We hypothesized that parents of child psychiatric patients would themselves show distorted psychological functioning. We tested this hypothesis by comparing the Adult Attachment Interviews (AAI) of 237 Italian parents (128 normative parents with no psychiatric involvement, 30 with a child in psychiatric care, and 79 in psychiatric care themselves). The AAIs were coded for self- and child-protective strategies, psychological trauma and unresolved loss, and signs of depression. The coders of the AAIs were blind to all information about the parents and their children. We found that parents of child psychiatric patients used more extreme strategies than the normative parents; they did not differ in strategy from parents who were psychiatric patients. With regard to trauma and loss, the parents of child patients differed from normative parents, but not adult patients, on trauma and had more unresolved losses than either normative parents or parents who were psychiatric patients. They also had more signs of depression than the other two groups of parents. We discuss the implications of these findings for treatment.

Landini A, Crittenden P, Landi G (2016) The Parents of Child Psychiatric Patients. Ann Psychiatry Ment Health 4(7): 1087.

•    Child psychiatry
•    Child psychiatric patients
•    Attachment
•    Adult attachment interview
•    DMM

AAI: Adult Attachment Interview; DMM: Dynamic-Maturational Model of Attachment and Adaptation

Individual treatment is frequently offered to children who have been diagnosed with psychosocial problems, with children’s symptoms both indicating the need for treatment and often being the focus of treatment. Instead, infants’ risk is usually defined by parents’ risk, such that early intervention with parents around parenting is widely used to prevent infants’ developmental dysfunction [1,2]. Although attachment has been a central issue in infant work, neither parenting nor attachment have been central to work with older children. In the face of this inconsistency, there have been calls for interpersonal and family level evaluation and treatment of children’s problems and even for a paradigm shift toward systemic processes in how we manage troubled children [3]. In this Study we explored whether or not parents of child psychiatric patients demonstrate the psychological characteristics indicative of adequate parenting. Such characteristics include protecting children both physically and emotionally (so that children can use all their physical and mental resources), enabling children to feel secure and comfortable (so that they can safely explore the world outside their home), and assisting children to learn to use their own minds to derive self-protective meaning from their experience (so that they can achieve independence by adulthood). Attachment theory describes individual differences in precisely these parental functions [4-6]. In addition, there is an attachment-based assessment tool, the Adult Attachment Interview (AAI) [7], that can identify parents who might place their children at risk through inadequate protection, comfort, and clarity of communication. We were interested to know whether or not parents of child psychiatric patients would use (1) attachment strategies associated with protecting and comforting their children (as opposed to themselves) and (2) patterns of information processing indicative of clarity of communication. The issues of protection and comfort are straightforward, whereas communication that promotes ‘meaning-making’ requires explanation. Through their non-verbal and verbal communication with children, parents give children information about what is meaningful and what meanings to apply. For example, some parents tell their children the truth whereas others trick children into compliance by promising treats they have no intention of giving; creating false predictions is a transformation of ‘cognitive’ information that can teach children not to believe what others say. Similarly, most parents use smiles to indicate that everything is safe and secure whereas some parents smile when they are angry or frightened; the latter transforms ‘affective’ information and teaches children that what they see can be false. In ways like these, children learn from their parents how to derive meaning from information. If the parents of child psychiatric patients showed either extreme protective strategies or transformed patterns of information processing, then there might be a need for such parents to be included in the assessment and treatment of their troubled children.

The Functioning of Parents of Troubled Children: We found only one study that evaluated the mental health of the parents of children referred for psychiatric evaluation; in that study 86% of the parents received one or more psychiatric diagnoses [8]. Notably, in the same sample [9] and another [10], the siblings of the identified child patients were found to have equally severe, albeit different, needs as the identified child patient. Similarly, siblings of children on whom there was a child protection complaint also show dysfunction, although not necessarily the same as the reported child [11,12]. These studies suggest that the problems were not located in the identified child psychiatric patient or victim of maltreatment; instead, the family as a whole appeared to be troubled. We wanted to know whether or not parents of child psychiatric patients had characteristics associated with adequate parenting. Specifically, we thought that they might have limited ability to protect and comfort their children when there was danger or the children felt unsafe and that they might transform information in ways that made it difficult for their children to make meaning of their parents’ behavior and words [5,13]. This would imply that the parents’ attachment strategies were more self-protective than child-protective prior to their children receiving psychiatric diagnoses [5,8]. Specifically, we wanted to know whether parents of children in psychiatric treatment demonstrated transformed information processing and extreme self- and child-protective strategies that might have either prevented their protecting and comforting their children or led to their endangering and confusing their children – or both. Observational assessments such as the AAI can reveal how information about danger is processed and how it disposes protective behavior for self and progeny [14].

