POSTPARTUM ANXIETY AND INTRAPSYCHIC VULNERABILITY REFLECTED BY EARLY MALADAPTIVE SCHEMAS

Background : The prevalence of postpartum anxiety is high and its negative consequences are significant. Most research studies have aimed at exploring the demographic and psychosocial risk factors while neglecting factors of intrapsychic vulnerability . The aim of our study is to examine postpartum anxiety from this aspect, that is, uncovering the early relational experiences of mothers, specifically regarding their early maladaptive schemas. Methods : 125 women participated in online data collection in the first year after giving birth. Maternal anxiety was measured with the postpartum version of STAI-S, while early maladaptive schemas were assessed with the Schema Questionnaire. Results : In terms of demographics, most of the mothers in our sample had middle-class characteristics, but 21.6% had clinical, and 34.4% had subclinical levels of anxiety. Our results confirmed a significant, direct effect of maladaptive schemas on postpartum anxiety in case of 13 of the 15 measured schemas. Discussion : This study draws attention to the intrapsychic vulnerability of mothers as a highly neglected etiological aspect of postpartum anxiety. Further research of intrapsychic and, especially, attachment-related vulnerability should be carried out to identify at-risk women and treat postpartum anxiety more adequately. Limitations : The sample consisted of highly educated, married women with planned children, so the results are not representative for mothers in general.


Introduction
Postpartum mental disorders are among the most common complications of childbirth. Navarro and colleagues found 18.1% prevalence of non-psychotic mental disorders at six weeks postpartum (NAVARRO et al. 2008). These disorders impede the successful fulfilment of the postnatal period viewed as normative crises, and have negative effects on the mother's self-confidence, self-efficacy (RECK et al. 2012), on the partnership (WENZEL et al. 2005;MERLE-FISHMAN 2010) and on the quality of the mother-child bonding (MOEHLER et al. 2006). According to the traditional classification, postpartum disorders can be divided into three categories: postpartum baby blues, postpartum depression and postpartum psychosis (SHARMA & BURT 2011). However, recently more experts have found this classification oversimplified (BROCKINGTON 2004) and have point to a greater variety of postpartum distress (BROCKINGTON 2004;MILLER et al. 2006). Postnatal anxiety disorders seem to be the most neglected mental states after childbirth, despite their high prevalence that is close to (or exceed) the 13% prevalence of postpartum depression (RECK et al. 2008;WENZEL et al. 2003;WENZEL et al. 2005;MATTHEY et al., 2003;O'HARA & SWAIN 1996). Pioneering studies of the topic aimed at raising awareness of these clinically relevant conditions by describing their symptomatology and course, and assessing their prevalence (eg. WENZEL et al. 2005;FAIRBROTHER & ABRAMOWITZ 2007;RECK et al. 2008). Others tried to identify the risk factors of postpartum anxiety, focusing mainly on demographic and psychosocial aspects. The most significant predictors that have been found are: maternal age (WENZEL et al. 2005;MARTINI et al. 2015), psychiatric disorders in the personal and family history (WENZEL et al. 2005;BRITTON 2008), maternal education (MARTINI et al. 2015), lack of pregnancy planning, perceived peripartum distress, high trait-anxiety and low education (BRITTON 2008). Further research have found that previous traumatic events played an important role in the etiology of postpartum PTSD (AYERS et al. 2009;VERREAULT et al. 2012), along with parity, delivery type (AYERS et al. 2009), anxiety sensitivity (VERREAULT et al. 2012), traumatic experiences at previous births, and fear of dying during childbirth (POLACHEK et al. 2012). HAAGEN and colleagues (2015) identified state anxiety, antenatal depressive symptoms, perinatal psychoform and somatoform dissociation as risk factors for PTSD symptoms at three months postpartum.
Only a few researchers have investigated intrapsychic vulnerability to pathologic al postpartum anxiety. FAIRBROTHER and ABRAMOWITZ (2007) interpreted the etiology of postpartum obsessive-compulsive (OCD) disorder within a cognitive-behavioural framework. They pointed to the role of increased responsibility, catastrophising of intrusive thoughts and the negative confirmatory effect of compulsive checking as significant contributors to the development or exacerbation of postpartum OCD.
However, little attention has been paid to some other intrapsychic factors that could be strongly activated in the process of becoming a mother, such as the woman's early relational experiences. According to MERLE-FISHMAN (2010), the complex psychological process of becoming a mother involves the reorganisation of the self. Its experience and outcome is influenced by the mother's own early relational experiences, representations and re-lived memories. Activation of the mother's possible attachment injuries and other negative experiences may be accompanied by intense emotional fluctuation, anxiety, and uncertainty related to motherhood and ambivalence.
Representations of early experiences with the primary caregiver, and the cognitive-affective content of internal working models (BOWLBY 1973) can well be examined through early maladaptive schemas. These are based on mother-infant interactions (YOUNG et al. 2003), organised through the dimensions of the self and the other. Maladaptive schemas are self-fulfilling in nature and stable over time, and deeply influence adult attachment. Schemas are activated by triggers that are similar to the original childhood experience in which they are rooted. Activated schemas flood individuals with intense, disruptive negative emotions, like fear, anger or shame (YOUNG et al. 2003).
Regarding postpartum depression, some studies verified a significant link between insecure attachment and depressive symptoms (SIMPSON & RHOLES, 2003;IKEDA et al. 2014;MCMAHON et al. 2005). Though these results provide empirical evidence for the effect of mothers' early relational experiences on the level of postpartum adjustment, this approach is still neglected in the research of pathological postpartum anxiety.

