Previous research has shown that infants whose mothers suffer from mental disorder are at increased risk for developmental delays, cognitive and functional impairments, physical symptoms and injuries, as well as behavioral and emotional problems in pre-school and school age
[1–6]. Affective and mood disorders, anxiety disorders, posttraumatic stress disorders, obsessive-compulsive disorders, and psychosis with elevated incidences of comorbid diagnoses, varying in onset, course and prognosis, have a higher prevalence in mothers, both postpartum and during early child-age years
[7–17]. In clinical populations, co-occurring personality disorders have to be considered as a crucial part of the global risk that mentally ill mothers experience
The quality of parenting has been viewed as one key mediator of the relationship between maternal psychopathology and the outcome for infants and young children. Maternal psychological and vegetative symptoms, such as social withdrawal, anxiety, emotional lability, impulsivity, and severe exhaustion or weariness, have been found to impinge on the interaction and the development of the relationship between mother and infant. Often, such mothers are emotionally, cognitively and/or behaviorally inhibited or impaired in their ability to recognize and react with appropriate “sensitivity” and “responsiveness”
 to their children’s needs. In regard to the infants of these mothers, higher incidences of behaviors, such as persistent crying, motoric restlessness, averted gaze or head position, physical neediness and a lack of expression of delight in the presence of their mothers, have been observed
[23–25]. These behaviors increase the mother’s experience of stress, which further contributes to the maintenance of maladaptive interactional behavioral patterns
[26–32]. Without improvement in these negative interaction cycles, for example, through early interventions, there is a greater likelihood for infants to develop an insecure-avoidant or an insecure-ambivalent or disorganized attachment strategy during the first year of life
[33, 34]. According to Ainsworth et al.
, in the Strange Situation, insecure-avoidant (Category A) infants do not actively seek proximity and physical contact to their mothers after separation, instead favoring exploration. Insecure-ambivalent (Category C) infants, on the other hand, tend to cling to their caregivers and are characterized by frantic appeals to establish and maintain close proximity to them. At the same time, they show anger and sulky or aggressive behavior towards their mothers. These children are difficult to comfort after separation, that is, they maximize their attachment neediness
. Secure infants (Category B) reach out to their mothers at times of separation distress and calm down easily when comforted so that they can resume play or exploration. If the context of care is additionally characterized by either fear-inducing or frightening parenting practices (e.g., abuse, neglect, aggressive behaviors) or by fearful behaviors on the part of the mother (e.g., signs of anxiety or avoidance, dissociation, etc.), the child is more likely to develop disorganized or disoriented behavioral patterns (Category D)
[35, 36]. These include contradictory behavioral tendencies, such as fearful/anxious vacillation between exploring and seeking closeness, temporally uncoordinated or slowed movements and occasional physical or mimic paralysis or “freezing” of the child toward the mother. A maternal mental illness is associated with the development of insecure organized (avoidant, ambivalent) and disorganized attachment styles in the child
[37–41]. Social-emotional maladjustment and related mental health problems in the child are common consequences
[42, 43]. Furthermore, mothers with mental disorders themselves are frequently characterized as insecure (dismissing Category Ds or preoccupied Category E), or unresolved (Category U) rather than autonomous (Category F) when attachment representations of their own childhood experiences with attachment figures are assessed
[44–47]. These mothers are troubled by the physical and affective states, needs and behaviors of their infants, which may correspond to “frightening and frightened”
 and/or “disrupted”
 parenting behaviors
According to the model of transgenerational transmission
, maternal attachment representations influence the child’s attachment more than maternal sensitivity does
[52–54]. Further clinical research on this issue provides evidence that the maternal capacity for mentalization appears to be more crucial than the expressed sensitivity of mothers in the attachment relationship
. The mental abilities of the mother to perceive and recognize her own and the child’s wishes, motives, needs, thoughts and feelings within the context of the attachment relationship and to communicate, as well as to reciprocate these through active engagement, kinesthetic expressions, words and play, represent key competencies for the development of secure attachment, self-regulation and mental health in the child – even if the mother had experienced unprocessed adverse experiences during her own childhood
. Conversely, a diminished capacity to mentalize brought about, for example, by maternal psychopathology, increases the likelihood of the child developing an insecure or disorganized attachment style, unless this process can be changed by an attachment-based intervention.
Research on the treatment of women with postpartum depression has shown that treatment of only the mother (i.e., medication, individual psychotherapy) is not sufficient to buffer against the negative impact of psychopathology on the child’s cognitive and psychosocial development, as well as attachment
[57–61]. Rather, there is a necessity to support mentally ill mothers in their specific needs in caring and relating to their infants. Attachment-based interventions rooted in empirical research on developmental psychopathology are, thus, promising approaches to address these multi-faceted treatment targets. However, to date, very few attachment-based interventions have been systematically tested in formal clinical settings using a randomized controlled trial (RCT) framework involving women with complex postpartum psychiatric disorders.
Attachment-based interventions are designed to promote maternal sensitivity, to change maternal mental representations, to promote attachment security in the child, and/or to support the family, examples include STEEP™
; Wait, Watch, and Wonder
; Circle of Security
; Video Intervention To Promote Positive Parenting
. Each of these interventions has been conducted in different settings (home-based, institutional, individually, group-based), with various at-risk populations (i.e., low SES, depressed mothers, adolescent mothers, preterm infants), and at a wide range of dosages and intensities. Video-analysis is a typical technique implemented as part of these interventions in order to facilitate change. Meta-analytic evidence identifies short-term approaches (< 16 sessions), targeting maternal sensitivity as being the most effective, and sensitivity-focused interventions conducted with clinically referred samples (i.e., DSM-III-R depressed mothers), as being more effective than interventions with other groups
. Attachment security, in particular, has been found to be readily influenced by sensitivity-focused interventions
. Yet, it remains unknown how brief attachment-focused interventions designed for at-risk populations differentially impact mothers’ mentalizing, mental representations of attachment, beyond sensitivity and mental illness, and children’s level of attachment security.
The Circle of Security intervention is a brief, behavioral and insight-oriented therapeutic group approach for promoting attachment and autonomy in the parent–child relationship. It combines psycho-educational, cognitive-behavioral, and psychodynamic understanding and intervention techniques. Given that women with various and co-occurring mental illnesses after childbirth differ greatly in their requirements, openness and motivation for, and compliance with mother-infant treatment, the COS intervention can allow individualized, flexible and deepened therapeutic access to each mother-infant dyad, which is further supported by its group character. Therefore, within our clinical context, COS seems to be a promising intervention to detect and treat early difficulties in the development of secure attachment relationships.
The current RCT is designed to evaluate the efficacy of the COS intervention for mentally ill mothers with infants, for the first time in Germany and in a clinical context. The main research question is whether COS, in comparison with treatment as usual (TAU), increases attachment security and prevents the development of insecure and/or disorganized attachment by promoting maternal sensitivity and mentalizing measured after treatment at follow-up when the children are aged 16–18 months. We hypothesize that following treatment there will be a higher proportion of secure child–mother attachment in the COS arm compared to TAU. This paper describes the design of the trial, the implementation of the study protocol in a child and adolescent psychiatric outpatient unit, and the data analysis strategies.