Assessing Individual Differences in Parents’ Attachment

We used the Dynamic-Maturational Model of Attachment and Adaptation (DMM) because it focused on individual differences in the protective function of parents. Specifically, the DMM defined three basic groups (Types A, B, and C) that differed on self- and child-protective strategies [5]. The three groups were defined by both a gradient of transformations in information and also the degree of integration of cognitive and affective information, where cognitive information referred to temporal contingencies between events (e.g., the sound of a bell preceding food) and affective information referred to the intensity of contextual stimulation (e.g., an extremely loud bell sound), see figure 1. The gradient of transformations of information, represented on the vertical axis of the model, ranged from information that truly predicted danger through omitted and distorted information to falsified, denied, and delusional information. Different kinds of transformation were associated with different protective strategies, see Figure (!). The horizontal axis of the model represents the degree of integration of cognitive and affective information. The prevalence of cognitive information (that is, sequential information about contingencies leading to dangerous or safe results) or affective information (that is, contextual information about probability of safety or danger) characterizes respectively the Type A and the Type C attachment protective strategies; the Type B strategy uses a balance of cognitive and affective information whereas Type A/C alternates between unintegrated A and C strategies and transformations of information [5].

Behaviorally, the Type A strategy is an inhibitory strategy of “not doing the wrong thing” and of pleasing the attachment figure by behaving as the attachment figure desires, that is, dismissing one’s own perspective when it doesn’t match the attachment figure’s. People using a Type A strategy tend to conform to other people’s expectations; therefore they are not usually identified as being troubled. The Type C strategy is an exaggerating strategy of emphatic showing one’s perspective about what is wrong, blaming the attachment figure and putting pressure on others to protect and comfort the self. People using a Type C strategy tend to attract attention from other people, sometimes eliciting caregiving and sometimes exasperating or irritating others. The Type B strategy is a balanced strategy of clear communication about one’s state in the expectation of the availability and cooperation of the attachment figure [5].

The original attachment strategies observed by Ainsworth in infants and subsequently found in adults (A1-2, B1-5, and C1-2) [15] indicate little or no transformation of information and low risk whereas the DMM strategies (A3-8, C3-8, and A/C), observed in preschoolers and also at later ages [5,16], indicate increasingly distorted transformations and greater risk. Strategies numbered ‘3-4’ indicate that parents sometimes transform information in ways that confuse their own needs with their children’s, at ‘5-6’ parents sometimes act self-protectively rather than child-protectively, and at ‘7-8’ parents sometimes delusionally construe their child as a threat to themselves [5]. The DMM-AAI strategies have been associated with differences in prosocial brain functioning using functional magnetic resonance imaging (fMRI) [17]; specifically, mothers using a Type A attachment strategy showed significantly less empathic arousal when their infants cried than mothers using a Type B strategy. The strategies derived from the AAI also differentiated several psychiatric disorders from normative functioning and from each other. These disorders included eating disorders [18,19], anxiety [20], borderline personality disorder [21], and PTSD [22]. In addition, the AAI can identify adults’ psychological trauma and unresolved loss. These have been associated with adult psychiatric disorders in several studies [18-22]. Similarly, the AAI can identify expectation of failure (a cognitive state) and pervasive low arousal and sadness (an affective state). These states are labeled ‘depression’ in AAI attachment classifications, but DMM depression is not identical to the psychiatric diagnosis of depression. DMM depression is restricted to specified cognitive and affective states. In the case of parents of child psychiatric patients, our concern was that individual treatment of child psychiatric patients might overlook problems in their families that contributed to the children’s distress, including the unidentified need of the parents for mental health treatment for themselves [23]. Finding that the parents of child psychiatric patients use more extreme strategies, with highly transformed information, than normative parents would support the idea that psychiatric problems presented by children warrant a psychological assessment of the whole family (especially the parents).