Aim of the study
The aim of this study was to investigate mothers' intrapsychic vulnerability to postpartum anxiety, specifically in the light of early maladaptive schemas. We supposed that 1) women with early maladaptive schemata are more vulnerable for clinicallevel anxiety in the postpartum period; 2) unmet core emotional needs, indicated by the maladaptive activity of certain schema-domains, increase mothers' vulnerability to postpartum anxiety. We hypothesised that the needs of 'secure attachment to others' (I. schema-domain), 'autonomy and competence' (II. schema domain) and 'freedom to express needs and emotions' (IV. domain) could have special impact on the development of postpartum anxiety because these needs are deeply activated in the mother-child relationship. 3) Finally, we hypothesised that women with clinicallevel anxiety, women with subclinical anxiety, and non-anxious mothers could be separated into distinct groups on the basis of schema-activity that would highlight the importance of paying attention not only to pathological anxiety but to subclinically anxious mothers as well. The relationship between core emotional needs, maladaptive schema domains (formed on the basis of the unfulfilment of early needs) and specific early maladaptive schemas are illustrated by

Participants
The sample consisted of 125 women who gave birth within a year. This was the only criterion for inclusion since we aimed to investigate postpartum anxiety in the general population. We created our response form on the website of the Institute of Psychology at the University of Debrecen (Hungary) for online data collection. Call for participation and the link of the online interface were posted at websites of online magazines aiming at mothers, and of perinatal foundations. We also prepared brochures including the aim of the study, information about postpartum anxiety, the link to our webpage and a contact email address. Brochures were made available at offices of health visitors and pediatricians.

Demographic data collection
Demographic data regarding the mother's age, marital status, and level of education were collected along with the number of children of the mother, and whether the last baby was planned or not.  YOUNG 1998;WELBURN et. al. 2002;in: KÖKÖNYEI 2008) consists of 75 items designed for the measurement of 15 early maladaptive schemas: Emotional Deprivation, Abandonment/Instability, Mistrust/Abuse, Social Isolation/Alienation, Defectiveness/Shame, Failure, Dependence/Incompetence, Vulnerability to Harm or Illness, Enmeshment/Undeveloped Self, Subjugation, Self-Sacrifice, Emotional Inhibition, Unrelenting Standards, Entitlement/Grandiosity, Insufficient Self-Contol/Self-Discipline. WELBURN and colleagues (2002) investigated the psychometric properties of the shortened version. Factor analysis confirmed fifteen schema subscales and their high internal consistency.

