ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1503

Perinatal Complications:Associations with Postpartumdepressive symptoms andNeuroticism

PATRICIA ECKERDAL

ISSN 1651-6206ISBN 978-91-513-0466-3urn:nbn:se:uu:diva-362566

Dissertation presented at Uppsala University to be publicly examined in Sal IX,Universitetshuset, Biskopsgatan 3, Uppsala, Friday, 30 November 2018 at 19:15 for thedegree of Doctor of Philosophy (Faculty of Medicine). The examination will be conductedin Swedish. Faculty examiner: senior professor Christina Hultman (Department of MedicalEpidemiology and Biostatistics, Karolinska Institutet).

AbstractEckerdal, P. 2018. Perinatal Complications: Associations with Postpartum depressivesymptoms and Neuroticism. Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1503. 87 pp. Uppsala: Acta Universitatis Upsaliensis.ISBN 978-91-513-0466-3.

Even though most pregnancies and deliveries are uncomplicated, still fifteen percent of allwomen in developed countries suffer pregnancy-related complications. The aim of this thesiswas to explore the associations between perinatal complications and perinatal maternal health,with emphasis on postpartum depressive symptoms (PPDS) and neuroticism taking into accountpotential confounding or mediating factors such as history of depression, antenatal depressivesymptoms and delivery experience.

In the first study (n=446), the association between heavy postpartum haemorrhage and PPDSat six weeks postpartum was delineated by using path-analysis in order to provide insight intothe complex mediating roles of several consequences of postpartum haemorrhage. There was nodirect association between postpartum haemorrhage and PPDS, only an indirect one via anaemiaat discharge and negative delivery experience.

The second study (n=3888) examined the association of mode of delivery with PPDS at 6weeks postpartum. The results indicate that the association between elective caesarean sectionand PPDS is highly confounded by history of depression and fear of delivery, while emergencycaesarean section and vacuum extraction increase odds for PPDS by leading to postpartumcomplications and negative delivery experience.

The third study (n=1503) investigated the association between the use of epidural analgesiaduring delivery and PPDS. A positive association in the crude analysis was no longer presentafter adjustment for sociodemographic, psychosocial and obstetrical variables, indicating thatpain relief through epidural analgesia is not likely to affect risk for PPDS.

In the last study (n=1969), the association between neuroticism and perinatal complicationswas explored. Neuroticism was not associated with adverse perinatal outcomes, except forgestational diabetes mellitus. The association, however, became statistically non-significantafter adjusting for psychiatric morbidity.

In summary, the current studies do no find evidence for a direct association between perinatalcomplications and postpartum depressive symptoms or neuroticism. However, several importantmediators have been identified, among which postpartum anaemia and negative deliveryexperience deserve special attention. Also, earlier psychiatric history needs to be addressed asan important confounder.

Keywords: antenatal depression, ceasarean section, delivery complications, Edinburghpostnatal depression scale, EPDS, epidural analgesia, gestational diabetes mellitus,instrumental delivery, neonatal complications, neuroticism, obstetric complications,personality, perinatal complications, postpartum depression, postpartum haemorrhage,pregnancy complications, vacuum extraction, vaginal delivery

Patricia Eckerdal, Department of Women's and Children's Health, Akademiska sjukhuset,Uppsala University, SE-75185 Uppsala, Sweden.

© Patricia Eckerdal 2018

ISSN 1651-6206ISBN 978-91-513-0466-3urn:nbn:se:uu:diva-362566 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-362566)

List of Papers

This dissertation is based on the following papers, which are referred to in the text by their Roman numerals.

I Eckerdal, P., Kollia, N., Lofblad, J., Hellgren, C., Karlsson, L., Hogberg, U., Wikstrom, A-K., Skalkidou, A. (2016) Delineating the association between heavy postpartum haemorrhage and postpartum depression. PLoS One, 2016; 11(1): e0144274

II Eckerdal, P., Georgakis, M., Kollia, N., Wikstrom A-K., Hog-berg, U., Skalkidou, A. (2018) Delineating the association be-tween mode of delivery and postpartum depression symptoms: a longitudinal study. Acta Obstet Gynecol Scand.; 2018; Mar;97(3):301-311. doi: 10.1111/aogs.13275.

III Eckerdal, P., Kollia, N., Karlsson, L., Skoog-Svanberg, A., Wik-

ström,A-K., Högberg, U., Skalkidou, A. (2018) The association of epidural analgesia during delivery and postpartum depressive symptoms. A Swedish cohort study. Submitted

IV Axfors C., Eckerdal, P., Volgsten H., Skoog-Svanberg A. Wik-

ström, A. K., Ekselius, L. Ramklint, M., Sundström-Poromaa, I., Skalkidou, A. (2018) Neuroticism is not associated with the risk of obstetric or neonatal outcomes. Submitted

Reprints were made with permission from the respective publishers.

Contents

Introduction ............................................................................................... 11 Normal delivery ....................................................................................11Perinatal complications .........................................................................12Antenatal complications ........................................................................13

Gestational Diabetes .........................................................................13Fear of delivery ................................................................................14

Delivery complications .........................................................................15Postpartum Haemorrhage ..................................................................15Mode of delivery ..............................................................................17Pain experience.................................................................................18Epidural Analgesia ...........................................................................18

Delivery experience ..............................................................................19Postpartum depression...........................................................................20

Definition .........................................................................................20Prevalence ........................................................................................21Comorbidity .....................................................................................22Aetiology and risk factors .................................................................22Screening..........................................................................................22Prevention ........................................................................................23Treatment .........................................................................................23Consequences ...................................................................................24

Personality ............................................................................................25Neuroticism ......................................................................................25

Rationale for the Thesis .............................................................................27

Aims .........................................................................................................28General Aim .........................................................................................28Specific Aims .......................................................................................28

Material and Methods ................................................................................30Study populations..................................................................................30

The BASIC project (Studies I-IV) .....................................................30The UPPSAT project (Study I) .........................................................30Studies including personality assessment (Study IV) .........................31

Study design .........................................................................................32Study I ..............................................................................................32Study II ............................................................................................32Study III ...........................................................................................33Study IV ...........................................................................................33

Variables...............................................................................................34Exposure ..........................................................................................34Outcome ...........................................................................................35Covariates ........................................................................................35

Psychometric instruments ......................................................................37EPDS ...............................................................................................37SSP ..................................................................................................37

Statistical analyses ................................................................................38Univariate analyses ...........................................................................38Logistic regression ............................................................................38Path Analysis ....................................................................................39

Ethical considerations ................................................................................41

Results ......................................................................................................42Study I ..................................................................................................42Study II .................................................................................................43Study III................................................................................................46Study IV ...............................................................................................47

Discussion .................................................................................................49Main Findings .......................................................................................49

Study I ..............................................................................................50Study II ............................................................................................50Study III ...........................................................................................51Study IV ...........................................................................................51

Methodological considerations ..............................................................51Strengths ..........................................................................................51Limitations .......................................................................................52

Clinical implications of the results .........................................................54

Summary ...................................................................................................56

Conclusion ................................................................................................57

Future studies ............................................................................................58

Sammanfattning på svenska .......................................................................60

Acknowledgments .....................................................................................62

References .................................................................................................65

Appendix...................................................................................................84Participants in BASIC ...........................................................................84Participants in UPPSAT ........................................................................85Conceptual Directed Acyclic Graph (Study III) .....................................86Directed Acyclic Graph including available variables (Study III) ...........87

Abbreviations

BMI Body mass index

CI Confidence Interval

CS Caesarean section

EDA Epidural analgesia

EPDS Edinburgh postnatal depression scale

GDM Gestational diabetes mellitus

GSEM Generalized structural equation modelling

Hb Haemoglobin concentration

HPA-axis Hypothalamic-pituitary-adrenal axis

LGA Large for gestational age

MBR Medical birth register

NEO-PI Neuroticism Extraversion Openness Personality Inventory

OR Odds ratio

PPD Postpartum depression

PPDS Postpartum depressive symptoms

PPH Postpartum haemorrhage

PTSD Posttraumatic stress disorder

SEM Structural equation modelling

SGA Small for gestational age

SSP Swedish university Scales of Personality

SSRIs Selective serotonin reuptake inhibitor

VE Vacuum extraction

W-DEQ Wijma Delivery Expectancy/Experience Questionnaire

9

Preface

To become a parent is an overwhelming experience. The perinatal period is a time of intense change and transition, in both somatic and psychological mo-dalities. The underlying hormonal and metabolic changes taking place in com-bination with psychosocial factors could increase the risk of depression after delivery. This may severely affect the mother, but also her partner and their child in both the short- and long-term perspective.

The multidisciplinary research field of perinatal depression has expanded during the last thirty years. Yet, there are still many unsolved questions on how to identify, treat, and even more desirably, prevent postpartum depres-sion.

Mainly thanks to social media and user’s organizations, the last years have allowed for more open discussions about mental illness and depression in gen-eral. Yet, unfortunately, getting depressed while having a newborn is still as-sociated with stigma.

As an obstetrician, I commonly encounter women who have just delivered a baby, sometimes under traumatic conditions. My clinical mission is to guide and help the individual woman who suffers from, or is at risk for, perinatal depression or other mental disorders. With my thesis, my intention has been to contribute to the collection of new knowledge about the association be-tween perinatal complications and maternal mental health and move the field a step closer towards a more integrative approach considering both somatic and psychological aspects in perinatal health.

10

11

Introduction

Normal delivery WHO defines normal delivery as follows: "spontaneous in onset, low-risk at the start of labor and remaining so throughout labor and delivery. The infant is born spontaneously in the vertex position between 37 and 42 completed weeks of pregnancy. After birth, mother and infant are in good condition" (1). The definition also stresses that “giving birth is not only safe but also a posi-tive experience for women and their families” (2).

When the woman presents to the delivery ward and the first stage of deliv-ery is affirmed, a risk assessment of the forthcoming delivery is performed. This assessment includes obstetric history, further medical history, blood pres-sure, and monitoring of the fetal heart rate by cardiotocography for at least 20 minutes (3).

There is no universal definition of the time point corresponding to the onset of delivery. The Swedish guidelines define the onset of the active first stage of delivery when two of the following three criteria are fulfilled: 1) cervix is dilated 4 cm or effaced and dilated >1 cm; 2) regular painful contractions with a frequency of 2-3/ 10 minutes; 3) rupture of membranes (4). Similarly to the definition of the onset of delivery, there is no well-established recommenda-tion for the duration of the active first stage. WHO recommends the duration from the time of cervix dilation of 5 cm to delivery not to extend 12 hours in primiparas and 10 hours in multiparas (2). If augmentation is needed, the first intervention is amniotomy if the membranes are unruptured, otherwise oxyto-cin-infusion.

The second stage starts when the cervix is fully dilated and continues until the child is born. A duration up to three hours is considered normal for nullip-arous women, as opposed to two hours for multiparous women (2).

The third stage of delivery starts after the child is born, and ends when the placenta is delivered, normally within 30 minutes (2).

Throughout the entire delivery process, foetal monitoring should regularly be performed by auscultation or cardiotocography (4). In almost all deliveries, the woman requires some form of pain-relief. As a first approach, non-phar-macological methods such as bath, massage acupuncture and transcutaneous nerve stimulation are preferred (5, 6). Later on, there is often a need for phar-macological interventions, like nitrous acid, paracervical blockade, local an-aesthetic blockade or epidural analgesia (EDA) (2, 5).

12

Perinatal complications In 2016, Lancet published a series of papers about maternal health, which af-firm that improvement of global maternal health has a central role in achieving the United Nations Sustainable Development Goals (7, 8). The authors also stress the diversity and divergence of maternal health (9-14). Fortunately, the global maternal mortality has significantly decreased during the last two dec-ades, but remains worryingly high in some low- and middle-income countries (7). Maternal morbidity is defined by WHO as “any health condition attributed to and/or aggravated by pregnancy and childbirth that has a negative impact on the woman's wellbeing”. This definition implies that not only physical con-ditions but also conditions related to mental health and sense of wellbeing may be considered as maternal morbidity.

The causes of maternal mortality and morbidity are increasingly diverse, reflecting large-scale demographic, epidemiological, socioeconomic, and en-vironmental transitions (10, 15). Poor maternal health often mirrors re-strictions in human rights (10). The WHO recently initiated the Morbidity Working Group (MMWG) that identified 121 diagnostic categories of mater-nal morbidity (15, 16). According to their report, the five main direct obstetric causes of maternal morbidity include postpartum haemorrhage (PPH) (preva-lence 6.2–10.8%), eclampsia (0.5%), pre-eclampsia (2.3%), severe abortion complications (0.6%), and puerperal sepsis (4.4%) (17). These prevalence rates are just rough estimations as the lack of common definitions and assess-ment methods may bias the estimates. For example, some studies included conditions existing before pregnancy in their definitions of maternal morbidity, whereas others excluded them (18, 19). Some studies even classified conditions like nausea, commonly experienced during pregnancy, as a type of morbidity, whereas others defined maternal morbidity strictly in terms of pregnancy-associated hospitalizations (18). There are also epidemiological limitations of studies on maternal morbidity. The most common limitation is facility-based instead of population-based studies, leading to selection bias (17). Hospital-based studies on maternal morbidity might only be representative of women who seek care (10). This is even more apparent in low- and middle-income countries, where organized registers for diagnosis are lacking (10) and the participation in perinatal care is lower, although fortunately increasing (20, 21). Commonly, studies rely on self-reported data, which might lead to an over-estimation of the complications, suggesting that self-perceived ill health is not only the result of biological changes but also depends on social support and social context (17). There is also a lack of globally representative systematic reviews on many conditions. Several of the published systematic reviews are based solely on studies from specific countries, which cast some doubt on the geographic representativeness of the findings (17). A further limitation for the comparability of the studies is the use of different methods for quantifying

13

morbidity rate. Overall prevalence is the most commonly used metric, but other studies may use point-prevalence or period-prevalence, and the assessment time-point may differ, thus leading to divergent results (17). The forthcoming paragraphs present a brief summary of the perinatal complications examined in this thesis.

Antenatal complications Gestational Diabetes Gestational diabetes mellitus (GDM) is one of the most common antenatal complications with potential adverse effects for both the woman and the foe-tus/newborn if not optimally treated (22). The prevalence is increasing, in line with the worldwide epidemic of obesity, and is now estimated to 16.2% (23). However, the prevalence broadly varies worldwide depending on diagnostic criteria (24, 25) and ethnicity (26), ranging from 1.8% in Sweden (27) up to 74% in Barbados (28). Even in Sweden, the prevalence differs across regions according to screening procedures, ethnic group composition and changed di-agnostic criteria during the last years (27, 29-31).

Numerous studies support an association between diabetes and depression in general (32), with a bidirectional causality (33-35). There is more scarce literature on associations between GDM and perinatal depression. Most stud-ies report an association between GDM and antenatal depression (36) as well as PPD (36-40), but there are also studies reporting no association (41). Stud-ies on the association between pre-existing diabetes and perinatal depression report contradicting results (39, 41).

There are several possible explanations for the reported associations of maternal mental health with GDM. These might include changes in the hypothalamic–pituitary–adrenal axis (HPA-axis), where heightened stress levels and chronic inflammation induce insulin resistance (42) via pregnancy-related intrinsic inflammatory mechanisms (43). Depression is further associ-ated with lower physical activity and a change in dietary patterns, which both are risk factors for GDM (44, 45). The distress of getting the diagnosis of GDM may also lead to reactive depression, for example when more advanced treatment is needed or complications have occurred (46). Additionally, some studies report an overrepresentation of obesity and metabolic syndrome in in-dividuals with high neuroticism (47, 48), which has been linked to perinatal depression. Some studies also report an association between obesity and PPD (49).

14

Fear of delivery A woman´s expectations and experiences of pregnancy and delivery are often complex with both positive and negative thoughts, and most women experi-ence some concerns about the forthcoming delivery. However, for many women the fear might affect the women´s quality of life and could also take phobic dimensions as to restrict their daily life and relationships (50, 51). The prevalence of fear of delivery varies between 4 and 40% depending on the definition of fear of delivery, study design, questionnaires and social context (52). A recent meta-analysis reports a prevalence of 14% for tokophobia, i.e. severe fear of delivery (51).

In Sweden, nearly all women participate in a system of regular visits at the maternal primary care unit (53). Except for health improvement and screening for pregnancy-related disorders, these visits include qualitative assignments like screening for fear of delivery, and parental education in order to improve strategies to deal with the forthcoming delivery and parenthood (53). In case of positive screening for fear of delivery, the woman is offered appointments to a midwife or consultant physician for helping strategies to handle the fear. However, there are no standardized guidelines for either definition, screening method or treatment of fear of delivery in Sweden (54). Eight percent of Swe-dish pregnant women obtain support and/or treatment for fear of delivery (27).

Fear of delivery is more common in nulliparous women, so called primary fear of delivery (51, 55). Secondary fear of delivery is associated with trau-matic experiences from previous delivery, these could either be based on ob-jective delivery complications or on psychological responses (50). Other fac-tors associated with fear of delivery include sociodemographic factors, history of abuse, affective disorders, neuroticism, long period of infertility and expo-sure to negative stories about delivery (50). Thereby, even a medical uncom-plicated vaginal delivery may be experienced as negative for the individual woman.

Women with fear of delivery have a higher risk of posttraumatic stress dis-order (PTSD) after delivery (56) and may also, in the worst case, decide not to become pregnant again (57, 58). Some studies indicate that women with fear of delivery have a higher risk of prolonged delivery, instrumental delivery and caesarean section (CS) (59-62), although some other studies show contra-dicting results (63, 64). A woman with fear of delivery is also more prone to request induction of delivery (65) or CS (62, 66-69). This is probably in an attempt to deal with the expectant humiliation and loss of control, the fear of pain, and the fear of sequels for herself and the foetus (70). Apart from the psychological consequences, women with fear of delivery often require more perinatal healthcare from both midwifes and physicians, which generates higher perinatal costs (69).

The Swedish national medical guidelines for CS on maternal request em-phasize the importance of weighting risks between a non-medical intervention

15

versus the potential risks of not performing the intervention (71). In the case of CS on maternal request, a weighting of short and long-term consequences for the individual woman and the child is in focus, but also ethical considera-tions regarding the woman´s autonomy and the principle of justice should be taken into account (71).

The most widely used validated questionnaire, both in the clinic and the research setting, for assessing maternal expectations about and experience of delivery is The Wijma Delivery Expectancy/Experience Questionnaire (W-DEQ) (version A for expectancy, version B for experience) which consists of 33 items (72). However, some studies use just one or a few questions with three to five response options on a Likert-scale (73-76).

Delivery complications Although diverse and divergent, the global maternal mortality and morbidity rates are fortunately decreasing (7). Nevertheless, the delivery process is in-evitably associated with a number of maternal and infant complications. The most common maternal complications include PPH, severe vaginal lacera-tions, and infection (77). The delivery process itself can be prolonged and complicated, and unfortunately, become a negative experience that might even cause posttraumatic stress symptoms (78, 79). The strongest delivery-related risk factors associated with PTSD are operative delivery, pain, negative deliv-ery experience, and infant-related complications (56, 79). There is also a strong co-morbidity between delivery-related PTSD and perinatal depression (56, 80, 81). Moreover, there are some studies reporting an association be-tween pregnancy complications, negative delivery experience and PPD (38, 40, 82, 83). Summarized, this gives a hypothetical framework of a plausible association of delivery complications with PPD, which has been only partly studied (84, 85).

Postpartum Haemorrhage PPH is a leading cause of global maternal mortality, accounting for 25% of all maternal deaths (86). The prevalence has a wide range, with estimations from 6 to11% (87, 88). The last estimates from Sweden were 4.7% for vaginal de-livery and 14% for CS (27).

There are several definitions of PPH. One of the most common ones, de-fines PPH as a blood loss of > 500 mL at vaginal delivery or > 1000 mL at CS (89). Another well-established definition, and the one used in Swedish clinical practice, defines PPH as a blood loss of > 1000 mL within 24 hours from delivery, regardless of mode of delivery (25). The reasoning for this cut-off is that bleeding below 1000 mL is likely to be well tolerated in women without chronic underlying medical disorders such as anaemia or cardiac disease (90).

16

Other definitions are based on the reduction in Haemoglobin concentration (Hb) (91) or the need for transfusion or uterotonics (92-95), hysterectomy (92), or embolization (96). It should be noted that the clinical estimation of the blood loss is usually underestimated in cases of bleeding > 500 mL (97-99).

The four most common causes of PPH are uterine atony, genital tract trauma, retained placenta, and coagulopathy (100) where atony is the most common cause. Multiparity, polyhydramnios, large foetus, chorioamnionitis obesity and induced or prolonged delivery predispose to uterine atony. (89) Other well-established risk factors for PPH are CS, especially emergency CS, instrumental delivery and retained placenta (100). Notably though, most women with PPH do not have any known risk factors (100).

During the last two decades, an increasing rate of PPH has been observed in high-income countries (97, 99, 101, 102). However, because of the hetero-geneity in PPH classification, an accurate estimation of this overall increase is not possible. Nevertheless, Briley et al. reports a three times higher risk for PPH >1500 mL between 1997/98 and 2008/09 in Scotland (99). Interestingly, the increase in PPH incidence cannot be fully explained by the time trends in currently known risk factors such as higher maternal age, obesity, and CS (97, 101). Other proposed factors that might explain this increase include a more liberal approach for longer deliveries, changes in the management of the third stage of delivery, increasing rates of induction, and a complex interaction be-tween several risk factors (97, 102, 103). Another hypothesis suggests the in-creasing use of antidepressants as an explanation for this increase (104-106). This hypothesis may not be the sole explanation, as only 2-3% of pregnant women in Europe and 10% of women in North America are treated with anti-depressants (107), but in the context of perinatal complications and PPD, the hypothesis is worth taking into consideration. Selective serotonin-reuptake in-hibitors (SSRIs) decrease thrombocyte aggregation and therefore contribute to an increased risk of bleeding (108), and thus PPH (106, 109). SSRIs may even have an atonic effect on the uterus (110, 111). Interestingly, however, even antidepressants other than SSRIs seem to have a positive association with PPH. Proposed explanations include an effect of these non-SSRIs drugs on coagulation by a shift in the neurotransmitter balance (105) or an effect of the underlying psychiatric disorder to prolonged delivery (105).

