulnerable to depressive symptoms early - but not late - postpartum in a "gene-dose" dependent manner. Further, DNA methylation and gene expression markers of estradiol sensitivity predispose to PND and are linked to the estradiol stimulation phase in the pharmacological manipulation of sex-steroids risk model, thus constituting a candidate biomarker for PND.

It is currently unknown if estradiol sensitivity during pregnancy confers to PND risk through mechanism that (transiently) affect serotonergic tone in susceptible women. Changes in brain function late in pregnancy may extend to the early postpartum and shape how the brain integrates additional neurobiological changes that are associated with the postpartum hormonal withdrawal phase. This study will examine these mechanisms in a group of pregnant women that are followed from late pregnancy across early to late postpartum up to 6 month.

Natural variation in SERT-genotypes provides a unique opportunity to specifically address the interaction between SERT-gene expression-capacity and estradiol exposure through pregnancy in processes driving changes in serotonergic tone, brain structure and activity, and mental health from late pregnancy to 6 months postpartum. The time-points comprise: basic program: 2-5 days postpartum, 6 weeks postpartum and 6 months postpartum for all participants and for the imaging program participants: 2-5 days postpartum, 3-5 weeks postpartum, 12 weeks postpartum, 6 months postpartum.

By including women who undergo planned caesarean section, cerebrospinal fluid (CSF) can be obtained and thus, for the first time combine CSF markers of serotonergic tone and other transmitter systems (serotonin, 5-hydroxyindolacetic acid, other monoamines, γ-aminobutyric acid) with molecular brain imaging methods that index serotonergic tone (i.e. serotonin 4 receptor binding)

Aims:

• Determine if depressive symptoms from late pregnancy to 6 months postpartum map onto molecular brain imaging markers of serotonin signaling early postpartum (week 3-5), and evaluate if such markers and/or symptoms are dependent on serotonin transporter genotype and/or predicted by candidate gene transcription biomarkers for estrogen sensitivity.
• Evaluate how markers of stress-regulation capacity, brain activity, brain structure (hippocampal volume) and central markers of neurotransmission are associated with the emergence of depressive symptoms in women postpartum.
• Map the association between serotonin-4 receptor binding and cerebrospinal fluid markers of serotonergic tone (serotonin and 5-hydroxyindolacetic acid levels).
• Determine if markers of mental distress in women during pregnancy and the postpartum period are associated with infant markers of stress-regulation and serotonergic signaling in placenta and umbilical cord blood.

Hypotheses:

• Women with high-expressing SERT genotypes are more sensitive to estradiol exposure in late pregnancy in terms of changes in proxies for serotonergic tone (PET imaging or csf based) and emergence of depressive symptoms in late pregnancy and/or postpartum and such an association will be stronger in the presence of candidate gene transcript PND biomarkers.
• CSF levels of 5-hydroxyindolacetic acid are associated with serotonin 4 receptor brain PET (Positron Emission Tomography) binding.

Study design:

150 pregnant women between 18-40 years of age who deliver by planned caesarean section, due to breech presentation of the fetus or previous caesarean section, will be included in a longitudinal study. Participants will be recruited at the midwife clinic of Rigshospitalet, Copenhagen, Denmark. Based on natural variation in European populations the expected distribution of high vs. low expressing SERT genotypes is 40/60, respectively, thus genotype status can be included in the analysis structure. Self-reported psychometrics and questionnaires will be collected online at inclusion, across the pre- to postpartum transition and up to 6 months postpartum (basic program: 2-5 days postpartum, 6 weeks postpartum and 6 months postpartum; imaging program: 2-5 days postpartum, 3-5 weeks postpartum, 12 weeks postpartum, 6 months postpartum). CSF will be collected as part of the anesthetic procedures for a planned caesarean section, thus avoiding any additional invasive procedures. CSF markers of serotonergic tone (serotonin and its main metabolite, 5-HIAA) will be measured by HPLC techniques. Corresponding blood samples for determining relevant biomarkers (sex-steroids, DNA, mRNA and microRNA) and saliva for hypothalamic-pituitary-adrenal axis dynamics, will be taken just before the planned caesarean section. Hair from mother and infant will be collected around delivery for further cortisol analyses. Placenta tissue and umbilical cord blood will also be collected for determining relevant markers of serotonergic and hypothalamic-pituitary-adrenal axis functioning.

A subgroup of the study cohort selected towards high (N=35) or low risk for later manifest PND (N=35), based on symptoms of mental distress 2-5 days postpartum (in-house interview, high-risk scores correspond to at least 12 on the Kennerley Maternity Blues Questionnaire and at least 8 on Stein's Maternity Blues Scale), will participate in an extended brain imaging program. This program will include 5-HT4R ([11C]SB207145) PET, structural MRI, functional MRI (including emotional processing, reward processing and resting state fMRI), neuropsychological testing and face to face rating of mental state with a semi-structured interview (HAM-D17).

The study includes long-term follow-up at six months. Collected data will enter the Center for Integrated Molecular Brain Imaging database, thus providing a basis for longitudinal follow-up, data sharing and crossvalidation.

Statistics:

Power calculations based on inter-subject variability of the 5-HT4R show that an imaging group size of 35 is required to detect a 15% difference with a power of 0.8 for the brain regions of interest. With the full cohort number of 150 and due to oversampling of high and low risk women, about 25 women are expected to develop manifest PND episodes and more will display subclinical depressive symptoms, which will allow for correlation analyses with relevant outcome parameters including the candidate gene transcript based biomarker of estrogen sensitivity.

Highly correlated self-reported psychometric outcomes will be included in a latent variable construct of self-reported mental state (composite measure) using structural equation modelling.

Ethics:

The PET scans convey no known risk for adults. Infants will not be exposed to radiation and will be nursed by special staff or a close relative while the mother is scanned. Participants who develop levels of mental distress or depressive symptoms that approach clinical thresholds will be referred to relevant and timely psychiatric care. The study has been approved by the local ethics committee.

Inclusion Criteria:

• Age 18-40 years
• Healthy pregnant women planned to deliver by caesarean section due to breech position of the fetus or previous caesarean section.

Exclusion Criteria:

• Current or previous severe psychiatric disorder such as psychotic disorders, eating disorder and bipolar disorder or current or previous psychiatric disorder requiring hospitalization.
• Current or previous neurological diseases, severe somatic disease, severe postpartum hemorrhage or use of medication that can interfere with study outcomes
• Severe disease or malformations in infants
• Obesity or underweight (pre-gestational BMI below 18 or above 35)
• Not fluent in Danish or severe visual or hearing impairments
• Earlier or present learning disabilities
• MRI contraindications (claustrophobia, metal implants)
• Previous exposure to radioactivity > 10 millisievert (mSv) within the last year
• Alcohol or drug abuse

Participants will be recruited from the midwife clinic at Rigshospitalet, Copenhagen, i.e., included women will be residents of the central Copenhagen area.

Via database of Center for Integrated Molecular Brain Imaging (Knudsen et al 2016, NeuroImage) data will be available for neuroscience research community contingent on approval by scientific board.

2020-2035

Approval by scientific board