Sample characteristics
The mean age of the sample was 30.2 years (SD = 4.2, range 20-41). Most of the mothers were married (78.4%), and only 3.2% of them were single. 71.2% were university or college graduates, 21.6% finished secondary school, and 14.4% had an educational level below secondary school. 84% of the mothers' last pregnancies were planned, and 52.5% were first pregnancies. Most of the mothers were middle-class based on their demographic characteristics, but 21.6% of them revealed clinical-level, and 34.4% of them subclinical levels of anxiety; only 44% of them proved to be nonanxious.

Statistics
Using the modified version of STAI-S, three groups could be created from our sample: mothers with clinical-level anxiety (n = 27), mothers with subclinical anxiety (n = 43), and non-anxious mothers (n = 55). To investigate the relationship of early maladaptive schemas with postpartum anxiety, the sample was divided into three categories based on YSQ having low, medium or high scores. Categorisation was based in terms of all early maladaptive schemas on the scale scores, distribution of the responses and statistical sample size. We determined the categories according to 1/3-2/3 percentiles so the sample size of each category was shared by a proportion of 1/3-1/3. We chose this method instead of comparing the groups by mean ranks because in a significant portion of the sample (40-80%), schema-values were 1. Therefore the study of linear connections became meaningless, but monotone connections could be tested. This statistical dispersion indicates that categories of the schema questionnaire show qualitative and not linear quantitative differences. An important consideration in interpreting the scales is that subjects without maladaptive schemas tend to give qualitative responses (score 1, that is 'definitely not' for all items), so no gradation could be supposed on the basis of their answers. However, subjects with maladaptive schemas give responses with definitely higher scores, but linearity is less expressed in their scores. Summarily, scores of the schemas can be interpreted as typical ordinal data, while possible relationships, regarding typical ordinary variables, can be studied with Somers' D.

Direct impact of early maladaptive schemas on the level of postpartum anxiety
Thirteen early maladaptive schemas proved to have significant impact on the level of postpartum anxiety in our sample. In case of Emotional Deprivation, Abandonment / Instability, Mistrust / Abuse, Social Isolation / Alienation, Defectiveness / Shame, Failure, Dependence / Incompetence, Vulnerability to Harm or Illness, Enmeshment / Undeveloped Self, Subjugation, Emotional Inhibition, Unrelenting Standards / Hypercriticalness and Insufficient Self-Control / Self-Discipline, significant differences emerged among the groups of mothers (p < 0.05): women with clinical-level anxiety tended to give the highest, while non-anxious mothers the lowest scores on each schema. In case of Entitlement, we found significant tendency (p < 0.1), while in case of Self-Sacrifice (p = 0.39) no significant differences among the groups. The effect-pattern of early maladaptive schemas on the level of anxiety is illustrated by Table 2, where the number of x indicates the general magnitude of the frequency. xx-xxx indicates that the magnitude may vary but the pattern and its significance remains. The values 1, 2, 3 demonstrate the categories of early maladaptive schemas defined by percentiles. Score 1 represents groups with low, and score 3 groups with high schema-values.
Our results confirmed our preliminary expectations. Mothers with lower scores for certain early maladaptive schemas (1) were most likely to have normal-level worry in the first postpartum year (xxx). It was much less likely that they would experience subclinical anxiety (xx), while having clinical-level anxiety had the lowest probability (x) in their cases. Subjects having higher scores (3) for maladaptive schemas were most likely to react with clinical-level anxiety in the first postpartum year (xxx), while the least likelihood in their case was to experience normal levels of worry (x). Subjects scoring with medium average values had the highest likelihood of subclinical or normal level of anxiety in the postpartum period. These patterns are illustrated in Table 3 Subjects with low scores (1) on Dependence / Incompetence schema belonged most likely into the non-anxious group (62.3%), while they had a very small chance of experiencing clinical-level anxiety (8.6%). Those who gave high scores for the same schema were most likely to develop clinical-level anxiety (54%), while the chance of feeling only normal worry was much smaller in their case (13.5%). Subjects having medium values on the given schema were most likely to get into the subclinical (57.8%) or normal groups (36.8%).

76
EJMH 13:1, June 2018 Table 2 Effect-pattern of early maladaptive schemas on the level of postpartum anxiety

Categories of maladaptive schema-values
Level of anxiety anxious subclinical non-anxious The same pattern is demonstrated in Table 4 via Failure and on Table 5 through Abandonment schemas.