Consequences of PPH include anaemia and traumatic experience of deliv-ery, which are both independently associated with increased risk for PPD (112-117). However, the literature on the association of PPH with PPD is sparse and the studies have diverging designs, definitions, and results (91, 96, 118). Importantly, the majority of the studies have not taken into account vul-nerability in the form of earlier psychiatric morbidity or psychological or so-matic consequences of PPH, for example the possible subjective experience of a threatening occurrence, or the following symptoms of anaemia.

17

Anaemia The most common definition of anaemia during pregnancy is Hb < 110g/L (119, 120) which differs from the standard definition in the general population due to the haemodilution and hypervolemia present during pregnancy. The definition of postpartum anaemia is more heterogeneous, in both time and cut-off in Hb values, which might range from 100 g/L to 120 g/L (114, 121, 122). After delivery, Hb falls slightly during the first 24 hours because of blood loss, and then starts to increase again over the next 2 to 5 days due to haemocon-centration (119). Hb is expected to return to pre-pregnancy concentration, i.e. Hb ≥120 g/L, at 6-8 weeks postpartum. The Hb is followed up at 6-12 weeks postpartum at the maternal primary care unit where all women are offered an appointment. There is no consistent evidence on the association between anae-mia and PPD, with some studies identifying anaemia as a risk factor (112-116), while others not (121).

Mode of delivery Mode of delivery may be categorized in four groups: spontaneous vaginal de-livery, instrumental vaginal delivery with vacuum extraction (VE) or forceps (the later uncommon in Swedish practice, according to the guidelines of the Swedish Society of Obstetrics and Gynaecology (123)), elective CS, and emergency CS. The distribution of the different modes of delivery is globally diverse, but there is a marked trend for increasing frequencies of CS world-wide with rates exceeding 40% in Latin America and 25% in Europe (124). There is also a substantial divergence within the countries, due to the popula-tion sample and the size of the study hospitals (27, 124). In Sweden, the re-spective figures in 2015 were 74 % for spontaneous vaginal delivery, 8% for VE, 8% for elective CS, and 10% for emergency CS (125).

The majority of studies examining the association between mode of deliv-ery and PPD report no direct association (85, 126-130). However, there is re-cent evidence suggesting that specific factors might act as potential mediators or confounders of this association (85, 131, 132). For example, a Swedish reg-ister study reported an association of vaginal instrumental delivery or CS with severe PPD for women without history of depression but not for women with a history of depression (38). In the Netherlands, where a substantial but de-clining number of deliveries is performed at home, a study reported that emer-gency CS was associated with PPD (133). A Canadian study reported that CS was associated with increased risk of PPD only among native-born mothers, and with a lower risk in non-native women (85). These results indicate that sociocultural factors may influence the beliefs about CS and may thereby have a confounding role in the interplay between mode of delivery and PPD. Other studies report that antenatal depression (134, 135) as well as discrepancy be-tween the women’s preference and actual mode of delivery might influence

18

the risk of PPD (134, 136). Finally, fear of delivery, which is associated with history of depression (137, 138) and might influence delivery experience (139) affects preferences for mode of delivery (66). Some studies indicate an association of fear of delivery with both elective and emergency CS (62, 75, 139-141).

Pain experience One of the major components of delivery experience is pain (142). Pain during delivery is often more tolerantly accepted by the women, as compared to other forms of pain (143). Yet, pain may contribute to posttraumatic stress symp-toms even though the overall delivery experience seems to be more important for the development of PTSD (144, 145). However, despite the central role of pain in delivery experience, there is no evidence that effective pain relief per se is associated with satisfaction with pain relief, sense of control, and general satisfaction with the delivery experience (6). This may seem a paradox, but the pain experience includes, apart from effective pharmacological analgesia, complex physiological, psychological, and cultural components (146-150). During the various stages of delivery, diverse physiological processes contrib-ute. At the first stage of delivery, the pain originates from the dilation of cervix and uterine contractions, so a visceral component dominates. In the second stage, the visceral pain is combined with somatic pain caused by distention of the vagina and perineal structures (5). Psychological and cultural components might include fear of delivery (148), depression and anxiety (148, 151, 152), expectation of pain and actual intensity (146, 153, 154), sense of control (143, 146, 155-157), and support from partner and midwife (143). As Hall and her colleagues conclude:”… pain and comfort were not opposites, but rather part of the whole emotion-sensation-environment landscape of birth” (158).

Moreover, the experience and memory of pain during delivery might change over time. For most women, the pain experience, is less intense in ret-rospect (152, 153, 159), but this is not the case for women with an overall negative delivery experience (160, 161). A theoretical framework for pain ex-perience is that sensory and emotional dimensions affect the momentary pain experience during delivery, whereas cognitive and evaluative dimensions in-fluence the retrospective memory of pain, which thereby is a rather mental process (160, 162).

Epidural Analgesia Most women require some sort of pain management during delivery. Non-pharmacological interventions aim to help women cope with pain, whereas the aim of the pharmacological interventions is to relieve pain (5). Among pharmacological interventions, EDA is the most effective method .Continuous

19

support and midwife-led care is associated with lower use of pharmacological pain relief (163, 164).

The frequency of EDA use during delivery varies across different settings, being as high as 90% in some hospitals for nulliparous women (165). The overall frequency of EDA use among nulliparous and multiparous women in Sweden is 53% and 21%, respectively (166), but with a high disparity between hospitals (27). Moreover, women receiving EDA more often report a history of depression and fear of delivery (167). EDA is associated with induction and high infant birth weight (167, 168). There seems to be an association between socioeconomic status and use of EDA, but whether low or high socioeconomic status is associated with a higher prevalence of EDA is not consistent across studies (167, 169, 170). A plausible explanation could include different health care systems(167, 169, 170).

Women with EDA report less pain in both the first and second stage of delivery, and are generally satisfied with pain relief (6). However, there are side effects associated with the use of EDA; hypotension (EDA may be used in the clinical practice if a women is hypertensive during delivery), motor blockade, maternal fever, longer second stage of delivery and thereby aug-mented need of oxytocin, higher risk of instrumental delivery, and CS (6).

As mentioned above, there is no evidence that women using EDA have a more positive delivery experience or that the sense of control is affected (6). Whether the intense pain during delivery is associated with PPD or other post-partum mental illnesses is still unclear (128, 171). Some studies have reported that effective pain relief leads to a reduction in the risk of PPD (172-174). Particularly, a Finnish study found that pain relief by EDA or paracervical blockade, but not by nitrousoxide, was associated with decreased odds of de-pressive symptoms directly after delivery (174). In the same context, a recent prospective study in China reported that use of EDA during delivery is asso-ciated with decreased rates of depressive symptoms 6 weeks postpartum (172). However, an English randomized trial and an Israeli cohort study did not find any association (175, 176), although the latter study reported a slightly increased risk of PPD when there was a mismatch between intended and actual use of EDA (175).

Delivery experience The delivery experience is a complex, multidimensional, and subjective phe-nomenon (157, 177) that has been thoroughly studied during the last twenty years, especially in the Nordic countries. There is no consensus about the def-inition of delivery experience and different aspects of the phenomenon pin-pointed in different contexts and studies. There are multiple instruments for scoring delivery experience with different psychometric values (178). W-

20

DEQ version B (72) is the most widely used one, which is an extensive ques-tionnaire with cultural validation and translations (178).

Because of the multidimensionality, subjectivity and dynamical pattern, some researchers argue that it is not possible to quantify delivery experience. Thereby, they advocate for a qualitative approach (157, 179). However, de-spite the ongoing discussion on methodology, there are some factors associ-ated with women´s overall experience of delivery that come up consistently; perceived control, social support and pain (157, 180). Women usually describe their delivery in a combination of negative and positive terms that may not be separated (158). But also, the delivery experience dynamically changes during delivery (158) as well as in the long-time course (181-183).

A negative delivery experience is a common reason for a request of elective CS in subsequent pregnancies (68, 184) whereas a positive delivery experi-ence may result in feelings of empowerment and self-efficiency (149, 182) during the transition to motherhood (157, 185).

Postpartum depression Like other psychiatric diseases, the diagnosis and treatment of PPD is ideally based on the LEAD procedure 185). This implies that the diagnosis is based on anamnestic history of the patient, and sometimes even from relatives, in combination with physical ex-amination and the use of questionnaires and diagnostic interviews. The results are compiled by a team of experts who discuss the diagnosis and the severity before taking a decision of treatment, which should always be discussed with the patient as well. Thereafter, the treatment is continuously evaluated, pref-erably with standardized questionnaires (186). However, this is a time-con-suming process. Therefore, the Swedish recommendation is to base the diag-nosis for affective disorders on structured or semi-structured interviews, such as MINI (Mini International Neuropsychiatric Interview), or SCID-I (Struc-tured Clinical Interview for DSM-IV) which have both higher specificity and sensitivity compared with a solely clinically-based diagnosis (186).

Definition There is no consensus about the diagnostic criteria for PPD between the two existing classification systems, ICD-10 (International statistical classification of diseases and related health problems, 10th revision), developed by the World Health Organization, and DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition), developed by the American Psychiatric Association. In ICD-10, PPD is included in the diagnosis of “Mild mental and behavioural disorders associated with the puerperium” (F53.0) if onset within the first six weeks postpartum. In DSM-5, PPD is not categorized as a unique

21

diagnosis but as a major depressive episode with peripartum onset defined as onset of symptoms during pregnancy or within four weeks after delivery. However, in clinical praxis and for research purposes, this period is often pro-longed up to a year postpartum (187-190). In Sweden, the DSM-5 is used for both psychiatric outpatient and inpatient care but the diagnostic codes are con-verted to ICD-10 for reporting to the National Patient Register.

Symptoms of PPD do not differ from those of depression in general. They include depressed mood, loss of interest or pleasure, change in weight or ap-petite, insomnia or hypersomnia, psychomotor retardation or agitation, loss of energy or fatigue, worthlessness or guilt, impaired concentration or indeci-siveness, and recurrent thoughts of death or suicidal ideation or attempt. Nev-ertheless, obsessions and ruminations focusing on the newborn and suitability for mothering are more commonly present in PPD symptomatology (191, 192). Furthermore, changes in sleep and energy pattern are common postpar-tum and may thereby have a lower diagnostic reliability for diagnosing PPD than depression in general. Noteworthy, PPD should not be confused with the so called “baby blues“, which is a self-limiting feeling of low mood in the first weeks after delivery affecting up to half of all women (193).

The severity of PPD ranges from mild symptoms with natural recovery within some weeks to long-lasting episodes with psychotic symptoms that re-quire in-hospital treatment (193).

Prevalence PPD is one of the most common perinatal complications (188) and is often a continuation of an antenatal depression, especially if the latter has been sub-optimally treated (187, 194, 195). PPD is also more common among women with a history of depression even before pregnancy (187, 190, 193). In high-income countries, the point-prevalence of PPD at 3 months postpartum is es-timated to 13%, (188), but up to 20% of the women experience some postpar-tum depressive symptoms (PPDS) (187). Some studies also indicate an in-creasing prevalence of PPD (112). However, the comparability of prevalence across different studies is difficult due to differences in definitions, study de-sign, study population, and geographical setting (196-199). The majority of studies define PPD using symptom-based questionnaires instead of a struc-tured clinical assessment. This approach may overestimate the prevalence be-cause common maternal stressors, such as lack of sleep, may overlap with surveyed symptoms (200). On the other hand, the use of clinical diagnosis may lead to an underestimation of prevalence as women often suffer from de-pression in silence (196). In the literature, the most common study design is clinical follow-up studies based on care-based clinics or referral centres, which could negatively affect the representability (188)Although register-based studies also exist, they may underestimate prevalence estimates since many women do not seek medical care (201). According to a WHO report ,

22

women from low and middle-income countries have a higher prevalence of PPD and other postnatal psychiatric morbidity (202).

Comorbidity Postpartum as well as antenatal depression are often comorbid with anxiety and PTSD (203, 204) .Experiencing the first-ever depressive episode during the first weeks postpartum is more often associated with a future diagnosis of bipolar disorder (205). Additionally, women with bipolar disorder have a very high risk for developing postpartum psychosis (190, 205).

Aetiology and risk factors Stressful life events, lack of social support, domestic violence or history of abuse and a burdened psychiatric history are risk factors with strong associa-tions with PPD (187, 190). Risk factors with moderate effect size are depres-sion and anxiety during pregnancy, neuroticism, postpartum blues, and poor marital relationship (82, 187, 190). Other risk factors include low socioeco-nomic status, absence of partner, unwanted pregnancy, obstetrical stressors (38, 82, 131), lack of breastfeeding, and illness/problems with the infant (190, 193). From a pathophysiological perspective, there is also evidence that sleep disturbances, hormonal changes, and genetic variations associate with the risk of PPD (190, 196). Hormonal changes include gonadal steroid hormones which might modulate neurotransmission and neuroplasticity, oxytocin, thy-roid hormones, changes in the HPA-axis, and neuroimmune pathways (190). Genetic factors might refer to either inherited variants that directly increase the risk of PPD or to epigenetic alterations that might influence expression of the genome and might be the consequence of environmental burden, such as psychological stressors (190). Also, some studies indicate differential gene expression in perinatal depression with an antenatal debut as compared to per-inatal depression with postnatal debut (206, 207).

Screening Although the diagnosis of PPD should be based on a clinical psychiatric inter-view evaluating the criteria for a major depressive episode (186), screening interviews, or even more common, self-reporting validated questionnaires are used in both research and in the clinical primary setting. A widely used ques-tionnaire is Edinburgh Postnatal Depression Scale (EPDS) (208). The effi-ciency of EPDS for screening for PPD has been debated and ethical consider-ations have been raised (209). There is low to moderate strength of evidence for the efficacy of screening for reduction of depressive symptoms postpartum and improved mental health (186, 210) and EPDS is the only questionnaire with sufficient evidence (186, 211). The potential effectiveness of screening

23

for PPD appears to be dependent on the adequacy of follow-up for women with positive screening (186, 210). Also, as already mentioned, it is important to identify women with bipolar diagnosis with postpartum debut (190, 194).

The Swedish guidelines recommend screening for PPD by using the EPDS at 6-8 weeks postpartum at the child primary care unit or at the postpartum control at the maternal primary care unit (212, 213); the latter, however, is less common in practice.

The purpose of screening is to identify both women at high risk for PPD as well as to diagnose those who already have clinical depression. In case of scoring 12 or more points in EPDS, or if a need of psychological support be-comes apparent during the consultation, the national guidelines recommend person-centred counselling at the paediatric nurse as a first step. The second line is referral to psychologist in the maternal or paediatric primary care, or physician in the primary care system (212, 213).

Prevention There is evidence for the use of psychosocial or psychological interventions to decrease the risk for developing PPD (214-216). However, due to the lim-ited number of high-quality studies, some uncertainty still remains (214, 215). Additionally, the notable diversity in study designs makes drawing firm con-clusions difficult (214, 215). In 2016, the National Institute of Health Care in the United Kingdom performed a systematic review and meta-analysis of stud-ies of antenatal and postnatal prevention of PPD. Based on the findings, they classified three levels of prevention: prevention for all pregnant women (uni-versal), prevention for pregnant women with social problems (selective), and prevention for women with psychological risk factors (indicative). Result- and cost-effective methods include midwifery redesigned care (universal preven-tion), person-centred approach (universal), cognitive-behavioural therapy (universal), education on preparing for pregnancy (selective), and interper-sonal therapy (universal and indicative). These interventions should be indi-vidually designed and could additionally include psychosocial interventions such as social support through group-based approaches (215), home visits by nursing staff close after the delivery, or lay-man based support by telephone (214). Overall, continuity of care is important for building a trustful relation-ship, which is an important determinant of the success of interventions (215).

A Cochrane systematic review in 2018 concluded that there is limited evi-dence about the efficacy of antidepressants for prevention of PPD in women at high risk for developing PPD (217).

Treatment Women with PPD have described that the process of seeking professional help induced feelings of shame and reported different barriers for help-seeking.

24

Nevertheless, they express gratefulness after receiving help (200, 201, 218). The treatment of choice in mild to moderate depression is psychosocial (peer support guided self-help, person-centred approach) or psychological therapy (189, 219), preferentially interpersonal therapy (220, 221), or cogni-tive behavioural therapy (222). Other effective non-medical therapies as self-help interventions (223) and internet-based cognitive-behavioural therapy (224) may also be effective, but there is less evidence on these methods. Again, there is a lack of high-quality studies, which limits the extraction of meaningful conclusions (189, 191, 219). Also, most trials have compared non-medical treatment with standard care of PPD, which might also include anti-depressants, thus making interpretations more difficult (189). Qualitative studies point out the importance of continuous and good relationship with the caregiver (201, 218).

Sometimes there is a need for pharmacological treatment, either alone or in combination with non-medical treatments (190). After ensuring that the woman does not suffer from bipolar depressive disorder, the first line medical treatment should be SSRIs, particularly sertraline that has only minimal pas-sage to the breastmilk (190, 225, 226). However, for women with a history of depression and a former effective treatment, the same treatment approach should again be considered (190, 226). Since anxiety is commonly seen with PPD and metabolism may be slower in the first few weeks postpartum, starting the antidepressant at a low dose may be considered (198, 227). In case of treat-ment failure, the diagnosis should be re-evaluated. Differential diagnosis of PPD might also include underlying somatic disorders such as thyroid disease, bipolar disorder or an early presentation of a first episode of psychosis (190). After ensuring the diagnosis, an augmentation with benzodiazepines, antipsy-chotics and mood stabilizers may be necessary (225). For treatment-resistant severe depression, treatment with electroconvulsive therapy may be consid-ered and is recommended by the Swedish Psychiatric Association (228-230).

Consequences PPD is one of the most common contributors to maternal morbidity (231), even though the diagnosis is often ignored in studies focusing on maternal somatic morbidity (232). PPD might have major consequences for both the mother’s future affective status and the child’s development. In particular, PPD might impair the mother’s relationship with the partner and infant, im-pact on mother-infant bonding and negatively influence the child’s emotional and cognitive development (187, 190). Suicide in the prolonged puerperium is the most severe and feared consequence with a ratio of 3.7 per 100000 live births (233). However, it is also important to note that the rate of suicide in the first year postpartum is lower when compared to the respective rate for women who have not given birth (234-236).

25

Apart from all consequences for the individual woman, her child and her family (187, 190), there is also a need to address economic considerations from a public health perspective. In 2014, the London School of Economics reported that the aggregated cost for perinatal depression and anxiety was £6.6 billion in the United Kingdom and thereby concluded the need to allocate more resources to support women with perinatal mental illness (237, 238).

Personality Personal characteristics may be described using concepts as temperament, character or personality. Temperament is closer to the ancient Greek “humoral theory” and represents the stable biological factors as genetics and inherent endowments that affect the individual´s activity level, mood, emotionality and sociability. Traditionally, temperament has been attributed to infant and chil-dren patterns (239). Character, on the other hand is more connected with nur-ture and refers to the competences acquired during development (46). Person-ality is a broader concept enclosing both constitutional and learned features associated with psychological and behavioural traits. Even if there is no uni-versal definition, personality can be defined as the individual´s characteristic pattern of cognitions, emotions and behaviours (240). Traditionally, personal-ity has been assumed to be relatively stable from adulthood and throughout life, but this assumption has been doubted during the last two decades (239, 241-244). The prevailing theory now is that personality starts to develop dur-ing childhood and although it becomes relatively stable over adulthood, it could still be influenced, for example, by current affective state (244). Later on, in the senescence, a normative change of the personality is common with higher emotional stability, sociability and conscientiousness (242, 245). Also, as the central nervous system ages and social circumstances change, the tem-perament may be enhanced (242).

There are several personality models and instruments that group personal-ity into larger categories, or factors (246). The most widely used model is the Five-Factor Model (FFM), also called the Big Five, which includes the factors of Neuroticism, Extraversion, Openness to experience, Agreeableness and Conscientiousness. NEO-PI (Neuroticism, Extraversion, Openness Personal-ity Inventory) is the most well-known personality questionnaire consisting of 240 items (247).

Neuroticism One of the most well-studied personality factors is neuroticism, which might also be termed as negative affectivity, low emotional stability or harm avoid-ance (239, 247, 248). Neuroticism is characterized by a tendency towards neg-ative emotions, such as anxiety, fear, irritability, and anger (239). It is also

26

closely related to self-catastrophizing and with the belief of one’s inability to manage or cope with challenging events (239). Moreover, neuroticism is as-sociated with problematic coping strategies, such as emotion-focused coping as opposed to problem-solving coping (249, 250), as well as withdrawal and wishful thinking (30). Individuals with high neuroticism have also been found to be more likely to have an apparent need for seeking health care (249).

It is well established that neuroticism is associated with anxiety and depres-sive disorders (239, 246) and has been suggested as a mediator between bio-logical and psychological vulnerability leading to affective disorders (239, 242). Pregnant women with neuroticism report more concerns about their pregnancies and higher rates of fear of delivery (251-253). They are also at a higher risk for postpartum blues (254) and PPD (254-256). In terms of physi-cal health, neuroticism is associated with smoking (255, 256), migraine (257), and cardiovascular disease (258, 259), whereas some studies point to a cluster of factors related to metabolic syndrome among individuals with high levels of neuroticism (260). Although neuroticism is one of the most extensively studies personality factors, studies exploring the role of neuroticism in perina-tal outcomes remain scarce. A few studies have found significant associations between neuroticism and adverse perinatal outcomes including preterm con-tractions (261), failure to progress, assisted delivery, emergency CS, severe lacerations, foetal distress and foetal growth restriction (262-264). Studies on the association between neuroticism and preterm birth have reported diverging results (251-253).

27

Rationale for the Thesis

Pregnancy and delivery are complex physiological processes with potential complications. It is also a period of life full of emotions, and even existential thoughts. We know that complications can affect not only somatic health but also psychological health, for example through traumatic memories from de-livery. However, it is still unclear if perinatal obstetric complications per se influence the development of PPD. The literature is contradictory, and some studies suggest that those diverging results could partly be explained by other factors, acting as confounders or mediators. The role of personality in this context, which is strongly associated with the risk for PPD, is sparsely studied. Therefore, the further exploration of the associations between perinatal com-plications, personality and perinatal depression is warranted.

28

Aims

General Aim The general aim was to explore the associations between perinatal complica-tions and perinatal mental health, with emphasis on postpartum depressive symptoms and neuroticism. Special focus has been put on major potential con-founding or mediating factors such as history of depression, antenatal depres-sive symptoms and delivery experience.

Specific Aims - to explore the association between postpartum haemorrhage and post-

partum depressive symptoms taking into account possible confound-ers and mediators, such as anaemia and negative delivery experience. (Study I)

- to explore the association between mode of delivery and postpartum depressive symptoms and explore the role of sociodemographic co-variates, history of depression, fear of delivery, obstetric complica-tions, and delivery experience, as confounders and mediators. (Study II)

- to investigate the association between use of epidural analgesia and

postpartum depressive symptom, considering several sociodemo-graphic, psychosocial and obstetrical variables. (Study III)

- to assess if maternal neuroticism is associated with adverse perinatal

complications. (Study IV)

29

Table 1. Overview of included studies Study Design Data Study

populations Exposure Outcome

I Population based, nested cohort study

Survey, medical records

446 women PPH

PPDS

II Population based, nested case-control study

Survey, medical records

3888 pregnan-cies

Mode of delivery

PPDS

III Population based, nested case-control study

Survey, medical records

1503 women EDA PPDS

IV Convenience samples, cross-sectional study

Survey, national registers

1969 women Neuroticism Adverse perinatal outcomes

PPH: Postpartum haemorrhage PPDS: Postpartum depressive symptoms EDA: Epidural analgesia

30

Material and Methods

Study populations All studies (studies I-IV) were based on the BASIC project, but the study pop-ulations in study I and study IV also included participants from other Uppsala-based studies.

The BASIC project (Studies I-IV) The BASIC (Biology, Affect, Stress, Imaging and Cognition) project is a pro-ject with a special focus on perinatal maternal wellbeing. This longitudinal project started in September 2009 and is still ongoing with over 5200 individ-ual women and 6100 pregnancies (March 2018). All pregnant women at 16th-18th gestational week attending the routine ultrasound examination in Uppsala University Hospital received written information about the study and were in-vited to participate. Exclusion criteria were inability to adequately communi-cate in Swedish, confidentially kept data, intrauterine demise, blood borne in-fectious disease, and age below 18 years. BASIC has a participation rate of 21%. Following informed consent, women were instructed to complete self-administered web-based questionnaires at recruitment (16th-18th gestational week), at the 32nd week of pregnancy and at 6 weeks, 6 months and 12 months postpartum. The response rate at 6 weeks postpartum was 81%.

Participating women differ from the background general pregnant women population as they report a higher educational level (61% with university ed-ucation vs. 52% in the Swedish Birth Register) and are more likely to be pri-miparas (54% vs 44%).

The UPPSAT project (Study I) The UPPSAT project (n=2318) is a population-based longitudinal study in-vestigating maternal mental wellbeing after delivery in Uppsala County, Swe-den. All women giving birth in Uppsala University Hospital between May 2006 and June 2007 were asked by their midwife at antenatal care to partici-pate in the study. Exclusion criteria included inability to adequately communi-cate in Swedish, confidentially kept data, intrauterine demise or admission of the newborn to the neonatal intensive care unit. Informed consent was ob-

31

tained by 65% of the target population. The women were instructed to com-plete self-administered questionnaires at 5 days, 6 weeks, 6 months and 12 months postpartum. Besides the 5-day-questionnaire, which was distributed directly to the women by the midwife, all subsequent questionnaires were sent by postal mail. The response rate at 6 weeks postpartum was 73%.

We compared the women who participated in the study with aggregated data for all women delivering in Uppsala University Hospital during the study period. The only difference was that the UPPSAT cohort recruited more nul-liparous women. No significant differences regarding maternal age, preg-nancy complications, delivery outcome, weight of the baby, or area of resi-dence were detected.

Studies including personality assessment (Study IV) Project of induced abortion A multi-center cohort study was conducted between September 2009 and June 2010 (n=1542). All women who requested an induced abortion before the end of gestational week 12 were approached for participation. Women were in-formed about the study during their registration for the first abortion visit. Those who agreed to participate received written information about the study and completed a questionnaire including Swedish universities Scales of Per-sonality (SSP) during the first visit. Inability to read and understand Swedish was the only exclusion criteria. The participation rate in the project was 59%.

Project on infertility This cohort study was conducted from August 2005 to April 2007 (n=323). All couples undergoing in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) treatment at the Centre of Reproduction at Uppsala University Hospital, Sweden, were approached for participation in the project. Exclusion criteria included inability to read and understand written Swedish, couples un-dergoing cycles with gamete donation, and earlier IVF treatment during the study period. The SSP questionnaire was distributed and completed at the day of oocytes retrieval. The participation rate was 76%.

Project on experience of combined oral contraceptive use For this case-control study women were recruited by announcement calls in newspapers and posters in health care centres between March and May 2006. In all, 285 women were screened by telephone interview. Exclusion criteria were ongoing pregnancy or breastfeeding, no presence of prior experience of combined oral contraceptives, treatment with hormonal contraceptives other than oral contraceptives, or current treatment with psychotropic drugs includ-ing SSRIs. After applying these exclusion criteria, 118 women participated in the project.

32

Project on Premenstrual Mood Disorder Participants for this project were recruited between October 2005 and Novem-ber 2006 among women seeking help for premenstrual symptoms in the out-patient ward of the Department of Obstetrics and Gynecology, Uppsala Uni-versity Hospital, and from newspaper advertisements (n=85). Thirty women met the criteria for premenstrual mood disorder, defined according to DSM-IV and were included in the project. An additional 55 asymptomatic healthy women were included as controls. Exclusion criteria were ongoing treatment with any hormonal compound and treatment with benzodiazepines or other psychotropic drugs including SSRIs in the last three months. After written in-formed content, all women answered SSP during the follicular phase.

Study design Study I Study I is a nested cohort study (n=446) including 168 participants from the UPPSAT project and 278 participants from the BASIC project. Data for this study were collected from September 2009 to November 2012.

Figure 1. Flowchart of included studies and participants in Study I

Study II Study II is a nested case-control using solely data from the BASIC project collected between September 2009 and December 2014. Exclusion criteria were duplex and women with missing data on mode of delivery, which re-sulted in a total of 3888 pregnancies. Some women participated more than once, resulting in 3508 unique individuals included in the analyses.

33

Figure 2. Flowchart of participants in Study II

Study III Study III is a nested case-control study based solely on data from the BASIC study that were collected from September 2009 to August 2017. Only nullip-arous women were included. Exclusion criteria were duplex, prematurity (ges-tational length <37 weeks), breech position, induction of delivery, elective CS, preeclampsia, and hypertension, resulting in 1503 participating women.

Figure 3. Flowchart of participants in Study III

Study IV Study IV is a convenience cross-sectional study with participants from several projects based in Uppsala, Sweden (n=1969). Data for the study were col-lected during the period 2005-2011 and were linked to national health regis-ters. Exclusion criteria were women who had not given birth or whose first delivery took place before 1984 or outside Sweden, and twin pregnancies.

34

Figure 4. Flowchart of the included projects and participants in Study IV

Variables The data were collected from self-administered questionnaires and medical journals (study I-III). For study IV, the data were also linked to several Swe-dish official health registers; the Medical Birth Register (MBR) (information about participants’ first delivery, somatic and psychiatric health, medication at first antenatal booking, pregnancy and delivery) (265), the Patient Register (diagnoses from hospital admissions and outpatient clinic visits from five years before delivery to one year after delivery) (266), the governmental agency Statistics Sweden (socioeconomic factors) and the Prescribed Drug Register (data for anxiolytic/sedative drugs and antidepressants) (267).

For the BASIC and UPPSAT projects, the first questionnaire (pregnancy week 17-19 and 5 days postpartum, respectively) contains socioeconomic background variables and information about earlier mental health. The major-ity of mental health variables were examined via validated questionnaires (see below). However, in order to achieve an acceptable questionnaire length (268), some factors were also examined by short questions specifically formu-lated for the current study.

Exposure The main exposure variables for the four study were PPH (study I), mode of delivery (study II), use of EDA during delivery (study III) and Neuroticism (study IV).

35

PPH was defined as bleeding ≥1000 mL within 24 hours after delivery, in accordance with the clinical practice in Sweden. Non-exposed individuals were randomly selected by including one every 12th woman in the register of participants. To reduce misclassification, women with postpartum bleeding estimated between 650 and 999 mL were not considered, and the next availa-ble participant was instead included.

Mode of delivery was classified in four categories; vaginal delivery, VE, elective CS and emergency CS. Elective CS was defined as planned CS during day-time. The definition of emergency CS was an unplanned CS performed before or after onset of delivery due to foetal or maternal compromise. Mode of delivery status was retrieved from the medical records. Similarly, infor-mation on the use of EDA during delivery was obtained from the medical rec-ords.

Neuroticism was assessed using the validated SSP. In contrast to the other exposures which were categorical (study I-III), the neuroticism score was ex-amined as a continuous variable.

Outcome The outcome in study I-III was presence of PPDS, as a proxy for PPD. PPDS was defined by a score of ≥12 in EPDS at 6 weeks postpartum (186).

For study IV, the outcomes included perinatal complications based on reg-ister linked data: mode of delivery (vaginal delivery, VE, any CS, elective CS, emergency CS), GDM, gestational hypertension and preeclampsia (grouped together due to low frequency and common aetiology), induction of delivery, dystocia (prolonged first or second stage of delivery), severe lacerations (grade 3-4, cervix rupture, or complex vaginal lacerations), placental reten-tion, and PPH. Neonatal outcomes included premature birth (before gesta-tional week 37+0), small for gestational age (SGA) (below the 10th percentile, according to the Swedish sex-specific reference curves for gestational age) (269), large for gestational age (LGA) (above the 90th percentile) and Apgar score below seven at 5 minutes. A composite worst-case variable was created, including intrauterine demise, eclampsia, severe preeclampsia, premature birth before gestational week 32+0, SGA < -2.5 SD (lowest 0.6%), and pla-cental abruption.

Covariates For study I-III, the questionnaires collected information on a wide range of pregnancy- and postpartum-related covariates, including socioeconomic fac-tors, history of depression, previous psychological contact (with psychologist or psychiatrist), mood during pregnancy, expectations of delivery, delivery experience, breastfeeding, sleeping habits in the postpartum period, and sup-port by partner.

36

Anaemia postpartum is closely linked to PPH and is thereby an important var-iable for study I. The definition of postpartum anaemia used in this study was a Hb below 110 g/L at discharge. In Sweden, Hb directly after delivery or at discharge is routinely assessed only in cases of PPH; thus, there were high rates of missing data concerning anaemia status at discharge. Hemodynamic changes induce a rise in Hb during the first days postpartum (119). Thus, an algorithm was created in order to assign a value for subjects not having anae-mia at discharge, even when not tested: the woman was classified as having no anaemia at discharge if the lowest antenatal Hb was 105 g/L and she had a postpartum bleeding less than 400 mL. Correspondingly, women with lowest antenatal Hb of >110 g/L and postpartum bleeding of <500mL and women with lowest antenatal Hb >125 g/L and postpartum bleeding of <650mL were considered as not suffering from anaemia at discharge. In all other cases, (i.e. in case of larger bleeding than the cut-off in the algorithm) women were clas-sified as having anaemia at discharge.

History of depression was self-reported by the single question “Have you ever had a depression?” at gestational week 17.

Presence of antenatal depressive symptoms was defined as a score of ³13 EPDS points at gestational week 32 (270).

Mood during pregnancy, i.e. a more general variable than antenatal depres-sive symptoms, was assessed at gestational week 32 (BASIC project) or ret-rospectively at 5 days postpartum (UPPSAT project) with the question “How have you been feeling during your pregnancy?”. The four alternative answers were “Better than usual”, “As usual”, “A little depressed” and “Depressed”. Based on these answers mood during pregnancy was dichotomized to “Good” (first two alternatives) vs “Depressed” (last two alternatives).

Expectations of delivery and fear of delivery were assessed by the question “How do you expect your forthcoming delivery?” with the possible answers: “okay”, “excited”, “good”, “longing for”, “worried”, “wish it was over”, “ter-rified”, “fear of vaginal delivery”, and “fear of caesarean section”. A dichoto-mized variable for expectations of delivery (positive: “okay”, “excited”, “good”, “longing for”; negative: “worried”, “wish it was over”, “terrified”) was created. Answers including both “positive” and “negative” choices, ex-cept for the answer “terrified” were classified as positive. If the woman an-swered that she was terrified, had fear of vaginal delivery or CS, or if she reported a visit to a consultant physician or midwife due to fear of delivery, she was classified as suffering from fear of delivery.

Delivery experience was assessed at 5 days postpartum in the UPPSAT study and at 6 weeks postpartum in the BASIC study with the question: “How would you describe your delivery experience?” with five alternative answers. The answers “Wonderful/Good/Ok” were grouped together, while the answers “Bad/Awful” were considered as suggestive of a negative self-reported deliv-ery experience.

37

In study IV, covariates were extracted from the national registers, including year of delivery, maternal age at delivery, maternal height, body mass index (BMI) at first antenatal care visit, educational level, smoking during preg-nancy, chronic somatic disease (pre-gestational hypertension, diabetes melli-tus, or chronic kidney disease), involuntary childlessness (≥1 year or in vitro fertilization), and psychiatric morbidity (including ICD codes for ‘mental dis-orders due to psychoactive substance use’, ‘affective disorders’, ‘anxiety, stress-related and somatoform disorders’, ‘eating disorders’, ‘personality dis-orders’, ‘disturbances of activity and attention’, and the prescription of anti-depressant or anxiolytic drugs during pregnancy).

Psychometric instruments EPDS EPDS is a self-administered instrument specifically developed for screening for postpartum depression. EPDS does not include any items concerning changes in appetite, sleeping disturbances, fatigue, and somatic symptoms that relate to depression in general but also normally occur in the puerperium. (208). Specifically, it consists of 10 items and women are asked to answer questions on a fourth-grade Likert-scale based on their feelings in the last 7 days (208). The questionnaire is translated to numerous languages and is con-sidered the only screening tool with sufficient evidence for clinical screening and for use in epidemiological studies focusing on depression in the postpar-tum setting (121, 186). The optimal cut-offs for diagnosis of PPD differ across countries, but cut-offs of 9/10 in Asian countries and of 12/13 in most Euro-pean countries seem to offer balanced rates of sensitivity and specificity (186, 271). Particularly, a cut-off of 12/13, has a weighted sensitivity of 72% and specificity of 88% with SCID-I as the reference standard (186). The Swedish version of EPDS, compared to Montgomery-Asberg Depression Rating Scale (MADRS) as the reference standard, has a sensitivity of 96%, a specificity of 49% (272) at a cut-off of 11/12.

The EPDS has also been validated as a screening tool for antepartum de-pression with a cut-off of 12/13 offering a sensitivity of 77% and specificity of 94% in the Swedish setting (270). In a review by the Swedish Council on Health Technology Assessment and Assessment of Social Services, the weighted specificity is 96%, but it was difficult to evaluate the sensitivity due to the heterogeneity of the included studies (186).

SSP The Swedish universities Scales of Personality (SSP) is a self-reporting vali-dated questionnaire assessing personality traits. SSP was further developed

38

from the Karolinska Scales of Personality (KSP) with the aim to have a high psychometric quality, to be suitable for both clinical work and research, and to be written in simple language for both psychiatric patients and healthy sub-jects. In contrast to many other personality inventories, SSP does not aim to cover the entire personality spectrum but to assess specific areas with research interest (273). The objective was to have a biological approach to personality traits (273). It consists of 91 items with the response format of a four-graded Likert-scale ranging from “does not apply at all” to “applies completely” (273). The answers are analysed into thirteen scales, all consisting of seven claims; Somatic trait anxiety, Psychic trait anxiety, Stress susceptibility, Lack of assertiveness, Impulsiveness, Adventure seeking, Detachment, Social de-sirability, Embitterment, Trait irritability, Mistrust, Verbal trait aggression, and Physical trait aggression. These scales could in turn be coalesced in three factors, including Neuroticism, Aggressiveness, and Extraversion. Because of mean gender differences in the trait anxiety, trait aggressiveness and detach-ment scales, the results on each scale have been gender-adjusted according to the Swedish reference population (273, 274), creating normative T-scores (mean =50, SD =10).

Neuroticism, the most well-studied factor, consists of 42 statements which format six traits; somatic trait anxiety, psychic trait anxiety, stress susceptibil-ity, lack of assertiveness, embitterment, and mistrust.

Statistical analyses Univariate analyses In the first step of the statistical analysis for all studies, all covariates were cross-tabulated by the main exposure variable (study I-IV), and in Study I-III also by the main outcome, PPDS. Associations were statistically assessed with the Pearson chi-square test or Fisher´s exact test (in case of n≤5) for categori-cal variables. For continuous and asymmetrically distributed variables the as-sociations were assessed by Mann-Whitney for binary variables, and Kruskal-Wallis for three or more categories. P-values ≤0.05 were interpreted as statis-tically significant.

Logistic regression Logistic regression models were designed with dichotomized EPDS score (≥12 vs. <12) at 6 weeks postpartum as the dependent variable in study I-III. The independent variable was the main exposure (PPH, Mode of delivery, and EDA, respectively). Furthermore, in Study I, logistic regression analysis was repeated with anaemia at discharge instead of PPH as the independent variable

39

due to strong co-linearity with PPH. Adjusted models further included rele-vant confounders.

In study IV, the logistic regression used neuroticism as the independent variable, in its continuous form. Odds ratios (ORs) for the dependent variables (perinatal complications), with 95% confidence intervals (CIs), were calcu-lated for a 63-unit increase in neuroticism, which equals the interquartile range (IQR). Since the model is linear, this may also be interpreted as the OR be-tween women at the 75th and 25th percentiles of neuroticism.

Path Analysis In order to explore an association characterized by a complex interplay be-tween a number of potentially confounding or mediating variables, new sta-tistical methods have been developed during the last decades. Path analysis is such an example. Although it has been mainly used in behavioural sciences, during the last decade an increasing number of studies in the fields of psychi-atry and even perinatal depression applies path analysis (275-280).

In the current thesis, path analysis was performed to delineate the associa-tions of PPH and mode of delivery with PPDS at 6 weeks postpartum. We applied Structural Equation Modelling (SEM) (study I) and Generalized Structural Equation Modelling (GSEM) (study II), which is a further develop-ment of SEM.

SEM is a multivariate statistical analysis for analysing structural associa-tions between measured variables and latent constructs by using a combina-tion of multiple regression analysis and factor analysis (281). Thereby, esti-mates of multiple and interrelated dependence are obtained.

Based on relevant literature and results from the univariate analyses, we designed conceptual models including associations between the main expo-sure and outcome, as well as potential confounders and mediators (fig 5, fig 6).

40

Figure 5. Graphic display of the conceptual path analysis model reflecting the direct and indirect effects on postpartum depressive symptoms (PPDS) at 6 weeks postpar-tum with postpartum haemorrhage (PPH) as exposure.

Figure 6. Graphic display of the conceptual path analysis model reflecting the direct and indirect effects on postpartum depressive symptoms (PPDS) at 6 weeks postpar-tum with mode of delivery as exposure (VE: Vacuum extraction, CS: Caesarean sec-tion, VaD: Vaginal delivery)

41

Ethical considerations

All women were given oral and written information about the course and aim of the study and gave their written informed consent. Further, they received information on the voluntary nature of their contribution and that they had the freedom to withdraw from the study whenever they wished it, in accordance with the Declaration of Helsinki. The study protocols were approved by the Regional Research and Ethics Committee of Uppsala (UPPSAT: Dnr 2006/150, August 2006. BASIC: Dnr 2009/171, July 2009. Study IV: Dnr 2014/092, June 2014.)

Projects such as BASIC and UPPSAT have considerable ethical considera-tions relating to respect for both the woman and the child. Several questions and questionnaires evaluate very personal information and may thus evoke diverse feelings or even leave women in a more vulnerable status. Thereby, after every questionnaire all participating women are given explicit infor-mation on how to get professional psychological or psychiatric help if needed. The women are further provided with the contact details of the researchers responsible for the projects, in order to discuss and get support and further information. In some cases, women were offered a visit in the research labor-atory for further evaluation and a referral for further care was sent, if needed. Because some of the questionnaires also include items about self-harm and suicidal attention, all women with affirmative answers in these questions were contacted for a clinical assessment and thereafter, if needed, referred to a ther-apist or physician. This procedure has in some cases allowed a rapid detection and treatment of PPD, which in other way may have been undiagnosed for a long time due to the shame and the experienced barriers for help-seeking. It is also possible that women, by completing the questionnaires, reflected on their psychological health and thereby sought help on their own.

42

Results

Study I Totally, 466 women were included in the analysis of Study I. Among them 196 women experienced PPH, defined as bleeding of ≥1000 mL. Fifty-three women had EPDS ≥12 p at 6 weeks postpartum, indicating presence of signif-icant depressive symptomatology.