Discussion
We investigated the role of early maladaptive schemas in the etiology of postpartum anxiety. According to our findings, mothers with early maladaptive schemas are significantly more vulnerable to having clinical-level anxiety in the postpartum period than women without maladaptive schemas.
Thirteen schemas, namely Emotional Deprivation, Abandonment / Instability, Mistrust / Abuse, Social Isolation / Alienation, Defectiveness / Shame, Failure, Dependence / Incompetence, Vulnerability to Harm or Illness, Enmeshment / Undeveloped   Self, Subjugation, Emotional Inhibition, Unrelenting Standards / Hypercriticalness and Insufficient Self-Control / Self-Discipline also proved to increase mothers' vulnerability to postpartum anxiety. It is safe to assume that core psychological needs, like 'secure attachment to others', 'autonomy and competence,' 'realistic limits and self-control', 'freedom to express needs and emotions', and 'spontaneity and play' may all play a role in the etiology of postpartum anxiety. These needs are deeply activated during the psychological and caring relationship with the child. Injuries of secure attachment could be activated during motherchild bonding, while deficits of the autonomy-competence domain might be triggered by the responsibility of caring for the child. Mothers with a compromised ability to express their needs and emotions can become more vulnerable to exhaustion and solitude. Regarding the also significant schema-domain of Overvigilance and Inhibition (V.) and Impaired Limits (III.), it can be supposed that overprotective mothers set strict standards to themselves so in their judgement all faults are defeats, and they are unable to relax. This may impair their self-esteem as mothers, leading to depletion. Mothers who struggle with exercising sufficient self-control and have lower frustration tolerance can become also more vulnerable in a period when they have to cope with new tasks and intense emotions.
The results confirm our hypothesis that intrapsychic vulnerability plays a role in the complex etiology of postpartum anxiety. According to our findings, mothers with early relational injuries -including representations of self and the other -are significantly more vulnerable to postpartum anxiety. Moreover, they tend to perceive and interpret reality distorted by past experiences, so they are highly vulnerable in a period when their relationship and attachment patterns become expressly activated.
Our findings verify that clinicians should pay more attention to intrapsychic factors in the treatment and research of mothers with postpartum anxiety. Exploring early maladaptive schemas and linking those to anxiety symptoms can make mothers understand their feelings and worries and can reduce their feelings of shame, disability or being a bad mother. In addition, the role of therapeutic relationship becomes highly significant in light of our results, as the mother's (early) needs and representations -due to their activation -are very likely to emerge in the relationship. The therapist can ensure limited reparenting (within therapeutic boundaries) by offering new experiences of secure attachment, acceptance, and opportunity to freely expressing needs and emotions (YOUNG et al. 2003). These could become not only corrective experiences for the mother, but, hopefully, patterns that she may transfer into her relationship with her child.
The postpartum period provides an opportunity not only to repetition but to reparation as well. Therapeutic work with the mothers may create unique chance to process their own story, life scripts and relationship patterns to others (MERLE-FISH-MAN 2010). Furthermore, the mother's improvement serves as a primary prevention for the child through healthier attachment experiences with the mother.

Limitations
The main limitation of the study is related to the online data collection due to which the sample is not representative for the general population. The questionnaire was available only to those who had internet access; this might explain the overrepresentation of married women with higher educational level and planned children in the sample. However, a possible advantage of our study is that the results confirm the high prevalence of postpartum anxiety even in mothers whose life circumstances are good, and it also indicates that demographic risk factors cannot fully explain postpartum anxiety.
It should also be noted that our study revealed no specific schema-patterns in the background of postpartum anxiety that could be the target of the treatment. In general, the consequences of early experiences and deficits are likely to be aspecific rather than specific. Our main purpose was to investigate the possible role of intrapsychic vulnerability in postpartum anxiety what we only scratched the surface of. Further studies in larger samples are needed to uncover the role of intrapsychic factors of postpartum anxiety in depth, which will be necessary not only for providing better treatment to these mothers but for implementing preventive measures by identifying those at highest risk.