In the multivariate logistic regression analysis, there was no significant as-sociation between PPH and PPDS. However, there was a significant associa-tion between anaemia at discharge and PPDS (aOR 2.29, 95%CI 1.15-4.58), even after adjustment for confounders (table 2). Adjustments were made for previous psychological contact, mood during pregnancy and lack of exclusive breastfeeding 6 weeks postpartum.

Table 2. Multivariate logistic regression derived odds ratios (ORs) and 95% Confi-dence Intervals (95% CI) for postpartum depressive symptoms (PPDS) at 6 weeks postpartum

Unadjusted OR (95%)

Model 1* aOR (95% CI)

Model 2** aOR (95% CI)

Postpartum haemorrhage (≥1000 mL vs ≤650 mL)

1.38 (0.78-2.45) 1.81 (0.91-3.57)

Anaemia at discharge from hos-pital (Hb<110g/L vs Hb≥110g/L)

2.10 (1.15-3.82) 2.29 (1.15-4.58)

Previous psychological contact 2.28 (1.22-4.25) 2.08 (1.05-4.10) 1.90 (0.95-3.80) Mood during pregnancy (Depressed vs good/ok)

3.52 (1.18-8.83) 3.02 (1.51-6.06) 2.87 (1.40-5.87)

Lack of exclusive breastfeeding (6 weeks pp)

2.85 (1.58-5.15) 2.30 (1.16-4.56) 2.41 (1.20-4.85)

*Model 1 includes PPH and possible confounders (previous psychological contact, mood dur-ing delivery and lack of exclusive breastfeeding) as independent variables **Model 2 includes anaemia and possible confounders (previous psychological contact, mood during delivery and lack of exclusive breastfeeding) as independent variables

Path analysis revealed direct and positive effects of previous psychological contact, depressed mood during pregnancy, anaemia at discharge, self-re-ported negative delivery experience, and postpartum stressors on odds for PPDS at six weeks postpartum (fig 7). PPH did not have a direct effect on PPDS but indirectly increased its odds through anaemia at discharge and neg-ative delivery experience.

43

Figure 7. Graphic display of the estimated path analysis models through which post-partum haemorrhage (PPH), anaemia and other delivery related variables as well as earlier psychological contact, lack of exclusive breastfeeding and inadequate sleep at 6 weeks postpartum influence postpartum depression symptoms (PPDS) at 6 weeks postpartum.

The arrows represent regression equations used to assess mediation. The strength of the rela-tionship between two variables was estimated as a standardised regression weight (i.e., path coefficient, a, corresponding to correlation coefficients). Pathways indicated with a continuous arrow were statistically significant (p<0.05). Dotted arrows represent pathways with a p=0.05-0.20).

Study II A total of 3888 pregnancies were included in the study. Out of them, 363 (9%) referred to women who participated more than once. Seventy-four percent of women had spontaneous vaginal delivery, 8% VE, 9% Elective CS and 9% Emergency CS. Compared to vaginal delivery, women, who delivered by Emergency CS had higher odds for PPDS six weeks after delivery in the crude analysis (OR 1.45, 95%CI 1.04-2.01), but the association was completely at-tenuated in the multivariate model (table 3). Adjustments were made for par-ity, history of depression, fear of delivery, negative delivery experience, ob-jective postpartum complications, and self-reported physical symptoms at 6 weeks postpartum.

44

Table 3. Multivariate logistic regression derived odds ratios (ORs) and 95% Confi-dence Interval (95% CI) for postpartum depressive symptoms (PPDS) at 6 weeks post-partum

Unadjusted OR (95% CI)

Adjusted Model aOR (95% CI)

Mode of delivery Vaginal delivery Reference Reference VE 1.36 (0.97-1.91) 1.10 (0.72-1.69) Elective CS 1.34 (0.94-1.92) 1.19 (0.73-1.92) Emergency CS 1.45 (1.04-2.01) 0.76 (0.48-1.21) Parity 1.09 (0.84-1.42) Self-reported history of depression 2.68 (2.09-3.45) Fear of delivery 1.82 (1.39-2.38) Negative delivery experience 2.46 (1.67-3.63) Objective postpartum complications 1.08 (0.79-1.49) Self-reported physical symptoms postpartum (6 w pp) 1.49 (1.15-1.92) Adjusted model includes possible confounders (parity, self-reported history of depression, fear of delivery) and mediators (negative delivery experience, objective postpartum complications, and self-reported physical symptoms postpartum) as independent variables

In the path analysis, elective CS was not associated with PPDS, compared to vaginal delivery (fig 8). Delivery fear and history of depression were associ-ated both with elective CS and the odds for PPDS at 6 weeks postpartum, thus acting as important confounders in the examined association. Elective CS was further associated with lower odds of negative delivery experience.

45

Figure 8. Graphic display of the estimated path analysis models of elective caesar-ean section (CS) vs vaginal delivery (VaD), reflecting the direct and indirect effects on postpartum depressive symptoms (PPDS) at 6 weeks postpartum.

The path coefficients represent the estimated odds ratios from the corresponding logistic re-gression models. Pathways indicated with a continuous arrow were statistically significant (p<0.05). Dotted arrows represent pathways with borderline significance (p 0.05-0.10). Red arrow represents a negative association.

When repeating the path analysis with emergency CS and VE as exposure, a strikingly different pattern appeared (fig 9). Both emergency CS and VE, as compared to vaginal delivery, increased 7-fold and 4-fold, respectively, the odds for negative delivery experience, which in turn increased the odds for PPDS. Other pathways through which emergency CS increased odds for PPDS included objective postpartum complications and self-reported postpar-tum physical symptoms, which also acted as mediators. Unlike elective CS, there was no association of VE and elective CS with fear of delivery or history of depression.

46

Figure 9. Graphic display of the estimated path analysis models of emergency cae-sarean section (CS) and vacuum extraction (VE) vs vaginal delivery (VaD), reflecting the direct and indirect effects on postpartum depressive symptoms (PPDS) at 6 weeks postpartum.

The path coefficients represent the estimated odds ratios from the corresponding logistic re-gression models. Pathways indicated with a continuous arrow were statistically significant (p<0.05). Dotted arrows represent pathways with borderline significance (p 0.05-0.10).

Study III 1503 nulliparas were included in study III, of whom 800 (53%) reported use of EDA during delivery.

In the crude analysis, use of EDA was associated with higher odds of PPDS at 6 weeks postpartum (OR 1.52, 95% CI 1.8-2.08). However, this association dissipated after adjusting for a set of a priori determined confounders includ-ing age, fear of delivery, and antenatal depressive symptoms, as well as when further adjusting for the mediators negative delivery experience and operative delivery (i.e. VE or emergency CS). In the fully adjusted model, significant associations with PPDS at 6 weeks postpartum were noted for fear of delivery (OR 2.29, 95 % CI 1.57-3.32), antenatal depressive symptoms (OR 7.36, 95% CI 4.80-11.26) and negative delivery experience (OR 2.51,95% CI 1.51- 4.18) (table 4).

47

Table 4. Univariate and multivariate logistic regression derived Odds Ratios (ORs) and 95% Confidence Intervals (CIs) for postpartum depressive symptoms (PPDS) at 6 weeks postpartum

Unadjusted Model 1* Model 2** OR (95% CI) aOR (95% CI) aOR (95% CI) Epidural analgesia 1.52 (1.18-2.08) 1.22 (0.87-1.72) 1.11 (0.76-1.62) Age ≥35 years 1.11 (0.75-1.74) 1.35 (0.82-2.20) 1.27 (0.75-2.15) Fear of delivery 2.76 (1.99-3.81) 2.12 (1.49-3.01) 2.29 (1.57-3.32) Antenatal depressive symptoms

8.81 (6.05-12.83) 7.36 (4.99-10.85) 7.36 (4.80-11.26)

Negative delivery experi-ence

3.02 (1.98-4.61) 2.51 (1.51-4.18)

Operative delivery 1.36 (0.98-1.91) 1.15 (0.73-1.70) *Model 1: Includes EDA and possible confounders (age, fear of delivery, antenatal depressive symptoms defined as EPDS ≥13 pregnancy week 32) as independent variables **Model 2: Adjusted for Model 1 + possible mediators (delivery experience, operative deliv-ery, i.e. vacuum extraction or emergency cesarean section) as independent variables

Study IV A total of 1969 woman were included in the study. Delivery started spontane-ously for 78% and ended with non-instrumental vaginal delivery for 71% of the participants. Young age, lower educational level, underweight, over-weight, smoking during pregnancy, and psychiatric morbidity were associated with neuroticism.

In this study, women with higher level of neuroticism (i.e. a 63-unit in-crease, corresponding to the IQR) did not have higher odds for adverse peri-natal outcomes except for a significant crude association with GDM (OR 2.53, 95% CI 1.20-5.33) (table 5). This association was enhanced after adjustment for the sociodemographic factors maternal age, educational level, height, BMI, smoking and year of delivery (OR 3.57, 95% CI 1.32-9.65). The effect estimate was similar even after further adjustment for psychiatric morbidity but was no longer statistically significant.

48

Table 5. Logistic regression-derived odds ratios (ORs) with 95% confidence intervals (CIs) for perinatal outcomes, by an increase of 63 units in neuroticism score in SSP (equaling the interquartile range)

Unadjusted OR(95% CI)

Model 1 aOR(95% CI)

Model 2 aOR(95%CI)

Vaginal delivery, non-instrumental 1.09 (0.95-1.26) 1.08 (0.92-1.28) 1.07 (0.91-1.26) Vaginal delivery, vacuum extrac-tion 0.79 (0.65-0.95) 0.82 (0.66-1.01) 0.85 (0.68-1.05)

Any caesarean section 1.06 (0.89-1.26) 1.06 (0.87-1.30) 1.05 (0.85-1.28)

Elective cesarean section 1.03 (0.76-1.41) 0.97 (0.68-1.39) 0.94 (0.65-1.34) Emergency caesarean section 1.07 (0.88-1.29) 1.09 (0.87-1.38) 1.09 (0.87-1.38)

Gestational diabetes mellitus 2.53 (1.20-5.33) 3.57 (1.32-9.65) 2.70 (0.92-7.95) Gestational hypertension or Preeclampsia 1.07 (0.82-1.39) 1.13 (0.85-1.50) 1.12 (0.84-1.50)

Induction of delivery 1.04 (0.88-1.23) 1.06 (0.88-1.28) 1.01 (0.84-1.22)

Dystocia 0.95 (0.82-1.11) 0.99 (0.84-1.17) 0.99 (0.83-1.17) Severe lacerations 0.91 (0.71-1.17) 0.88 (0.67-1.17) 0.92 (0.70-1.22)

Placental retention 0.66 (0.44-0.99) 0.70 (0.45-1.09) 0.71 (0.46-1.11) Postpartum haemorrhage 0.95 (0.74-1.23) 0.97 (0.73-1.29) 0.94 (0.70-1.26)

Premature birth < 37 w a 0.87 (0.65-1.16) 0.95 (0.69-1.33) 0.96 (0.69-1.35) SGA <10th percentile a, 1.12 (0.87-1.45) 1.15 (0.86-1.54) 1.11 (0.82-1.49)

LGA >90th percentile a, 1.14 (0.90-1.45) 1.07 (0.82-1.41) 1.06 (0.80-1.39) Apgar 5 min < 7 0.77 (0.42-1.41) 0.92 (0.48-1.77) 0.87 (0.45-1.70)

Composite worst-case variable b 0.86 (0.57-1.30) 0.86 (0.54-1.38) 0.83 (0.51-1.34) SGA: Small for gestational age (according to Swedish sex-specific reference curves) LGA: Large for gestational age (According to Swedish sex-specific reference curves) *Model 1 adjusted for maternal age at delivery, educational level, height, body-mass index at first antenatal care visit, smoking at first antenatal care visit and/or at gestational week 32, year of delivery. * Model 2 adjusted for Model 1 + psychiatric morbidity a Excluding intrauterine demise (n=3). b intrauterine demise, eclampsia, severe PE, premature birth < 32 w, SGA below -2.5 SD (0.6%), placental abruption

Imputations for missing values for BMI, maternal height, and smoking during pregnancy, yielded similar results except that the association between neurot-icism and GDM remained significant after adjustment for psychiatric morbid-ity.

49

Discussion

Main Findings The current thesis elucidates the complex interplay between perinatal compli-cations and perinatal mental health, with a special focus on PPDS at 6 weeks postpartum and neuroticism.

After adjustment for relevant confounders, no direct association was iden-tified between the examined perinatal complications and PPDS at 6 weeks postpartum. However, this thesis identified a number of indirect effects through various mediators, thus allowing the delineation of the mechanisms connecting perinatal complications with PPD. In particular, the studies sug-gest that PPH increases the odds for PPDS via causing anaemia (study I), whereas an emergency setting delivery (VE or emergency CS) indirectly in-creases the risk of PPDS by leading to postpartum obstetrical complications, self-reported postpartum physical symptoms, and a negative delivery experi-ence. On the other hand, history of depression or other psychiatric disease seem to act as confounders in the associations between elective CS and PPD, EDA and PPD, and even in the association between neuroticism and GDM. These findings can be explained by psychological (study I-IV) and pathophys-iological (study I, IV) mechanisms.

Perinatal complications like PPH and operative delivery may cause psy-chological stress and even lead to psychological trauma, especially if women have been afraid for her own and/or the baby´s life (55, 79, 81). Additionally, during an acute event, when the staff at the delivery ward must act fast, com-munication gaps might emerge between the healthcare staff and the woman or her partner (282), which could further negatively influence the experience of delivery. Moreover, recent studies have shown that the mismatch between ex-pectations or plans for the mode of delivery or pain relief method on the one side, and the actual delivery process on the other, might be an important con-tributor to delivery experience (175, 283). Thereby, negative delivery experi-ence, a well-known risk factor for PPD may act as a key mediator of the effect of the perinatal stressors on PPD (83, 284).

Several biological mechanisms integrating stress regulation and pregnancy physiology have been suggested for explaining the association between peri-natal complications and PPD. A recurring hypothesis includes the HPA-axis (190, 239), which may mediate the identified association of neuroticism with GDM (study IV). It may even be relevant in studies I-III, where a disruption

50

of the HPA-axis might occur as a result of the stress caused by complications coupled with negative delivery experience. Another pathophysiological mech-anism highlighted in this thesis is anaemia, as an important mediator in the association between PPH and PPD.

Study I We found that PPH increases risk of PPD by often leading to postpartum anae-mia and to a negative delivery experience. Postpartum distress symptoms (285) and a negative experience of delivery (286) are known psychological consequences of PPH and have also been associated with PPD. A somatic consequence of severe bleeding is anaemia, which presents with a sympto-matology including fatigue, reduced cognitive abilities, emotional instability, and subsequently depression (119, 287, 288). Both anaemia and traumatic ex-periences increase the risk for depression in general and should therefore be considered in studies assessing the association between PPH and PPD.

Study II The path analysis verified no direct link between mode of delivery and PPDS at 6 weeks postpartum but indicated several indirect associations. Interest-ingly, two distinct patterns emerged for different modes of delivery. Women delivering by elective CS were not generally at increased risk of PPDS, if one considers that their burdened psychiatric history already predisposed them to depression. History of depression, an established strong predictor of PPD (289), has been associated with fear of delivery (290), which affects prefer-ence of mode of delivery mode and could lead to choice of elective CS (68), as verified in the path analysis.

The second pattern includes women who delivered by emergency CS or VE. These women neither had a psychosocial background predisposing to de-pression nor self-reported fear of delivery. However, the emergency setting of their delivery and the accompanying obstetric and neonatal complications in-creased the risk for PPDS (291, 292). Emergency CS is known to be associated with more complications and a more prolonged recovery time than vaginal delivery and elective CS (292-295). Women delivering by emergency CS may suffer both the physical fatigue of the delivery, as well as the psychological burden of the subjectively experienced inability to give birth vaginally (61, 282, 296, 297). Nevertheless, there are also qualitative studies where women also express gratefulness and relief (295, 298). Similarly to emergency CS, VE is also characterized by its emergency nature and similar consequences (299). All those parameters contribute to a negative delivery experience that could in itself trigger depressive symptomatology (83, 291).

51

Study III There was no association between use of EDA during delivery and PPDS at 6 weeks postpartum after adjustments for antenatal depression and fear of de-livery, both associated with low delivery self-efficacy, i.e. the woman´s own belief of act and master over the delivery process (300, 301). Women with low self-efficacy are more likely to choose EDA. Furthermore, antenatal depres-sion is associated with low expectations and disappointment with delivery (300). Together, those factors could partially explain the interplay between antenatal depression, fear of delivery, EDA and PPDS.

Another contributing factor to the crude association between EDA and PPD could be the fact that the pain experience of delivery differs from other forms of extreme pain. Specifically, many women have a positive outlook on and a pragmatic acceptance of delivery pain (143), and might even express a wish to have a “natural delivery” without pharmacological analgesia (302). Never-theless, there might be a mismatch between the woman´s expectation of pain and pain relief and the actual pain experience (154, 175) with the majority of nulliparas being administered EDA after all. This paradox and potential dis-appointment could partially explain the higher incidence of PPDS in women reporting use of EDA during delivery (175).

Study IV No associations between high levels of neuroticism and a range of perinatal outcomes were found except for an association between high levels of neurot-icism and GDM. However, this association became statistically non-signifi-cant after adjustment for psychiatric comorbidity. A possible explanation is that high stress levels cause an overactivation of the HPA-axis, which in turn induces insulin resistance and thus the development of GDM (42). Moreover, GDM has a strong association with obesity (31). In our cohort as well as in others (47, 48), there was a positive correlation between neuroticism and obe-sity. Therefore, large weight gain during pregnancy, which was not adjusted for in the analyses, could mediate the association of neuroticism with GDM.

Methodological considerations Strengths A principal strength of all included studies is the longitudinal prospective de-sign and the relatively large sample sizes, which offer sufficient statistical power in order to extract robust conclusions. The prospective design reduces the risk of recall bias. The detailed questionnaires at multiple timepoints ena-bled us to obtain detailed information on many sociodemographic, psychiatric,

52

and obstetric variables at the individual level. This made it possible to exhaust-ively examine the effects of potential confounders and mediators.

The large sample sizes further enabled the use of more sophisticated statis-tical approaches, like path analysis, which are particularly designed to eluci-date the complex interplays between several variables, as was the case in two of our research questions (study I-II). These studies are among the first to use such an approach to explore the complex association of delivery complications with PPDS.

Additionally, the detailed classification of mode of delivery (study II), which was not commonly applied in previous studies, enabled us to feature the differences between elective CS and emergency CS and also highlighted the similar psychological consequences of emergency CS and VE.

The study on use of EDA during delivery (study III) is one of the largest studies in the field and has also adjusted for a number of important covariates, something that had been explicitly indicated to be missing from the literature (303, 304). The covariates were selected by conceptualising an interaction framework and analysing it as a directed acyclic graph (see appendices 3 and 4), which reduces the risk of choosing non-appropriate covariates for adjust-ment (305).

Study IV is one of the few and largest studies on neuroticism and perinatal complications. By using a combination of self-report personality measures and by extracting data for perinatal complications from the national register, we reduced the risk of recall bias and secured high internal validity.

Limitations General limitations Both the BASIC and UPPSAT projects are population-based cohorts where nearly all pregnant women (BASIC) or new mothers (UPPSAT) in Uppsala County were invited to participate. However, the participation rate in BASIC was only 21%, possibly because of the comprehensive nature of the study, including multiple questionnaires and extended biological sampling. Thus, the question arises whether this sample is representative. Analyses have shown that there is an overrepresentation of first-time, highly educated mothers in the study. If the decision to participate was not clearly related to exposures and outcomes in this project, the lower participation rate is not expected to greatly influence associations reported in the current studies but may limit the gener-alizability of the findings. On the other hand, the two other dominant studies, the UPPSAT-project (study I) and the project on abortion (study IV), had rel-atively high participation rates, (56% and 58%, respectively), when compared to other cohort-studies (306).

In both the BASIC and the UPPSAT-projects, women were beforehand in-formed about the focus of the studies on perinatal mental health, which may

53

also have led to selection bias. For example, women with an own history of mental illness, might be more motivated to participate, thus introducing a form of self-selection bias (306). On the other side, depressive symptoms (307) and neuroticism (308) have been previously shown to be more common among non-participants in cohort studies. Also, women with a psychiatric history may decline participation because of their integrity and reluctance towards exten-sive biological sampling and questionnaires. Nevertheless, it is important to note that the prevalence of women reporting PPDS in these studies does not greatly differ from other Swedish studies (309, 310). No analyses on the prev-alence of PPDS or neuroticism could be performed in any of the included studies, as we did not have data on either non-participants or non-responders.

The sociodemographic distribution of participants in all of the contributing studies was comparable to the general Swedish statistics of pregnant women (the Swedish Birth Register), except for a slightly higher age, and an overrepresentation of primiparas and women of higher educational level (306). Due to the exclusion of women unable to communicate in Swedish, there was also an underrepresentation of women not born in Sweden.

Furthermore, PPD was not diagnosed by a clinical interview, but by a screening questionnaire, EPDS, which may introduce information bias and a possible overestimation of the prevalence of PPD. However, EPDS has high sensitivity and specificity (186). It is also the most established instrument in clinical studies, thereby allowing for comparison between studies if using the recommended cut-offs, as was the case in the current studies (311).

Similarly, history of depression was assessed by self-report and despite the high reliability reported in previous studies (312, 313), recall bias cannot be ruled out. Also, in studies with focus on mental health, some degree of re-sponse bias can be anticipated.

Furthermore, in order to shorten the already extensive questionnaires, fear of delivery and delivery experience were assessed by single questions and not validated instruments, which could raise concerns about possible misclassifi-cation (74-76, 171). In the same context, the dichotomization of delivery ex-perience assessed by a four-scale score might not be fully justifiable (314).

Despite the detailed questionnaires enabling adjusting for numerous co-variates, residual confounding due to unknown factors cannot be ruled out. Finally, it is worth noting the substantial proportion of missing answers in some items. Some of these items may be perceived as sensitive information or as having moral connotations, for example smoking or experience of domestic violence, which might introduce response bias.

Specific limitations of individual studies In study I, women were categorized in exposed and unexposed for PPH which might entail some risk for misclassification bias. To reduce that risk, all women with an estimated bleeding between 649 and 999 mL were excluded. The amount of bleeding was in most cases assessed by visual estimation,

54

which often leads to underestimation of blood loss (98, 99). From clinical ex-perience, one may also presume that in situations where bleeding approaches 1000 mL, there is a tendency to report a bleeding below 1000 mL, in a, more or less, conscious attempt to normalize the delivery. Thereby, the potential association of exposure and outcome would be diluted. Another issue to be considered is the creation of an algorithm to reduce the proportion of missing data on anaemia at discharge. Nevertheless, the algorithm was strict and thus entailed low risk of misclassification.

The observational design in study III implicates that the use of EDA was subject to the preferences of women per se and may have thus been affected by personal characteristics as depression, fear of delivery (167), socioeco-nomic status (169, 170), and wish of a pain-free delivery (315). However, by designing a conceptual directed acyclic graph (305), we adjusted for all the necessary confounders, thus minimizing the risk of confounding by indica-tion.

Even though study IV is the largest study in the field, the analyses for some outcomes like GDM, were still underpowered. Finally, the evaluation of out-comes through registry data and not through active assessment might intro-duce misclassification due to underreporting of some diagnoses in the regis-ters.

Clinical implications of the results The current studies indicate that perinatal complications are not directly asso-ciated with PPDS, but that history of depression in combination with negative delivery experience and somatic complications may confound or mediate an association, especially in the light of earlier psychiatric history, an important risk factor that should be assessed in the clinical setting.

The results from the first study indicate that postpartum anaemia and neg-ative delivery experience increase the risk for PPDS. This in turn points to the importance for carefully monitoring and considering active treatment after heavy postpartum bleeding to prevent anaemia in the prolonged postpartum period. Timely interventions aiming at preventing an unresolved negative self-reported delivery experience should also be considered. These two measures are expected to decrease odds for PPD.

Women undergoing elective CS are, as a group, more vulnerable, with a burdened psychiatric history and more frequently reporting fear of delivery. By beforehand deciding to undergo a CS, these women have a positive deliv-ery experience that possibly protects them against PPDS at 6 weeks postpar-tum. This implies that, in counselling women with fear of vaginal delivery, the clinician must carefully consider and weight the individual woman´s risk for psychological complications such as PPD in the case of vaginal delivery versus the risk of somatic complications in the case of elective CS. However,

55

it is also important to note that some women with fear of delivery may process their fear of delivery during counselling and may thereby be prepared for a vaginal delivery and be strengthened of that effort (50). Thereby, every con-sultation and decision based on an individual woman’s request for elective CS is an intricate task where the individual caregiver’s capability to tune into the woman’s balance of needs, fears and resilience level is of central importance (71).

On the contrary, emergency CS and VE more often lead to a negative de-livery experience and postpartum physical complications that predispose to PPD. This pinpoints the need to focus on the prevention of unresolved nega-tive delivery experience among women with an emergency setting delivery, as well as the importance of promptly attending to physical complications.

In contrast to our hypothesis, the use of EDA did not impact the risk for PPDS when adjusting for relevant sociodemographic, psychosocial, and ob-stetrical variables. This should be taken into account by midwives, obstetri-cians and the delivering couple, who are advising women on the use of epi-dural analgesia.

It is also important to note that no statistically significant associations were found between neuroticism and adverse perinatal outcomes among nullipa-rous women. This is valuable information to caregivers. The association be-tween neuroticism and GDM was possibly compromised in the final adjusted model by power issues and thus no clinical implications can be drawn before new replication studies are conducted.

56

Summary

In the first study (n=446), we explored the association between PPH and PPDS at 6 weeks postpartum, addressing the role of postpartum anaemia, delivery experience, and psychiatric history. There was no direct and independent as-sociation between PPH and PPDS. However, anaemia at discharge from the maternity ward more than doubled the odds for PPD symptoms and mediated the association between PPH and PPD. Other principal factors in the associa-tion between delivery complications and PPDS were negative delivery expe-rience and depressed mood during pregnancy.

The second study (n=3888) delineated the association of mode of delivery with PPDS at 6 weeks postpartum. The results indicate that there is no inde-pendent association between elective CS and risk of PPD, but the association is highly confounded by history of depression and fear of delivery. On the other hand, emergency CS and VE increase odds for PPD through the media-tors postpartum complications and negative delivery experience.

The third study (n=1503) investigated the association between the use of EDA during delivery and PPDS. A positive association in the crude analysis was not apparent after adjustment for sociodemographic, psychosocial, and obstetrical variables, indicating that pain relief through epidural analgesia is not likely to affect the risk of PPDS.

In the last study (n=1969), the association of neuroticism and perinatal complications was explored. Neuroticism was not associated with adverse perinatal outcomes, except for an association with higher odds for GDM which, however, became statistically non-significant after adjusting for psy-chiatric morbidity.

57

Conclusion

The aim of the current thesis was to explore the complex association between perinatal complications and perinatal maternal health with focus on PPD and neuroticism. The debate on the potential association of perinatal complica-tions and PPD has been on-going for a long time, but studies on personality traits and perinatal complications are sparse. These studies integrating somatic and psychological aspects, give us insight in possible mediating steps in these processes, which we can target by preventive efforts.

In the current studies, special emphasis has been given to delineate the ex-amined associations by considering important covariates as history of depres-sion, antenatal depression, and delivery experience. The results indicate no direct association of the delivery complications under investigation, and PPDS or neuroticism. However, there are indirect associations through modifiable and non-modifiable mediators. An important modifiable mediator seems to be anaemia postpartum. In today’s praxis, there is no clear guideline on which women to test for anaemia before discharge after delivery. The first and only follow up of Hb-levels postpartum takes place at the maternal primary care setting at 6-12 weeks postpartum. The results of the current thesis highlight that it is important to target postpartum anaemia before discharge from the maternity ward in order to prevent PPD. In parallel, women with psychologi-cal vulnerability or a negative delivery experience also need to be identified, as these variables were also identified as confounder and mediator, respec-tively, in the association between PPH and PPD. Those women are at higher risk to develop PPD and thereby need extra psychological care and support in the early postpartum period to prevent the development of PPD, especially after a complicated delivery. Collecting information on earlier history of de-pression or other psychiatric disorders as part of the obstetric care would guide the selection of women at high risk for PPD particularly in cases when peri-natal complications occurred.

58

Future studies

The first study that mentioned depression following delivery was published in 1968, but it was not before the 1980s that the research-field begun to flourish. The development of the EPDS questionnaire, focused on perinatal depression, was an important contribution towards this direction. Since then, the field has widely expanded in a broad spectrum of perspectives, including identification of sociological, psychological, and biological risk factors, qualitative and quantitative research approaches, and development of methods for screening and treatment in the clinics. But still, over 12,000 women and their families are affected by PPD in Sweden every year, with no decreasing trend. Many of these women suffer in silence due to the perceived stigma of getting depressed when having a newborn, whereas others misinterpret their symptoms as of somatic origin (201, 218).

Currently, there is good evidence for several major risk factors for PPD, but there are still numerous potential risk factors, for which the association with PPD remains ambiguous. More advanced methods including path analy-sis and multifactorial approaches are expected to elucidate these associations.

In this thesis I have tried to delineate the association of PPH, mode of de-livery, and EDA with PPD, but there are still many other common perinatal factors and complications that might be interesting to investigate, for example severe lacerations and postpartum infection. During the last years, lacerations and their potential sequels have been spot-lighted as leading major concerns for pregnant women. This is something that midwifes and doctors in the de-livery ward and antenatal care face every day when they meet women with worries about the forthcoming delivery and possible laceration. Thereby, it would also be interesting to compare the rates of PPD in women with severe lacerations in two to three different time-periods, for example 2006-2007 (UPPSAT), 2009-2014 (BASIC) and 2015-2018 (BASIC).

Risk factors for PPD differ between different contexts, especially between low- and high-risk countries. Therefore, it would be interesting to examine in future studies whether the results presented in this thesis could be replicated in low-income countries.

Study I implies that treatment of postpartum anaemia may decrease the risk of PPD. The next step could thereby be a randomised controlled trial (RCT) on different treatment strategies for anaemia, comparing current practice with a more vigorous treatment of postpartum anaemia.

59

In contrast to the literature on PPD, studies on personality and pregnancy are scarce. The present study reports a crude association with GDM, which is however no longer statistically significant after adjusting for sociodemo-graphic factors and psychiatric morbidity. This association has to be explored further due to the small number of diabetic cases in the current studies limiting the statistical power.

At the moment, the BASIC project includes more than 4000 women as-sessed with the psychometric instrument Vulnerable Personality Style Ques-tionnaire (VPSQ) (316), which would allow for a replication study. Future studies should also investigate the association between perinatal outcomes and selected subscales of neuroticism, as different factors may mediate different outcomes (239).

Moreover, a subsequent study arising from the results of this thesis would be to incorporate perinatal complications in a path analysis on the association between neuroticism and PPD.

60

Sammanfattning på svenska

Att bli förälder är en överväldigande upplevelse. Graviditet, förlossning och barnsängstid innebär såväl stora kroppsliga förändringar som psykologiska och existentiella förändringar. Trots att det stora flertalet av graviditeter och förlossningar i västvärlden förlöper väl, mycket tack vare den omfattande mödrahälsovården, så drabbas femton procent av de västerländska kvinnorna av graviditets- och förlossningsrelaterade komplikationer. Syftet med denna avhandling var att utforska och beskriva de komplexa associationerna mellan perinatala komplikationer och psykiska ohälsa, med fokus på förlossningsde-pressioner och neuroticism. Neuroticism är ett personlighetsdrag som känne-tecknas av ökad ängslighet och en tendens att oftare uppleva obehagliga käns-lor som ångest, oro och rädsla. Speciell vikt har lagts vid att utforska viktiga bidragande faktorer, såsom tidigare depression, depression under graviditeten och förlossningsupplevelse.

Den första studien utforskade huruvida kvinnor som haft en stor förloss-ningsblödning oftare drabbades av förlossningsdepression. Sammanlagt 446 kvinnors enkätsvar och förlossningsjournaler granskades. En speciell sorts analytisk metod där det går att få en uppfattning om hur olika faktorer påver-kar varandra i flera steg användes, en så kallad steg-analys. På det sättet kunde vi få en inblick i vilka konsekvenser en stor förlossningsblödning har och hur den skulle kunna påverka risken att utveckla en depression. Sammanfattnings-vis ökade inte en stor förlossningsblödning i sig risken att drabbas av en för-lossningsdepression. Däremot talar resultaten för att det är en ökad risk för förlossningsdepression om kvinnan haft en stor förlossningsblödning och om hon dessutom haft blodbrist vid utskrivningen från BB eller haft en dålig för-lossningsupplevelse.

I den andra studien undersökte vi om det fanns något samband mellan för-lossningssätt och förlossningsdepression. Även här använde vi oss av steg-analys. 3888 förlossningar följdes upp 6 veckor efter förlossningen. Vi kunde inte identifiera någon direkt association mellan förlossningssätt och förloss-ningsdepression. Däremot utkristalliserade sig flera faktorer som bidrog till en indirekt association mellan förlossningssätt och förlossningsdepression. Kvin-nor vars förlossningar avslutats med ett akut kejsarsnitt eller med sugklocka hade, jämfört med kvinnor med spontan vaginal förlossning, en ökad risk att drabbas av förlossningsdepression ifall de även hade en negativ förlossnings-upplevelse, drabbades av förlossningskomplikationer eller rapporterade att de hade fysiska besvär efter förlossningen. Även kvinnor med planerat kejsarsnitt

61

fick oftare förlossningsdepression, men detta kunde helt förklaras med att de oftare var förlossningsrädda och tidigare i livet haft en depression.

I den tredje studien undersökte vi om ryggbedövning under förlossningen minskade risken att drabbas av förlossningsdepression. 1503 förstföderskor deltog i studien. Vi fann inga belägg för att ryggbedövning i sig minskar risken för att drabbas av en förlossningsdepression.

I den sista studien undersökte vi om neuroticism ökade risken för gravidi-tets- och förlossningskomplikationer samt komplikationer hos den nyfödde. I de tidigare studierna använde vi oss av enkäter till kvinnorna i kombination med journalgranskning. I den här studien utgick vi från kvinnornas svar på ett frågeformulär om personlighet och kopplade resultaten till nationella register istället. 1969 kvinnor från flera tidigare insamlade mindre studier deltog. Inga associationer mellan neuroticism och komplikationer hittades bortsett från ett visst samband mellan neuroticism och graviditetsdiabetes. Detta samband kvarstod dock inte efter att vi tagit hänsyn till annan psykisk ohälsa.

Sammanfattningsvis har avhandlingen belyst åtskilliga aspekter av sam-bandet mellan komplikationer under graviditet och förlossning och kvinnans psykiska ohälsa. Huruvida komplikationer är associerat med förlossningsde-pression har länge debatterats, men forskningen om neuroticism och kompli-kationer har varit mycket sparsam. I de aktuella studierna har vi inte funnit några belägg för att det finns ett direkt samband mellan graviditets- och för-lossningskomplikationer och psykisk hälsa. Däremot finns det andra viktiga faktorer som bidrar till en ökad risk att drabbas av en förlossningsdepression, däribland kvinnans förlossningsupplevelse. Likaså är det viktigt att beakta om kvinnan tidigare haft problem med psykisk ohälsa.

62

Acknowledgments

First of all, the very largest THANKS to my supervisor Alkistis Skalkidou who has guided and supported me, and sometimes even convinced me to continue, during this long journey. And thanks for your endless hospitality where you not only invite me to your home, but also my kids. And thanks for all delicious, and overwhelming dinners (as my daughter never will forget). Thanks to my co-supervisor Anna-Karin Wikström for all your support and substantial advises. And your always so fast responses on e-mails! Also, thanks for pushing me to prioritize the doctoral studies during the last year. And I will always remember chats in early mornings, running in the astonish-ing Greek surroundings. Ulf Högberg, my second co-supervisor, thanks for your always so thoughtful and wise opinions, from study-design to problematizing single words. Linnea Carlsson, my co-supervisor from the other side of the Baltic sea. It has been a pleasure to talk and communicate with you by e-mail, even but the focus of the thesis changed from the first plans. But your comments and ad-vices has always been fruitful. Masoumeh Rezapour Isfahani (Head of the Department of Obstetrics and Gy-naecology), Gunilla Hallberg (Head of the Obstetric Department), Karin Eurenius, Ajlana Mulic-Lutvica, Gudlaug Sverrisdottir (former Head of Res-ident Doctors). The fundament for all research is to have allocated time. You have been generous with giving me this time enabling me to accomplish my doctoral studies. And Gunilla, you have also given my so much support for both the doctoral studies and my clinical work but also to enable time for my family. This approach has been so important for me! Eva Bergman, my clinical supervisor during my resident years. ‘Thanks’ is a too small word for all your support. You still have a dinner waiting for you! Thanks to all coauthors; Marios Georgakis, Natasa Kollia, Agneta Skoog-Svanberg, Mia Ramklint, Lisa Ekselius, Charlotte Hellgren, Johanna Löfblad and Helena Volgsten, Ulf, Anna-Karin and Alkistis. Without you, the studies have still been in some sort of crude shape. And Cathrine Axfors, without you,

63

the paper on neuroticism would just still have been some sort of rudimentary core. Even when you´re coming directly from workout, you´re always so bril-liant. I´m so impressed of your sharpness and your athleticism Without statistic, no quantitative research. So, thanks to the statisticians Rich-ard White and Natasa Kollia for all your help with this tricky part. (even but I sometimes had a little bit hard to understand the Greek conversation…) Marios Georgakis, except co-authoring, you have helped me with language revisions and time-consuming submissions. Thanks! Hanna Henriksson, your gentleness with offering reading my half-time semi-nar and now my thesis is so much worth. And all your talks, sms and support. It´s payback-time soon! All other PhD-students and colleagues at the Department of Women´s and Children´s health, for a warmth, open and always supporting atmosphere. Thanks to all colleagues, physicians, midwifes, nurses and assistant nurses. Thanks for such good teamwork and that you contribute to that I always enjoy going to work Liljana Kočoska-Maraš, Tansim Akhter and Hildur Gudjonsdottir: you are my absolutely best room-mates. You have the key to both my heart and to my room. And Mariella Öberg, welcome to our room! Cecilia Johansson and Carolina Wikström. After half a year of sealing enve-lopes you came as angels of mercy and saved me from the Sisyphean task of sending information of the U-BIRTH study to all participants of the BASIC and UPPSAT projects. Jennie Bystedt, my dear, dear friend. Thanks to all long walks in winter and summer, rain, snow, blow and sun. Nothing can stop us on our walks, discuss-ing daily things as well as meaning with life. OK, you´re right, an approaching dissertation can temporary halt us… But soon, very soon…. Here we go again!

Teija Rantakeisu, my sunshine. It doesn´t matter how dull and concerned I am. After 10 minutes with you, I am full of energy and joy!

Karin Blomquist, my “aunt Hulda”. You are so important for me. Karin and Tilda Eckerdal, my goddaughter, without you, we´re not a whole

family. Thanks for all, all time I have spent with you. Mom and Dad. Thanks for always being there for me and my family. And

thanks for all red marks in the proof. Victoria, thanks for being my sister.

64

Julia, David and Nils. Even but you may have distrusted the last two months, you are the very, very most important for me. Now, after the disser-tation, I will play “Nya Finans” with you, follow you to the MTB-track, do excursions and picnics, start to play the violin again, go to town shopping. Just be with you and for you…. I love you!

Johan, my beloved husband, Jean le Vrai. During all our years together, we have experienced and shared so much, both joy and sadness. I know that I´m not always so easy to live with and sometimes hard to understand, espe-cially during the last months. I wouldn´t have managed this without your sup-port, love and sense of humour. Jag älskar dig så mycket!

65

References

1. Maternal and Newborn Health/Safe Motherhood Unit: Care in normal birth: a practical guide. Geneva: World Health Organization.

2. WHO recommendations: Intrapartum care for a positive childbirth experience. Geneva: World Health Organization, Research DoRHa; 2018. Report No.: 978-92-4-155021-5.

3. Vladic Stjernholm Y, Weichselbarun M, Amer-Wåhlin I. Riskbedömning ankomst förlossningsenhet. Råd för utfärdande av klinikrutiner/PM. Land-stingens Ömsesidiga Försäkringsbolag; 2015-10-25.

4. Herbst A, Amer-Wåhlin I, Vladic Stjernholm Y, Weichselbraun M, Domellöf M. Fosterövervakning vid aktiv förlossning. Råd för utfärdande av klinikru-tiner/PM. Landstingens Ömsesidiga Försäkringsbolag; 2015 2015-10-25.

5. Jones L, Othman M, Dowswell T, Alfirevic Z, Gates S, Newburn M, et al. Pain management for women in labour: an overview of systematic reviews. The Cochrane database of systematic reviews. 2012(3):Cd009234.

6. Anim-Somuah M, Smyth RM, Jones L. Epidural versus non-epidural or no analgesia in labour. The Cochrane database of systematic reviews. 2011(12):CD000331.

7. Graham W, Woodd S, Byass P, Filippi V, Gon G, Virgo S, et al. Diversity and divergence: the dynamic burden of poor maternal health. Lancet. 2016.

8. United Nations: Sustainable Development Goal 3. Ensure healty lifes and pro-mote well-being for all at all ages [Internet]. https://sustainabledevelop-ment.un.org/sdg3: United Nations; [

9. Campbell OM, Calvert C, Testa A, Strehlow M, Benova L, Keyes E, et al. The scale, scope, coverage, and capability of childbirth care. Lancet. 2016;388(10056):2193-208.

10. Graham W, Woodd S, Byass P, Filippi V, Gon G, Virgo S, et al. Diversity and divergence: the dynamic burden of poor maternal health. Lancet. 2016;388(10056):2164-75.

11. Koblinsky M, Moyer CA, Calvert C, Campbell J, Campbell OM, Feigl AB, et al. Quality maternity care for every woman, everywhere: a call to action. Lan-cet. 2016;388(10057):2307-20.

12. Kruk ME, Kujawski S, Moyer CA, Adanu RM, Afsana K, Cohen J, et al. Next generation maternal health: external shocks and health-system innovations. Lancet. 2016;388(10057):2296-306.

13. Miller S, Abalos E, Chamillard M, Ciapponi A, Colaci D, Comande D, et al. Beyond too little, too late and too much, too soon: a pathway towards evidence-based, respectful maternity care worldwide. Lancet. 2016;388(10056):2176-92.

14. Shaw D, Guise JM, Shah N, Gemzell-Danielsson K, Joseph KS, Levy B, et al. Drivers of maternity care in high-income countries: can health systems support woman-centred care? Lancet. 2016;388(10057):2282-95.

66

15. Chou D, Tuncalp O, Firoz T, Barreix M, Filippi V, von Dadelszen P, et al. Constructing maternal morbidity - towards a standard tool to measure and monitor maternal health beyond mortality. BMC Pregnancy Childbirth. 2016;16:45.

16. Firoz T, Chou D, von Dadelszen P, Agrawal P, Vanderkruik R, Tuncalp O, et al. Measuring maternal health: focus on maternal morbidity. Bull World Health Organ. 2013;91(10):794-6.

17. Gon G, Leite A, Calvert C, Woodd S, Graham WJ, Filippi V. The frequency of maternal morbidity: A systematic review of systematic reviews. Interna-tional journal of gynaecology and obstetrics: the official organ of the Interna-tional Federation of Gynaecology and Obstetrics. 2018;141 Suppl 1:20-38.

18. Bacak SJ, Callaghan WM, Dietz PM, Crouse C. Pregnancy-associated hospi-talizations in the United States, 1999-2000. Am J Obstet Gynecol. 2005;192(2):592-7.

19. Mutihir JT, Utoo BT. Postpartum maternal morbidity in Jos, North-Central Ni-geria. Niger J Clin Pract. 2011;14(1):38-42.

20. The Millennium Development Goals Report 2015. New York: United Nations, Nations U; 2015.

21. Doctor HV, Nkhana-Salimu S, Abdulsalam-Anibilowo M. Health facility de-livery in sub-Saharan Africa: successes, challenges, and implications for the 2030 development agenda. BMC public health. 2018;18(1):765.

22. Martis R, Crowther CA, Shepherd E, Alsweiler J, Downie MR, Brown J. Treat-ments for women with gestational diabetes mellitus: an overview of Cochrane systematic reviews. The Cochrane database of systematic reviews. 2018;8:Cd012327.

23. IDF Diabetes Atlas. 8th Edition. 8th ed. Brussel, Belgium: International Dia-betes Federation; 2017 2017. 147 p.

24. Schneider S, Bock C, Wetzel M, Maul H, Loerbroks A. The prevalence of ges-tational diabetes in advanced economies. Journal of perinatal medicine. 2012;40(5):511-20.

25. Hosseini E, Janghorbani M. Systematic review and meta-analysis of diagnos-ing gestational diabetes mellitus with one-step or two-step approaches and as-sociations with adverse pregnancy outcomes. International journal of gynae-cology and obstetrics: the official organ of the International Federation of Gy-naecology and Obstetrics. 2018.

26. Yuen L, Wong VW, Simmons D. Ethnic Disparities in Gestational Diabetes. Current diabetes reports. 2018;18(9):68.

27. Graviditetsregistret. Årsrapport 2017. Stockholm; 2017. 28. Sacks DA, Hadden DR, Maresh M, Deerochanawong C, Dyer AR, Metzger

BE, et al. Frequency of gestational diabetes mellitus at collaborating centers based on IADPSG consensus panel-recommended criteria: the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Diabetes Care. 2012;35(3):526-8.

29. Fadl HE, Simmons D. Trends in diabetes in pregnancy in Sweden 1998-2012. BMJ open diabetes research & care. 2016;4(1):e000221.

30. Ignell C, Claesson R, Anderberg E, Berntorp K. Trends in the prevalence of gestational diabetes mellitus in southern Sweden, 2003-2012. Acta Obstet Gy-necol Scand. 2014;93(4):420-4.

31. Lindqvist M, Persson M, Lindkvist M, Mogren I. No consensus on gestational diabetes mellitus screening regimes in Sweden: pregnancy outcomes in rela-tion to different screening regimes 2011 to 2012, a cross-sectional study. BMC Pregnancy Childbirth. 2014;14:185.

67

32. Mezuk B, Eaton WW, Albrecht S, Golden SH. Depression and type 2 diabetes over the lifespan: a meta-analysis. Diabetes Care. 2008;31(12):2383-90.

33. Pan A, Lucas M, Sun Q, van Dam RM, Franco OH, Manson JE, et al. Bidirec-tional association between depression and type 2 diabetes mellitus in women. Arch Intern Med. 2010;170(21):1884-91.

34. Roy T, Lloyd CE. Epidemiology of depression and diabetes: a systematic re-view. J Affect Disord. 2012;142 Suppl:S8-21.

35. Sumlin LL, Garcia TJ, Brown SA, Winter MA, Garcia AA, Brown A, et al. Depression and adherence to lifestyle changes in type 2 diabetes: a systematic review. The Diabetes educator. 2014;40(6):731-44.

36. Hinkle SN, Buck Louis GM, Rawal S, Zhu Y, Albert PS, Zhang C. A longitu-dinal study of depression and gestational diabetes in pregnancy and the post-partum period. Diabetologia. 2016;59(12):2594-602.

37. Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS. Ef-fect of treatment of gestational diabetes mellitus on pregnancy outcomes. The New England journal of medicine. 2005;352(24):2477-86.

38. Silverman ME, Reichenberg A, Savitz DA, Cnattingius S, Lichtenstein P, Hultman CM, et al. The risk factors for postpartum depression: A population-based study. Depress Anxiety. 2017;34(2):178-87.

39. Kozhimannil KB, Pereira MA, Harlow BL. Association between diabetes and perinatal depression among low-income mothers. Jama. 2009;301(8):842-7.

40. Meltzer-Brody S, Maegbaek ML, Medland SE, Miller WC, Sullivan P, Munk-Olsen T. Obstetrical, pregnancy and socio-economic predictors for new-onset severe postpartum psychiatric disorders in primiparous women. Psychol Med. 2017;47(8):1427-41.

41. Katon JG, Russo J, Gavin AR, Melville JL, Katon WJ. Diabetes and Depres-sion in Pregnancy: Is There an Association? Journal of Women's Health. 2011;20(7):983-9.

42. Joseph JJ, Golden SH. Cortisol dysregulation: the bidirectional link between stress, depression, and type 2 diabetes mellitus. Annals of the New York Acad-emy of Sciences. 2017;1391(1):20-34.

43. Osborne LM, Monk C. Perinatal depression--the fourth inflammatory morbid-ity of pregnancy?: Theory and literature review. Psychoneuroendocrinology. 2013;38(10):1929-52.

44. Mijatovic-Vukas J, Capling L, Cheng S, Stamatakis E, Louie J, Cheung NW, et al. Associations of Diet and Physical Activity with Risk for Gestational Di-abetes Mellitus: A Systematic Review and Meta-Analysis. Nutrients. 2018;10(6).

45. Shepherd E, Gomersall JC, Tieu J, Han S, Crowther CA, Middleton P. Com-bined diet and exercise interventions for preventing gestational diabetes melli-tus. The Cochrane database of systematic reviews. 2017;11:Cd010443.

46. Gois C, Akiskal H, Akiskal K, Figueira ML. The relationship between temper-ament, diabetes and depression. J Affect Disord. 2012;142 Suppl:S67-71.

47. Phillips AC, Batty GD, Weiss A, Deary I, Gale CR, Thomas GN, et al. Neu-roticism, cognitive ability, and the metabolic syndrome: The Vietnam Experi-ence Study. J Psychosom Res. 2010;69(2):193-201.

48. Gerlach G, Herpertz S, Loeber S. Personality traits and obesity: a systematic review. Obesity reviews : an official journal of the International Association for the Study of Obesity. 2015;16(1):32-63.

49. Milgrom J, Skouteris H, Worotniuk T, Henwood A, Bruce L. The association between ante- and postnatal depressive symptoms and obesity in both mother

68

and child: a systematic review of the literature. Women's health issues : official publication of the Jacobs Institute of Women's Health. 2012;22(3):e319-28.

50. Striebich S, Mattern E, Ayerle GM. Support for pregnant women identified with fear of childbirth (FOC)/tokophobia - A systematic review of approaches and interventions. Midwifery. 2018;61:97-115.

51. O'Connell MA, Leahy-Warren P, Khashan AS, Kenny LC, O'Neill SM. World-wide prevalence of tocophobia in pregnant women: systematic review and meta-analysis. Acta Obstet Gynecol Scand. 2017;96(8):907-20.

52. Nilsson C, Hessman E, Sjoblom H, Dencker A, Jangsten E, Mollberg M, et al. Definitions, measurements and prevalence of fear of childbirth: a systematic review. BMC Pregnancy Childbirth. 2018;18(1):28.

53. Kunskapsstöd för Mödravården Stockholm, Sweden: Socialstyrelsen (National Board of Health and Welfare); 2014. Report No.: 2014-2-2.

54. Larsson B, Karlstrom A, Rubertsson C, Hildingsson I. Counseling for child-birth fear - a national survey. Sexual & reproductive healthcare : official jour-nal of the Swedish Association of Midwives. 2016;8:82-7.

55. Bastos MH, Furuta M, Small R, McKenzie-McHarg K, Bick D. Debriefing interventions for the prevention of psychological trauma in women following childbirth. The Cochrane database of systematic reviews. 2015(4):Cd007194.

56. Ayers S, Bond R, Bertullies S, Wijma K. The aetiology of post-traumatic stress following childbirth: a meta-analysis and theoretical framework. Psychol Med. 2016;46(6):1121-34.

57. Sydsjo G, Angerbjorn L, Palmquist S, Bladh M, Sydsjo A, Josefsson A. Sec-ondary fear of childbirth prolongs the time to subsequent delivery. Acta Obstet Gynecol Scand. 2013;92(2):210-4.

58. Bahl R, Strachan B, Murphy DJ. Outcome of subsequent pregnancy three years after previous operative delivery in the second stage of labour: cohort study. BMJ. 2004;328(7435):311.

59. Bayrampour H, Ali E, McNeil DA, Benzies K, MacQueen G, Tough S. Preg-nancy-related anxiety: A concept analysis. International journal of nursing studies. 2016;55:115-30.

60. Adams SS, Eberhard-Gran M, Eskild A. Fear of childbirth and duration of la-bour: a study of 2206 women with intended vaginal delivery. Bjog. 2012;119(10):1238-46.

61. Karlstrom A. Women's self-reported experience of unplanned caesarean sec-tion: Results of a Swedish study. Midwifery. 2017;50:253-8.

62. Ryding EL, Lukasse M, Parys AS, Wangel AM, Karro H, Kristjansdottir H, et al. Fear of childbirth and risk of cesarean delivery: a cohort study in six Euro-pean countries. Birth. 2015;42(1):48-55.

63. Jespersen C, Hegaard HK, Schroll AM, Rosthoj S, Kjaergaard H. Fear of child-birth and emergency caesarean section in low-risk nulliparous women: a pro-spective cohort study. J Psychosom Obstet Gynaecol. 2014;35(4):109-15.

64. Fenwick J, Gamble J, Nathan E, Bayes S, Hauck Y. Pre- and postpartum levels of childbirth fear and the relationship to birth outcomes in a cohort of Austral-ian women. Journal of clinical nursing. 2009;18(5):667-77.

65. Sydsjo G, Sydsjo A, Gunnervik C, Bladh M, Josefsson A. Obstetric outcome for women who received individualized treatment for fear of childbirth during pregnancy. Acta Obstet Gynecol Scand. 2012;91(1):44-9.

66. Waldenstrom U, Hildingsson I, Ryding EL. Antenatal fear of childbirth and its association with subsequent caesarean section and experience of childbirth. BJOG. 2006;113(6):638-46.

69

67. Stutzer PP, Berlit S, Lis S, Schmahl C, Sutterlin M, Tuschy B. Elective Cae-sarean section on maternal request in Germany: factors affecting decision mak-ing concerning mode of delivery. Arch Gynecol Obstet. 2017;295(5):1151-6.

68. Storksen HT, Garthus-Niegel S, Adams SS, Vangen S, Eberhard-Gran M. Fear of childbirth and elective caesarean section: a population-based study. BMC Pregnancy Childbirth. 2015;15:221.

69. Nieminen K, Wijma K, Johansson S, Kinberger EK, Ryding EL, Andersson G, et al. Severe fear of childbirth indicates high perinatal costs for Swedish women giving birth to their first child. Acta Obstet Gynecol Scand. 2017;96(4):438-46.

70. Haines HM, Rubertsson C, Pallant JF, Hildingsson I. The influence of women's fear, attitudes and beliefs of childbirth on mode and experience of birth. BMC Pregnancy Childbirth. 2012;12:55.

71. Andolf EL, H, Hildingsson I, Wiklund I, Blomdahl M, Lundberg F, Månsson M, et al. Indikation för kejsarsnitt på moderns önskan. Stockholm: Samarbetet Nationella medicinska indikationer (Svenska läkaresällskapet, Sveriges Kom-muner och Landsting, Socialstyrelsen, SBU); 2011.

72. Wijma K, Wijma B, Zar M. Psychometric aspects of the W-DEQ; a new ques-tionnaire for the measurement of fear of childbirth. J Psychosom Obstet Gy-naecol. 1998;19(2):84-97.

73. Elvander C, Cnattingius S, Kjerulff KH. Birth experience in women with low, intermediate or high levels of fear: findings from the first baby study. Birth. 2013;40(4):289-96.

74. Geissbuehler V, Eberhard J. Fear of childbirth during pregnancy: a study of more than 8000 pregnant women. J Psychosom Obstet Gynaecol. 2002;23(4):229-35.

75. Laursen M, Johansen C, Hedegaard M. Fear of childbirth and risk for birth complications in nulliparous women in the Danish National Birth Cohort. BJOG. 2009;116(10):1350-5.

76. Nilsson C, Lundgren I, Karlstrom A, Hildingsson I. Self reported fear of child-birth and its association with women's birth experience and mode of delivery: a longitudinal population-based study. Women Birth. 2012;25(3):114-21.

77. Mathai M, Sanghvi H, Guidotti R, Broekhuizen F, Chalmers B, Johnson R, et al. Managing Complications in Pregnancy and Childbirth: A guide for mid-wives and doctors: World Health Organization WHO Library Cataloguing-in-Publication Data; 2007.

78. Olde E, van der Hart O, Kleber R, van Son M. Posttraumatic stress following childbirth: a review. Clinical psychology review. 2006;26(1):1-16.

79. Andersen LB, Melvaer LB, Videbech P, Lamont RF, Joergensen JS. Risk fac-tors for developing post-traumatic stress disorder following childbirth: a sys-tematic review. Acta Obstet Gynecol Scand. 2012;91(11):1261-72.

80. Vignato J, Georges JM, Bush RA, Connelly CD. Post-traumatic stress disorder in the perinatal period: A concept analysis. Journal of clinical nursing. 2017;26(23-24):3859-68.

81. Grekin R, O'Hara MW. Prevalence and risk factors of postpartum posttrau-matic stress disorder: a meta-analysis. Clinical psychology review. 2014;34(5):389-401.

82. Robertson E, Grace S, Wallington T, Stewart DE. Antenatal risk factors for postpartum depression: a synthesis of recent literature. Gen Hosp Psychiatry. 2004;26(4):289-95.

83. Bell AF, Andersson E. The birth experience and women's postnatal depression: A systematic review. Midwifery. 2016;39:112-23.

70

84. Johnstone SJ, Boyce PM, Hickey AR, Morris-Yatees AD, Harris MG. Obstet-ric risk factors for postnatal depression in urban and rural community samples. The Australian and New Zealand journal of psychiatry. 2001;35(1):69-74.

85. Sword W, Landy CK, Thabane L, Watt S, Krueger P, Farine D, et al. Is mode of delivery associated with postpartum depression at 6 weeks: a prospective cohort study. BJOG. 2011;118(8):966-77.

86. Kassebaum NJ, Bertozzi-Villa A, Coggeshall MS, Shackelford KA, Steiner C, Heuton KR, et al. Global, regional, and national levels and causes of maternal mortality during 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014;384(9947):980-1004.

87. Calvert C, Thomas SL, Ronsmans C, Wagner KS, Adler AJ, Filippi V. Identi-fying regional variation in the prevalence of postpartum haemorrhage: a sys-tematic review and meta-analysis. PLoS One. 2012;7(7):e41114.

88. Carroli G, Cuesta C, Abalos E, Gulmezoglu AM. Epidemiology of postpartum haemorrhage: a systematic review. Best Pract Res Clin Obstet Gynaecol. 2008;22(6):999-1012.

89. Chandraharan E, Krishna A. Diagnosis and management of postpartum haem-orrhage. BMJ. 2017;358:j3875.

90. Francois KE FM. Antepartum and postpartum hemorrhage. 5th ed. Philadel-phia, PA: Churchill Livingstone Elsevier,; 2007. 1391 p.

91. Thompson JF, Roberts CL, Ellwood DA. Emotional and physical health out-comes after significant primary post-partum haemorrhage (PPH): a multicentre cohort study. The Australian & New Zealand journal of obstetrics & gynaecol-ogy. 2011;51(4):365-71.

92. Bryant A, Mhyre JM, Leffert LR, Hoban RA, Yakoob MY, Bateman BT. The association of maternal race and ethnicity and the risk of postpartum hemor-rhage. Anesthesia and analgesia. 2012;115(5):1127-36.

93. Al-Zirqi I, Vangen S, Forsen L, Stray-Pedersen B. Prevalence and risk factors of severe obstetric haemorrhage. BJOG : an international journal of obstetrics and gynaecology. 2008;115(10):1265-72.

94. Mousa HA, Cording V, Alfirevic Z. Risk factors and interventions associated with major primary postpartum hemorrhage unresponsive to first-line conven-tional therapy. Acta obstetricia et gynecologica Scandinavica. 2008;87(6):652-61.

95. Kramer MS, Berg C, Abenhaim H, Dahhou M, Rouleau J, Mehrabadi A, et al. Incidence, risk factors, and temporal trends in severe postpartum hemorrhage. Am J Obstet Gynecol. 2013;209(5):449.e1-7.

96. Sentilhes L, Gromez A, Clavier E, Resch B, Descamps P, Marpeau L. Long-term psychological impact of severe postpartum hemorrhage. Acta Obstet Gy-necol Scand. 2011;90(6):615-20.

97. Knight M, Callaghan WM, Berg C, Alexander S, Bouvier-Colle MH, Ford JB, et al. Trends in postpartum hemorrhage in high resource countries: a review and recommendations from the International Postpartum Hemorrhage Collab-orative Group. BMC Pregnancy Childbirth. 2009;9:55.

98. Bose P, Regan F, Paterson-Brown S. Improving the accuracy of estimated blood loss at obstetric haemorrhage using clinical reconstructions. BJOG : an international journal of obstetrics and gynaecology. 2006;113(8):919-24.

99. Briley A, Seed PT, Tydeman G, Ballard H, Waterstone M, Sandall J, et al. Reporting errors, incidence and risk factors for postpartum haemorrhage and progression to severe PPH: a prospective observational study. Bjog. 2014;121(7):876-88.

71

100. McLintock C, James AH. Obstetric hemorrhage. J Thromb Haemost. 2011;9(8):1441-51.

101. Mehrabadi A, Hutcheon JA, Lee L, Kramer MS, Liston RM, Joseph KS. Epi-demiological investigation of a temporal increase in atonic postpartum haem-orrhage: a population-based retrospective cohort study. Bjog. 2013;120(7):853-62.

102. Ford JB, Roberts CL, Simpson JM, Vaughan J, Cameron CA. Increased post-partum hemorrhage rates in Australia. International journal of gynaecology and obstetrics: the official organ of the International Federation of Gynaecol-ogy and Obstetrics. 2007;98(3):237-43.

103. Joseph KS, Sheehy O, Mehrabadi A, Urquia ML, Hutcheon JA, Kramer M, et al. Can drug effects explain the recent temporal increase in atonic postpartum haemorrhage? Paediatric and perinatal epidemiology. 2015;29(3):220-31.

104. Jiang HY, Xu LL, Li YC, Deng M, Peng CT, Ruan B. Antidepressant use dur-ing pregnancy and risk of postpartum hemorrhage: A systematic review and meta-analysis. J Psychiatr Res. 2016;83:160-7.

105. Heller HM, Ravelli ACJ, Bruning AHL, de Groot CJM, Scheele F, van Pampus MG, et al. Increased postpartum haemorrhage, the possible relation with sero-tonergic and other psychopharmacological drugs: a matched cohort study. BMC Pregnancy Childbirth. 2017;17(1):166.

106. Hanley GE, Smolina K, Mintzes B, Oberlander TF, Morgan SG. Postpartum Hemorrhage and Use of Serotonin Reuptake Inhibitor Antidepressants in Preg-nancy. Obstet Gynecol. 2016;127(3):553-61.

107. Zoega H, Kieler H, Norgaard M, Furu K, Valdimarsdottir U, Brandt L, et al. Use of SSRI and SNRI Antidepressants during Pregnancy: A Population-Based Study from Denmark, Iceland, Norway and Sweden. PLoS One. 2015;10(12):e0144474.

108. Mahdanian AA, Rej S, Bacon SL, Ozdin D, Lavoie KL, Looper K. Sero-tonergic antidepressants and perioperative bleeding risk: a systematic review. Expert opinion on drug safety. 2014;13(6):695-704.

109. Lindqvist PG, Nasiell J, Gustafsson LL, Nordstrom L. Selective serotonin reuptake inhibitor use during pregnancy increases the risk of postpartum hem-orrhage and anemia: a hospital-based cohort study. J Thromb Haemost. 2014;12(12):1986-92.

110. Andersen J, Kristensen AS, Bang-Andersen B, Stromgaard K. Recent ad-vances in the understanding of the interaction of antidepressant drugs with ser-otonin and norepinephrine transporters. Chemical communications (Cam-bridge, England). 2009(25):3677-92.

111. Cordeaux Y, Pasupathy D, Bacon J, Charnock-Jones DS, Smith GC. Charac-terization of serotonin receptors in pregnant human myometrium. The Journal of pharmacology and experimental therapeutics. 2009;328(3):682-91.

112. Bansil P, Kuklina EV, Meikle SF, Posner SF, Kourtis AP, Ellington SR, et al. Maternal and fetal outcomes among women with depression. J Womens Health (Larchmt). 2010;19(2):329-34.

113. Corwin EJ, Brownstead J, Barton N, Heckard S, Morin K. The impact of fa-tigue on the development of postpartum depression. Journal of obstetric, gyne-cologic, and neonatal nursing : JOGNN / NAACOG. 2005;34(5):577-86.

114. Corwin EJ, Murray-Kolb LE, Beard JL. Low hemoglobin level is a risk factor for postpartum depression. The Journal of nutrition. 2003;133(12):4139-42.

115. Bloch M, Rotenberg N, Koren D, Klein E. Risk factors for early postpartum depressive symptoms. General hospital psychiatry. 2006;28(1):3-8.

72

116. Alharbi AA, Abdulghani HM. Risk factors associated with postpartum depres-sion in the Saudi population. Neuropsychiatric disease and treatment. 2014;10:311-6.

117. Wilkie GL, Deligiannidis KM. Effects of perinatal depression and anxiety on labor and delivery outcomes. Obstetrics and gynecology. 2014;123 Suppl 1:82S-3S.

118. Mohammad KI, Gamble J, Creedy DK. Prevalence and factors associated with the development of antenatal and postnatal depression among Jordanian women. Midwifery. 2011;27(6):e238-45.

119. Milman N. Postpartum anemia I: definition, prevalence, causes, and conse-quences. Annals of hematology. 2011;90(11):1247-53.

120. McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B. Worldwide preva-lence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005. Public health nutrition. 2009;12(4):444-54.

121. Armony-Sivan R, Shao J, Li M, Zhao G, Zhao Z, Xu G, et al. No relationship between maternal iron status and postpartum depression in two samples in China. J Pregnancy. 2012;2012:521431.

122. Barroso F, Allard S, Kahan BC, Connolly C, Smethurst H, Choo L, et al. Prev-alence of maternal anaemia and its predictors: a multi-centre study. European journal of obstetrics, gynecology, and reproductive biology. 2011;159(1):99-105.

123. http://backenbottenutbildning.se/: Swedish Association of Obstetrics and Gynaecology, Swedish Association of Midwives; [

124. Betrán AP, Ye J, Moller A-B, Zhang J, Gülmezoglu AM, Torloni MR. The Increasing Trend in Caesarean Section Rates: Global, Regional and National Estimates: 1990-2014. PLoS ONE. 2016;11(2):e0148343.

125. Statistikdatabas för graviditet, förlossning och nyfödda http://www.so-cialstyrelsen.se/statistik/statistikdatabas/graviditeter-forlossninga-rochnyfodda: Socialstyrelsen (The National Board of Health and Welfare);

126. Carter FA, Frampton CM, Mulder RT. Cesarean section and postpartum de-pression: a review of the evidence examining the link. Psychosomatic medi-cine. 2006;68(2):321-30.

127. Hannah ME, Hannah WJ, Hodnett ED, Chalmers B, Kung R, Willan A, et al. Outcomes at 3 months after planned cesarean vs planned vaginal delivery for breech presentation at term: the international randomized Term Breech Trial. JAMA : the journal of the American Medical Association. 2002;287(14):1822-31.

128. Eisenach JC, Pan PH, Smiley R, Lavand'homme P, Landau R, Houle TT. Se-verity of acute pain after childbirth, but not type of delivery, predicts persistent pain and postpartum depression. Pain. 2008;140(1):87-94.

129. Patel RR, Murphy DJ, Peters TJ. Operative delivery and postnatal depression: a cohort study. BMJ. 2005;330(7496):879.

130. Nielsen Forman D, Videbech P, Hedegaard M, Dalby Salvig J, Secher NJ. Postpartum depression: identification of women at risk. BJOG : an interna-tional journal of obstetrics and gynaecology. 2000;107(10):1210-7.

131. Blom EA, Jansen PW, Verhulst FC, Hofman A, Raat H, Jaddoe VW, et al. Perinatal complications increase the risk of postpartum depression. The Gen-eration R Study. BJOG. 2010;117(11):1390-8.

132. Rauh C, Beetz A, Burger P, Engel A, Haberle L, Fasching PA, et al. Delivery mode and the course of pre- and postpartum depression. Archives of gynecol-ogy and obstetrics. 2012;286(6):1407-12.

73

133. Blom EA, Jansen PW, Verhulst FC, Hofman A, Raat H, Jaddoe VWV, et al. Perinatal complications increase the risk of postpartum depression. The Gen-eration R Study. BJOG: An International Journal of Obstetrics & Gynaecol-ogy. 2010;117(11):1390-8.

134. Kringeland T, Daltveit AK, Moller A. How does preference for natural child-birth relate to the actual mode of delivery? a population-based cohort study from Norway. Birth. 2010;37(1):21-7.

135. Bayrampour H, Salmon C, Vinturache A, Tough S. Effect of depressive and anxiety symptoms during pregnancy on risk of obstetric interventions. The journal of obstetrics and gynaecology research. 2015;41(7):1040-8.

136. Houston KA, Kaimal AJ, Nakagawa S, Gregorich SE, Yee LM, Kuppermann M. Mode of delivery and postpartum depression: the role of patient prefer-ences. American journal of obstetrics and gynecology. 2015;212(2):229 e1-7.

137. Raisanen S, Lehto SM, Nielsen HS, Gissler M, Kramer MR, Heinonen S. Fear of childbirth in nulliparous and multiparous women: a population-based anal-ysis of all singleton births in Finland in 1997-2010. BJOG : an international journal of obstetrics and gynaecology. 2014;121(8):965-70.

138. Schwartz L, Toohill J, Creedy DK, Baird K, Gamble J, Fenwick J. Factors associated with childbirth self-efficacy in Australian childbearing women. BMC pregnancy and childbirth. 2015;15:29.

139. Ryding EL, Wijma B, Wijma K, Rydhstrom H. Fear of childbirth during preg-nancy may increase the risk of emergency cesarean section. Acta obstetricia et gynecologica Scandinavica. 1998;77(5):542-7.

140. Ryding EL, Wirfelt E, Wangborg IB, Sjogren B, Edman G. Personality and fear of childbirth. Acta obstetricia et gynecologica Scandinavica. 2007;86(7):814-20.

141. Johnson R, Slade P. Does fear of childbirth during pregnancy predict emer-gency caesarean section? BJOG : an international journal of obstetrics and gy-naecology. 2002;109(11):1213-21.

142. Bryanton J, Gagnon AJ, Johnston C, Hatem M. Predictors of women's percep-tions of the childbirth experience. J Obstet Gynecol Neonatal Nurs. 2008;37(1):24-34.

143. Van der Gucht N, Lewis K. Women's experiences of coping with pain during childbirth: a critical review of qualitative research. Midwifery. 2015;31(3):349-58.

144. Garthus-Niegel S, Knoph C, von Soest T, Nielsen CS, Eberhard-Gran M. The role of labor pain and overall birth experience in the development of posttrau-matic stress symptoms: a longitudinal cohort study. Birth. 2014;41(1):108-15.

145. Yildiz PD, Ayers S, Phillips L. The prevalence of posttraumatic stress disorder in pregnancy and after birth: A systematic review and meta-analysis. J Affect Disord. 2017;208:634-45.

146. Karlsdottir SI, Sveinsdottir H, Kristjansdottir H, Aspelund T, Olafsdottir OA. Predictors of women's positive childbirth pain experience: Findings from an Icelandic national study. Women Birth. 2017.

147. Rowlands S, Permezel M. Physiology of pain in labour. Baillieres Clin Obstet Gynaecol. 1998;12(3):347-62.

148. Junge C, von Soest T, Weidner K, Seidler A, Eberhard-Gran M, Garthus-Nie-gel S. Labor pain in women with and without severe fear of childbirth: A pop-ulation-based, longitudinal study. Birth. 2018.

149. Lundgren I, Dahlberg K. Women's experience of pain during childbirth. Mid-wifery. 1998;14(2):105-10.

74

150. Lowe NK. The nature of labor pain. Am J Obstet Gynecol. 2002;186(5 Suppl Nature):S16-24.

151. Lang AJ, Sorrell JT, Rodgers CS, Lebeck MM. Anxiety sensitivity as a pre-dictor of labor pain. Eur J Pain. 2006;10(3):263-70.

152. Curzik D, Jokic-Begic N. Anxiety sensitivity and anxiety as correlates of ex-pected, experienced and recalled labor pain. J Psychosom Obstet Gynaecol. 2011;32(4):198-203.

153. Aksoy H, Yucel B, Aksoy U, Acmaz G, Aydin T, Babayigit MA. The relation-ship between expectation, experience and perception of labour pain: an obser-vational study. Springerplus. 2016;5(1):1766.

154. Lally JE, Murtagh MJ, Macphail S, Thomson R. More in hope than expecta-tion: a systematic review of women's expectations and experience of pain relief in labour. BMC medicine. 2008;6:7.

155. Hodnett ED. Pain and women's satisfaction with the experience of childbirth: a systematic review. Am J Obstet Gynecol. 2002;186(5 Suppl Nature):S160-72.

156. Klomp T, Mannien J, de Jonge A, Hutton EK, Lagro-Janssen AL. What do midwives need to know about approaches of women towards labour pain man-agement? A qualitative interview study into expectations of management of labour pain for pregnant women receiving midwife-led care in the Netherlands. Midwifery. 2014;30(4):432-8.

157. Larkin P, Begley CM, Devane D. Women's experiences of labour and birth: an evolutionary concept analysis. Midwifery. 2009;25(2):e49-59.

158. Hall PJ, Foster JW, Yount KM, Jennings BM. Keeping it together and falling apart: Women's dynamic experience of birth. Midwifery. 2018;58:130-6.

159. Tinti C, Schmidt S, Businaro N. Pain and emotions reported after childbirth and recalled 6 months later: the role of controllability. J Psychosom Obstet Gynaecol. 2011;32(2):98-103.

160. Waldenstrom U, Schytt E. A longitudinal study of women's memory of labour pain--from 2 months to 5 years after the birth. BJOG. 2009;116(4):577-83.

161. Niven CA, Murphy-Black T. Memory for labor pain: a review of the literature. Birth. 2000;27(4):244-53.

162. Wall PD, McMahon SB, Koltzenburg M. Wall and Melzack's textbook of pain. 5th ed. Philadelphia: Elsevier/Churchill Livingstone; 2006. xviii, 1239 p. p.

163. Sandall J, Soltani H, Gates S, Shennan A, Devane D. Midwife-led continuity models versus other models of care for childbearing women. The Cochrane database of systematic reviews. 2016;4:Cd004667.

164. Hodnett ED, Gates S, Hofmeyr GJ, Sakala C. Continuous support for women during childbirth. The Cochrane database of systematic reviews. 2013;7:CD003766.

165. Echevarria GC, Grant GJ, Chung Y, Lax J. Survey of nulliparous parturients' attitudes regarding timing of epidural analgesia initiation. J Clin Anesth. 2017;41:106-11.

166. Graviditeter, förlossningar och Pregnancies. Medicinska födelseregistret 1971-2014. Assisterad befruktning 1991-2013. Socialstyrelsen Stockholm: So-cialstyrelsen (Swedish National Board of Health and Welfare); 2015

167. Raisanen S, Kokki M, Kokki H, Gissler M, Kramer MR, Heinonen S. The use of epidural analgesia for intrapartum pain relief in publicly funded healthcare. Acta Anaesthesiol Scand. 2014;58(3):291-7.

168. Ekeus C, Hjern A, Hjelmstedt A. The need for epidural analgesia is related to birthweight - a population-based register study. Acta Obstet Gynecol Scand. 2009;88(4):397-401.

75

169. Le Ray C, Goffinet F, Palot M, Garel M, Blondel B. Factors associated with the choice of delivery without epidural analgesia in women at low risk in France. Birth. 2008;35(3):171-8.

170. Stark MA. Exploring Women's Preferences for Labor Epidural Analgesia. The Journal of perinatal education. 2003;12(2):16-21.

171. Saisto T, Salmela-Aro K, Nurmi JE, Halmesmaki E. Psychosocial predictors of disappointment with delivery and puerperal depression. A longitudinal study. Acta Obstet Gynecol Scand. 2001;80(1):39-45.

172. Ding T, Wang DX, Qu Y, Chen Q, Zhu SN. Epidural labor analgesia is asso-ciated with a decreased risk of postpartum depression: a prospective cohort study. Anesth Analg. 2014;119(2):383-92.

173. Lim G, Farrell LM, Facco FL, Gold MS, Wasan AD. Labor Analgesia as a Predictor for Reduced Postpartum Depression Scores: A Retrospective Obser-vational Study. Anesth Analg. 2018;126(5):1598-605.

174. Hiltunen P, Raudaskoski T, Ebeling H, Moilanen I. Does pain relief during delivery decrease the risk of postnatal depression? Acta obstetricia et gyneco-logica Scandinavica. 2004;83(3):257-61.

175. Orbach-Zinger S, Landau R, Harousch AB, Ovad O, Caspi L, Kornilov E, et al. The Relationship Between Women's Intention to Request a Labor Epidural Analgesia, Actually Delivering With Labor Epidural Analgesia, and Postpar-tum Depression at 6 Weeks: A Prospective Observational Study. Anesth An-alg. 2017.

176. Howell CJ, Kidd C, Roberts W, Upton P, Lucking L, Jones PW, et al. A ran-domised controlled trial of epidural compared with non-epidural analgesia in labour. BJOG. 2001;108(1):27-33.

177. Waldenstrom U, Borg IM, Olsson B, Skold M, Wall S. The childbirth experi-ence: a study of 295 new mothers. Birth. 1996;23(3):144-53.

178. Nilver H, Begley C, Berg M. Measuring women's childbirth experiences: a systematic review for identification and analysis of validated instruments. BMC Pregnancy Childbirth. 2017;17(1):203.

179. Bell AF, Carter CS, Davis JM, Golding J, Adejumo O, Pyra M, et al. Childbirth and symptoms of postpartum depression and anxiety: a prospective birth co-hort study. Arch Womens Ment Health. 2016;19(2):219-27.

180. Waldenstrom U. Experience of labor and birth in 1111 women. J Psychosom Res. 1999;47(5):471-82.

181. Stadlmayr W, Amsler F, Lemola S, Stein S, Alt M, Burgin D, et al. Memory of childbirth in the second year: the long-term effect of a negative birth expe-rience and its modulation by the perceived intranatal relationship with caregiv-ers. J Psychosom Obstet Gynaecol. 2006;27(4):211-24.

182. Simkin P. Just another day in a woman's life? Women's long-term perceptions of their first birth experience. Part I. Birth. 1991;18(4):203-10.

183. Takehara K, Noguchi M, Shimane T, Misago C. A longitudinal study of wom-en's memories of their childbirth experiences at five years postpartum. BMC Pregnancy Childbirth. 2014;14:221.

184. Pang MW, Leung TN, Lau TK, Hang Chung TK. Impact of first childbirth on changes in women's preference for mode of delivery: follow-up of a longitudi-nal observational study. Birth. 2008;35(2):121-8.

185. Nelson AM. Transition to motherhood. J Obstet Gynecol Neonatal Nurs. 2003;32(4):465-77.

76

186. Ekselius L. Diagnostik och uppföljning av förstämningssyndrom : en system-atisk litteraturöversikt. Stockholm: Statens beredning för medicinskt utvärder-ing (Swedish agency for helath technology assessment and assessment of so-cial services); 2012.

187. O'Hara MW, McCabe JE. Postpartum depression: current status and future di-rections. Annu Rev Clin Psychol. 2013;9:379-407.

188. Gaynes BN, Gavin N, Meltzer-Brody S, Lohr KN, Swinson T, Gartlehner G, et al. Perinatal depression: prevalence, screening accuracy, and screening out-comes. Evidence report/technology assessment (Summary). 2005(119):1-8.

189. Dennis CL, Hodnett E. Psychosocial and psychological interventions for treat-ing postpartum depression. The Cochrane database of systematic reviews. 2007(4):Cd006116.

190. Meltzer-Brody S, Howard LM, Bergink V, Vigod S, Jones I, Munk-Olsen T, et al. Postpartum psychiatric disorders. Nature reviews Disease primers. 2018;4:18022.

191. Abramowitz JS, Meltzer-Brody S, Leserman J, Killenberg S, Rinaldi K, Ma-haffey BL, et al. Obsessional thoughts and compulsive behaviors in a sample of women with postpartum mood symptoms. Arch Womens Ment Health. 2010;13(6):523-30.

192. Brockington IF, Macdonald E, Wainscott G. Anxiety, obsessions and morbid preoccupations in pregnancy and the puerperium. Arch Womens Ment Health. 2006;9(5):253-63.

193. Howard LM, Molyneaux E, Dennis CL, Rochat T, Stein A, Milgrom J. Non-psychotic mental disorders in the perinatal period. Lancet. 2014;384(9956):1775-88.

194. Wisner KL, Sit DK, McShea MC, Rizzo DM, Zoretich RA, Hughes CL, et al. Onset timing, thoughts of self-harm, and diagnoses in postpartum women with screen-positive depression findings. JAMA psychiatry. 2013;70(5):490-8.

195. Munk-Olsen T, Laursen TM, Mendelson T, Pedersen CB, Mors O, Mortensen PB. Risks and predictors of readmission for a mental disorder during the post-partum period. Arch Gen Psychiatry. 2009;66(2):189-95.

196. Brummelte S, Galea LA. Postpartum depression: Etiology, treatment and con-sequences for maternal care. Horm Behav. 2016;77:153-66.

197. Woody CA, Ferrari AJ, Siskind DJ, Whiteford HA, Harris MG. A systematic review and meta-regression of the prevalence and incidence of perinatal de-pression. J Affect Disord. 2017;219:86-92.

198. Di Florio A, Putnam K, Altemus M, Apter G, Bergink V, Bilszta J, et al. The impact of education, country, race and ethnicity on the self-report of postpar-tum depression using the Edinburgh Postnatal Depression Scale. Psychol Med. 2017;47(5):787-99.

199. Borgstrom A, Odlind V, Ekselius L, Sundstrom-Poromaa I. Adverse mood ef-fects of combined oral contraceptives in relation to personality traits. Eur J Obstet Gynecol Reprod Biol. 2008;141(2):127-30.

200. Gjerdingen DK, Yawn BP. Postpartum depression screening: importance, methods, barriers, and recommendations for practice. J Am Board Fam Med. 2007;20(3):280-8.

201. Button S, Thornton A, Lee S, Shakespeare J, Ayers S. Seeking help for peri-natal psychological distress: a meta-synthesis of women's experiences. The British journal of general practice : the journal of the Royal College of General Practitioners. 2017;67(663):e692-e9.

202. Shidhaye R, Mendenhall E, Sumathipala K, Sumathipala A, Patel V. Associa-tion of somatoform disorders with anxiety and depression in women in low and

77

middle income countries: a systematic review. Int Rev Psychiatry. 2013;25(1):65-76.

203. Agius A, Xuereb RB, Carrick-Sen D, Sultana R, Rankin J. The co-existence of depression, anxiety and post-traumatic stress symptoms in the perinatal period: A systematic review. Midwifery. 2016;36:70-9.

204. Falah-Hassani K, Shiri R, Dennis CL. The prevalence of antenatal and postna-tal co-morbid anxiety and depression: a meta-analysis. Psychol Med. 2017;47(12):2041-53.

205. Munk-Olsen T, Laursen TM, Meltzer-Brody S, Mortensen PB, Jones I. Psy-chiatric disorders with postpartum onset: possible early manifestations of bi-polar affective disorders. Arch Gen Psychiatry. 2012;69(4):428-34.

206. Guintivano J, Arad M, Gould TD, Payne JL, Kaminsky ZA. Antenatal predic-tion of postpartum depression with blood DNA methylation biomarkers. Mo-lecular psychiatry. 2014;19(5):560-7.

207. Osborne L, Clive M, Kimmel M, Gispen F, Guintivano J, Brown T, et al. Rep-lication of Epigenetic Postpartum Depression Biomarkers and Variation with Hormone Levels. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology. 2016;41(6):1648-58.

208. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Develop-ment of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782-6.

209. Krantz I, Eriksson B, Lundquist-Persson C, Ahlberg BM, Nilstun T. Screening for postpartum depression with the Edinburgh Postnatal Depression Scale (EPDS): an ethical analysis. Scandinavian journal of public health. 2008;36(2):211-6.

210. Myers ER, Aubuchon-Endsley N, Bastian LA, Gierisch JM, Kemper AR, Swamy GK, et al. Efficacy and Safety of Screening for Postpartum Depres-sion. AHRQ Comparative Effectiveness Reviews. Rockville (MD)2013.

211. Hewitt CE, Gilbody SM. Is it clinically and cost effective to screen for postna-tal depression: a systematic review of controlled clinical trials and economic evidence. Bjog. 2009;116(8):1019-27.

212. Socialstyrelsen. Vägledning för barnhälsovården: Edita Bobergs; 2014. 169 p. 213. Förebyggande av postpartumdepression. Psykosocial och psykologisk

profylax mot depression efter förlossning. Stockholm: Statens beredning för medicinsk och social utvärdering. (Swedish Agency for Health Technology Assessment and Assessment of Social Services); 2014. 6 p.

214. Dennis CL, Dowswell T. Psychosocial and psychological interventions for pre-venting postpartum depression. Cochrane Database Syst Rev. 2013(2):Cd001134.

215. Morrell CJ, Sutcliffe P, Booth A, Stevens J, Scope A, Stevenson M, et al. A systematic review, evidence synthesis and meta-analysis of quantitative and qualitative studies evaluating the clinical effectiveness, the cost-effectiveness, safety and acceptability of interventions to prevent postnatal depression. Health technology assessment (Winchester, England). 2016;20(37):1-414.

216. Werner E, Miller M, Osborne LM, Kuzava S, Monk C. Preventing postpartum depression: review and recommendations. Arch Womens Ment Health. 2015;18(1):41-60.

217. Molyneaux E, Telesia LA, Henshaw C, Boath E, Bradley E, Howard LM. An-tidepressants for preventing postnatal depression. The Cochrane database of systematic reviews. 2018;4:Cd004363.

218. Hadfield H, Wittkowski A. Women's Experiences of Seeking and Receiving Psychological and Psychosocial Interventions for Postpartum Depression: A

78

Systematic Review and Thematic Synthesis of the Qualitative Literature. Jour-nal of midwifery & women's health. 2017;62(6):723-36.

219. Stephens S, Ford E, Paudyal P, Smith H. Effectiveness of Psychological Inter-ventions for Postnatal Depression in Primary Care: A Meta-Analysis. Ann Fam Med. 2016;14(5):463-72.

220. Fitelson E, Kim S, Baker AS, Leight K. Treatment of postpartum depression: clinical, psychological and pharmacological options. International journal of women's health. 2010;3:1-14.

221. Sockol LE. A systematic review and meta-analysis of interpersonal psycho-therapy for perinatal women. J Affect Disord. 2018;232:316-28.

222. Sockol LE. A systematic review of the efficacy of cognitive behavioral therapy for treating and preventing perinatal depression. J Affect Disord. 2015;177:7-21.

223. Lin PZ, Xue JM, Yang B, Li M, Cao FL. Effectiveness of self-help psycholog-ical interventions for treating and preventing postpartum depression: a meta-analysis. Arch Womens Ment Health. 2018.

224. Lau Y, Htun TP, Wong SN, Tam WSW, Klainin-Yobas P. Therapist-Sup-ported Internet-Based Cognitive Behavior Therapy for Stress, Anxiety, and Depressive Symptoms Among Postpartum Women: A Systematic Review and Meta-Analysis. Journal of medical Internet research. 2017;19(4):e138.

225. Kim DR, Epperson CN, Weiss AR, Wisner KL. Pharmacotherapy of postpar-tum depression: an update. Expert opinion on pharmacotherapy. 2014;15(9):1223-34.

226. Molyneaux E, Howard LM, McGeown HR, Karia AM, Trevillion K. Antide-pressant treatment for postnatal depression. The Cochrane database of system-atic reviews. 2014(9):Cd002018.

227. Wisner KL, Hanusa BH, Perel JM, Peindl KS, Piontek CM, Sit DK, et al. Post-partum depression: a randomized trial of sertraline versus nortriptyline. J Clin Psychopharmacol. 2006;26(4):353-60.

228. Burrell-Ward H, Fromson JA, Cooper JJ, De Oliveira G, Almeida M. Recom-mendations for the use of ECT in pregnancy: literature review and proposed clinical protocol. Arch Womens Ment Health. 2018.

229. Rundgren S, Brus O, Bave U, Landen M, Lundberg J, Nordanskog P, et al. Improvement of postpartum depression and psychosis after electroconvulsive therapy: A population-based study with a matched comparison group. J Affect Disord. 2018;235:258-64.

230. ECT: Kliniska riktlinjer för elektrokonvulsiv behandling. Stockholm. Svenska Psykiatriska föreningen; 2014.

231. Filippi V, Chou D, Ronsmans C, Graham W, Say L. Levels and Causes of Maternal Mortality and Morbidity. In: Black RE, Laxminarayan R, Temmer-man M, Walker N, editors. Reproductive, Maternal, Newborn, and Child Health: Disease Control Priorities, Third Edition (Volume 2). Washington (DC)2016.

232. Koblinsky M, Chowdhury ME, Moran A, Ronsmans C. Maternal morbidity and disability and their consequences: neglected agenda in maternal health. Journal of health, population, and nutrition. 2012;30(2):124-30.

233. Esscher A, Essen B, Innala E, Papadopoulos FC, Skalkidou A, Sundstrom-Poromaa I, et al. Suicides during pregnancy and 1 year postpartum in Sweden, 1980-2007. Br J Psychiatry. 2016;208(5):462-9.

234. Lysell H, Dahlin M, Viktorin A, Ljungberg E, D'Onofrio BM, Dickman P, et al. Maternal suicide - Register based study of all suicides occurring after de-livery in Sweden 1974-2009. PLoS One. 2018;13(1):e0190133.

79

235. Johannsen BM, Larsen JT, Laursen TM, Bergink V, Meltzer-Brody S, Munk-Olsen T. All-Cause Mortality in Women With Severe Postpartum Psychiatric Disorders. The American journal of psychiatry. 2016;173(6):635-42.

236. Orsolini L, Valchera A, Vecchiotti R, Tomasetti C, Iasevoli F, Fornaro M, et al. Suicide during Perinatal Period: Epidemiology, Risk Factors, and Clinical Correlates. Frontiers in psychiatry. 2016;7:138.

237. Bauer A, Knapp M, Parsonage M. Lifetime costs of perinatal anxiety and de-pression. J Affect Disord. 2016;192:83-90.

238. Bauer AP, M., Knapp, M., Iemmi, V.Adjela, B. The Costs of Perinatal Mental Health Problems. London: London School of Economics and the Centre for Mental Health, Health LSoEatCfM; 2014.

239. Barlow DH, Ellard KK, Sauer-Zavala S, Bullis JR, Carl JR. The Origins of Neuroticism. Perspect Psychol Sci. 2014;9(5):481-96.

240. Corr PJM, Gerald The Cambridge handbook of personality psychology Cam-bridge, U.K,: Cambridge University Press.; 2009.

241. Sanchez-Roige S, Gray JC, MacKillop J, Chen CH, Palmer AA. The genetics of human personality. Genes, brain, and behavior. 2018;17(3):e12439.

242. Newton-Howes G, Clark LA, Chanen A. Personality disorder across the life course. Lancet. 2015;385(9969):727-34.

243. McCrae RR, Costa PT, Jr., Ostendorf F, Angleitner A, Hrebickova M, Avia MD, et al. Nature over nurture: temperament, personality, and life span devel-opment. J Pers Soc Psychol. 2000;78(1):173-86.

244. De Pauw SS, Mervielde I. Temperament, personality and developmental psy-chopathology: a review based on the conceptual dimensions underlying child-hood traits. Child psychiatry and human development. 2010;41(3):313-29.

245. Roberts BW, Walton KE, Viechtbauer W. Patterns of mean-level change in personality traits across the life course: a meta-analysis of longitudinal studies. Psychol Bull. 2006;132(1):1-25.

246. Kotov R, Gamez W, Schmidt F, Watson D. Linking "big" personality traits to anxiety, depressive, and substance use disorders: a meta-analysis. Psychol Bull. 2010;136(5):768-821.

247. McCrae RR, Costa PT, Jr. Personality trait structure as a human universal. Am Psychol. 1997;52(5):509-16.

248. Cloninger CR. A systematic method for clinical description and classification of personality variants. A proposal. Arch Gen Psychiatry. 1987;44(6):573-88.

249. Seekles WM, Cuijpers P, van de Ven P, Penninx BW, Verhaak PF, Beekman AT, et al. Personality and perceived need for mental health care among primary care patients. J Affect Disord. 2012;136(3):666-74.

250. Connor-Smith JK, Flachsbart C. Relations between personality and coping: a meta-analysis. J Pers Soc Psychol. 2007;93(6):1080-107.

251. Saisto T, Salmela-Aro K, Nurmi JE, Halmesmaki E. Psychosocial characteris-tics of women and their partners fearing vaginal childbirth. BJOG. 2001;108(5):492-8.

252. Handelzalts JE, Becker G, Ahren MP, Lurie S, Raz N, Tamir Z, et al. Person-ality, fear of childbirth and birth outcomes in nulliparous women. Arch Gyne-col Obstet. 2015;291(5):1055-62.

253. Peñacoba-Puente C, Carmona Monge FJ, Carretero Abellán I, Marín Morales D. Effects of Personality on Psychiatric and Somatic Symptoms in Pregnant Women. Psychology of Women Quarterly. 2011;35(2):293-302.

254. Henshaw C. Mood disturbance in the early puerperium: a review. Arch Wom-ens Ment Health. 2003;6 Suppl 2:S33-42.

80

255. Lahey BB. Public health significance of neuroticism. Am Psychol. 2009;64(4):241-56.

256. Hakulinen C, Hintsanen M, Munafo MR, Virtanen M, Kivimaki M, Batty GD, et al. Personality and smoking: individual-participant meta-analysis of nine co-hort studies. Addiction. 2015;110(11):1844-52.

257. Davis RE, Smitherman TA, Baskin SM. Personality traits, personality disor-ders, and migraine: a review. Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophys-iology. 2013;34 Suppl 1:S7-10.

258. Shipley BA, Weiss A, Der G, Taylor MD, Deary IJ. Neuroticism, extraversion, and mortality in the UK Health and Lifestyle Survey: a 21-year prospective cohort study. Psychosom Med. 2007;69(9):923-31.

259. Terracciano A, Lockenhoff CE, Zonderman AB, Ferrucci L, Costa PT, Jr. Per-sonality predictors of longevity: activity, emotional stability, and conscien-tiousness. Psychosom Med. 2008;70(6):621-7.

260. Mommersteeg PM, Pouwer F. Personality as a risk factor for the metabolic syndrome: a systematic review. J Psychosom Res. 2012;73(5):326-33.

261. Johnston RG, Brown AE. Maternal trait personality and childbirth: the role of extraversion and neuroticism. Midwifery. 2013;29(11):1244-50.

262. Handelzalts JE, Krissi H, Levy S, Freund Y, Carmiel N, Ashwal E, et al. Per-sonality, preterm labor contractions, and psychological consequences. Arch Gynecol Obstet. 2016;293(3):575-82.

263. Vollrath ME, Sengpiel V, Landolt MA, Jacobsson B, Latal B. Is maternal trait anxiety a risk factor for late preterm and early term deliveries? BMC Preg-nancy Childbirth. 2016;16(1):286.

264. Chatzi L, Koutra K, Vassilaki M, Vardiampasis A, Georgiou V, Koutis A, et al. Maternal personality traits and risk of preterm birth and fetal growth re-striction. European psychiatry : the journal of the Association of European Psy-chiatrists. 2013;28(4):213-8.

265. The Swedish Medical Birth Register: a summary of content and quality. Stock-holm: Swedish National Board of Health and Welfare (Socialstyrelsen); 2003. Report No.: 2003-112-3.

266. Kvalitet och innehåll i patientregistret. Utskrivningar från slutenvården 1964-2007 och besök i specialiserad öppenvård (exklusive primärvårdsbesök) 1997-2007. Stockholm: Socialstyrelsen (Swedish National Board of Health and Wel-fare); 2009.

267. Wettermark B, Hammar N, Fored CM, Leimanis A, Otterblad Olausson P, Bergman U, et al. The new Swedish Prescribed Drug Register--opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiology and drug safety. 2007;16(7):726-35.

268. Galesic M, Bosnjak M. Effects of Questionnaire Length on Participation and Indicators of Response Quality in a Web Survey. Public Opinion Quarterly. 2009;73(2):349-60.

269. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr. 1996;85(7):843-8.

270. Rubertsson C, Borjesson K, Berglund A, Josefsson A, Sydsjo G. The Swedish validation of Edinburgh Postnatal Depression Scale (EPDS) during pregnancy. Nordic journal of psychiatry. 2011;65(6):414-8.

81

271. Gibson J, McKenzie-McHarg K, Shakespeare J, Price J, Gray R. A systematic review of studies validating the Edinburgh Postnatal Depression Scale in ante-partum and postpartum women. Acta psychiatrica Scandinavica. 2009;119(5):350-64.

272. Wickberg B, Hwang CP. The Edinburgh Postnatal Depression Scale: valida-tion on a Swedish community sample. Acta psychiatrica Scandinavica. 1996;94(3):181-4.

273. Gustavsson JP, Bergman H, Edman G, Ekselius L, von Knorring L, Linder J. Swedish universities Scales of Personality (SSP): construction, internal con-sistency and normative data. Acta Psychiatr Scand. 2000;102(3):217-25.

274. Willebrand M, Kildal M, Andersson G, Ekselius L. Long-term assessment of personality after burn trauma in adults. J Nerv Ment Dis. 2002;190(1):53-6.

275. Lydsdottir LB, Howard LM, Olafsdottir H, Einarsson H, Steingrimsdottir T, Sigurdsson JF. Adverse life experiences and common mental health problems in pregnancy: a causal pathway analysis. Arch Womens Ment Health. 2018.

276. Maleki A, Faghihzadeh S, Niroomand S. The relationship between paternal prenatal depressive symptoms with postnatal depression: The PATH model. Psychiatry Res. 2018;269:102-7.

277. Lowndes TA, Egan SJ, McEvoy PM. Efficacy of brief guided self-help cogni-tive behavioral treatment for perfectionism in reducing perinatal depression and anxiety: a randomized controlled trial. Cognitive behaviour therapy. 2018:1-15.

278. Iliadis SI, Comasco E, Hellgren C, Kollia N, Sundstrom Poromaa I, Skalkidou A. Associations between a polymorphism in the hydroxysteroid (11-beta) de-hydrogenase 1 gene, neuroticism and postpartum depression. J Affect Disord. 2017;207:141-7.

279. Gutierrez-Zotes A, Labad J, Martin-Santos R, Garcia-Esteve L, Gelabert E, Jover M, et al. Coping strategies and postpartum depressive symptoms: A structural equation modelling approach. European psychiatry : the journal of the Association of European Psychiatrists. 2015;30(6):701-8.

280. Fawcett J, Aber C, Haussler S, Weiss M, Myers ST, Hall JL, et al. Women's perceptions of caesarean birth: a Roy international study. Nursing science quarterly. 2011;24(4):352-62.

281. Anderson J.C Gerbing D.W. Structural equation modeling in practice. A re-view and recommended two-step approach. PSychological Bulletin. 1998;103(3):411-23.

282. Guittier MJ, Cedraschi C, Jamei N, Boulvain M, Guillemin F. Impact of mode of delivery on the birth experience in first-time mothers: a qualitative study. BMC Pregnancy Childbirth. 2014;14:254.

283. Garthus-Niegel S, von Soest T, Knoph C, Simonsen TB, Torgersen L, Eber-hard-Gran M. The influence of women's preferences and actual mode of deliv-ery on post-traumatic stress symptoms following childbirth: a population-based, longitudinal study. BMC Pregnancy Childbirth. 2014;14:191.

284. Gurber S, Baumeler L, Grob A, Surbek D, Stadlmayr W. Antenatal depressive symptoms and subjective birth experience in association with postpartum de-pressive symptoms and acute stress reaction in mothers and fathers: A longi-tudinal path analysis. European journal of obstetrics, gynecology, and repro-ductive biology. 2017;215:68-74.

285. Furuta M, Sandall J, Cooper D, Bick D. The relationship between severe ma-ternal morbidity and psychological health symptoms at 6-8 weeks postpartum: a prospective cohort study in one English maternity unit. BMC pregnancy and childbirth. 2014;14:133.

82

286. Weisman O, Granat A, Gilboa-Schechtman E, Singer M, Gordon I, Azulay H, et al. The experience of labor, maternal perception of the infant, and the moth-er's postpartum mood in a low-risk community cohort. Archives of women's mental health. 2010;13(6):505-13.

287. Beard JL, Hendricks MK, Perez EM, Murray-Kolb LE, Berg A, Vernon-Feagans L, et al. Maternal iron deficiency anemia affects postpartum emotions and cognition. The Journal of nutrition. 2005;135(2):267-72.

288. Corwin EJ, Arbour M. Postpartum fatigue and evidence-based interventions. MCN The American journal of maternal child nursing. 2007;32(4):215-20; quiz 21-2.

289. Beck CT. Predictors of postpartum depression: an update. Nurs Res. 2001;50(5):275-85.

290. Haines HM, Pallant JF, Fenwick J, Gamble J, Creedy DK, Toohill J, et al. Identifying women who are afraid of giving birth: A comparison of the fear of birth scale with the WDEQ-A in a large Australian cohort. Sex Reprod Healthc. 2015;6(4):204-10.

291. Tuohy A, McVey C. Experience of pregnancy and delivery as predictors of postpartum depression. Psychology, health & medicine. 2008;13(1):43-7.

292. Jansen AJ, Duvekot JJ, Hop WC, Essink-Bot ML, Beckers EA, Karsdorp VH, et al. New insights into fatigue and health-related quality of life after delivery. Acta obstetricia et gynecologica Scandinavica. 2007;86(5):579-84.

293. Bergholt T, Stenderup JK, Vedsted-Jakobsen A, Helm P, Lenstrup C. In-traoperative surgical complication during cesarean section: an observational study of the incidence and risk factors. Acta obstetricia et gynecologica Scan-dinavica. 2003;82(3):251-6.

294. Porter M, van Teijlingen E, Chi Ying Yip L, Bhattacharya S. Satisfaction with cesarean section: qualitative analysis of open-ended questions in a large postal survey. Birth. 2007;34(2):148-54.

295. Redshaw M, Hockley C. Institutional processes and individual responses: women's experiences of care in relation to cesarean birth. Birth. 2010;37(2):150-9.

296. Lobel M, DeLuca RS. Psychosocial sequelae of cesarean delivery: review and analysis of their causes and implications. Soc Sci Med. 2007;64(11):2272-84.

297. Somera MJ, Feeley N, Ciofani L. Women's experience of an emergency cae-sarean birth. Journal of clinical nursing. 2010;19(19-20):2824-31.

298. Nystedt A, Hogberg U, Lundman B. Some Swedish women's experiences of prolonged labour. Midwifery. 2006;22(1):56-65.

299. Uotila JT, Taurio K, Salmelin R, Kirkinen P. Traumatic experience with vac-uum extraction-influence of personal preparation, physiology, and treatment during labor. J Perinat Med. 2005;33(5):373-8.

300. Carlsson IM, Ziegert K, Nissen E. The relationship between childbirth self-efficacy and aspects of well-being, birth interventions and birth outcomes. Midwifery. 2015;31(10):1000-7.

301. Salomonsson B, Gullberg MT, Alehagen S, Wijma K. Self-efficacy beliefs and fear of childbirth in nulliparous women. J Psychosom Obstet Gynaecol. 2013;34(3):116-21.

302. Karlsdottir SI, Sveinsdottir H, Olafsdottir OA, Kristjansdottir H. Pregnant women's expectations about pain intensity during childbirth and their attitudes towards pain management: Findings from an Icelandic national study. Sexual & Reproductive Healthcare. 2015;6(4):211-8.

83

303. Chaput KH, Vinturache A. Methodologic Concerns Regarding a Study Con-cluding That Epidural Labor Analgesia Is Associated with a Decreased Risk of Postpartum Depression. Anesth Analg. 2015;121(6):1682-3.

304. Lim G, Chelly JE. More Research Is Required to Demonstrate a Relationship Between Intrapartum Pain Management and Postpartum Mood Disorders. Anesth Analg. 2015;121(2):575.

305. Shrier I, Platt RW. Reducing bias through directed acyclic graphs. BMC med-ical research methodology. 2008;8:70.

306. Nohr EA, Liew Z. How to investigate and adjust for selection bias in cohort studies. Acta Obstet Gynecol Scand. 2018;97(4):407-16.

307. Bergman P, Ahlberg G, Forsell Y, Lundberg I. Non-participation in the second wave of the PART study on mental disorder and its effects on risk estimates. Int J Soc Psychiatry. 2010;56(2):119-32.

308. Saßenroth D. The Impact of Personality on Participation Decisions in Surveys. A contribution to the Discussion on Unit Nonresponse: Carl von Ossietzky University, Oldenburg, Germany; 2012.

309. Josefsson A, Berg G, Nordin C, Sydsjo G. Prevalence of depressive symptoms in late pregnancy and postpartum. Acta Obstet Gynecol Scand. 2001;80(3):251-5.

310. Rubertsson C, Wickberg B, Gustavsson P, Radestad I. Depressive symptoms in early pregnancy, two months and one year postpartum-prevalence and psy-chosocial risk factors in a national Swedish sample. Arch Womens Ment Health. 2005;8(2):97-104.

311. Moraes GP, Lorenzo L, Pontes GA, Montenegro MC, Cantilino A. Screening and diagnosing postpartum depression: when and how? Trends in psychiatry and psychotherapy. 2017;39(1):54-61.

312. Sanchez-Villegas A, Schlatter J, Ortuno F, Lahortiga F, Pla J, Benito S, et al. Validity of a self-reported diagnosis of depression among participants in a co-hort study using the Structured Clinical Interview for DSM-IV (SCID-I). BMC Psychiatry. 2008;8:43.

313. Stuart AL, Pasco JA, Jacka FN, Brennan SL, Berk M, Williams LJ. Compari-son of self-report and structured clinical interview in the identification of de-pression. Compr Psychiatry. 2014;55(4):866-9.

314. Kesmodel US. Information bias in epidemiological studies with a special focus on obstetrics and gynecology. Acta Obstet Gynecol Scand. 2018;97(4):417-23.

315. Van den Bussche E, Crombez G, Eccleston C, Sullivan MJ. Why women prefer epidural analgesia during childbirth: the role of beliefs about epidural analgesia and pain catastrophizing. Eur J Pain. 2007;11(3):275-82.

316. Dennis CL, Boyce P. Further psychometric testing of a brief personality scale to measure vulnerability to postpartum depression. J Psychosom Obstet Gy-naecol. 2004;25(3-4):305-11.

84

Appendix

Participants in BASIC Appendix 1. Flowchart of participants in BASIC September 2009 until March 2018

85

Participants in UPPSAT Appendix 2. Flowcharts of participants in UPPSAT June 2006 to May 2007

86

Conceptual Directed Acyclic Graph (Study III) Appendix 3. Conceptual Directed Acyclic Graph of the association of epidural anal-gesia (EDA) during delivery and postpartum depression (PPD). Green arrows repre-sent casual paths. Red arrows represent biasing paths.

87

Directed Acyclic Graph including available variables (Study III) Appendix 4. Directed Acyclic Graph of the association of epidural analgesia (EDA) during delivery and postpartum depressive symptoms (PPDS) and variables included in the current study. Green arrows represent casual paths. Red arrows represent bias-ing paths.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1503

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-362566

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018