Hypotheses

We compared the parents of child psychiatric patients to normative parents and to parents who themselves were psychiatric patients. We expected that the parents of child psychiatric patients would (1) more often use the high risk DMM strategies than the low risk Ainsworth normative strategies, (2) show more psychological trauma and unresolved loss than normative parents, and (3) show more evidence of depression than normative parents. We did not expect the parents of child patients to differ functionally from parents who were psychiatric patients, except that, (4) parents of child patients would use more often the inhibitory, conforming Type A strategies (including Type A/C) than parents who were patients, who, we expected, would use more often the more attention-eliciting, emotionally exaggerated Type C strategies.

Participants

We accessed an archive of 405 Italian Adult Attachment Interviews collected by the Family Relations Institute, Inc. over a period of two decades. From the full archive we selected all parents with normally developing children (N=128), all parents of children in psychiatric treatment (N=30), and all parents who themselves were in psychiatric treatment (N=79.) Non-parents, criminals, and psychotherapists were excluded. Informed consent was obtained from all parents.

Procedure

Psychotherapists taking the AAI course each administered two AAIs to normative adults and a third to a patient in psychotherapy. In 30 cases, the psychotherapists chose the parent of a child psychiatric patient as a normative adult. For each interviewed subject, the psychotherapist interviewer was asked to provide demographic information and a health history for the participants, including their job, if the participant was a parent, how many children she or he had, if the participant or her/his children were now or had been in the past in psychiatric treatment, if the participant was institutionalized or had a criminal record.

Assessment

The Adult Attachment Interview is a one-hour semi-structured interview that queries about adults’ childhood relationships with their parents, particularly the protective and comforting function of the parents. The coding of the interviews yields a classification of the protective strategy used by the speaker, as well as psychological trauma, unresolved loss and signs of depression [24]. The Ainsworth strategies (A1-2, B1-5, and C1-2) were considered low risk whereas the DMM strategies ranged from moderate risk (A3-6 and C3-6) to high risk (A7-8 and C7-8), including A/C combinations of these; see Figure (1).

Classification Procedure

The AAIs were transcribed verbatim and then coded for instances of dysfluency by five reliable coders without knowledge of parent or child status, who used the DMM method of discourse analysis [24]. This allowed the coders to extract information about parents’ transformations of information, leading to classification of parents’ protective strategies, psychological trauma, unresolved loss, and signs of depression [24]. These variables were categorical, so we used a Chi-square as the basic statistic, reporting effect sizes with a Cramer’s V and using the square of that as an index of the percent of variance accounted for by the hypothesis.

Hypothesis 1

The parents of child psychiatric patients differed in DMM-AAI strategy from normative parents, but not from parents in psychiatric treatment (χ²(2) = 77.90, p< .001; Cramer’s V = .57, p< .001); see figure 2. Strikingly, all the strategies were used by at least one parent, but there were no normative B, A1-2, or C1-2 classifications among either parents of child psychiatric patients or the parents in psychiatric treatment. Roughly a third of the variance in strategy was accounted for by knowing whether the parent was normative or not.

Hypothesis 2

Psychological trauma was found in 64% of parents of child psychiatric patients, 54% of adult psychiatric patients, and 9% of parents of normally developing children; the difference was statistically significant (χ²(2) = 62.77, p< .001; Cramer’s V = .52, p< .001); parent group reflected about a quarter of the variance in trauma. A significant difference in unresolved loss was found (χ²(2) = 31.33, p< .001; Cramer’s V = .36, p< .001), with 77% of parents of child psychiatric patients showing unresolved loss compared to 42% of adult psychiatric patients and 23% of normative parents. Loss typified parents of child patients and reflected about a third of the variance.

Hypothesis 3

Parents of child psychiatric patients had significantly more evidence of depression than the other two groups (χ²(2) = 17.87, p< .001; Cramer’s V = .28, p< .001). In spite of being statistically significant, parental depression accounted for only 7% of the variance.

Hypothesis 4

There was a significant association between parent referral group and the A, B, C, A/C strategies χ²(6) = 50.37, p< .001; Cramer’s V = .33, p< .001), but percent of variance accounted for was small (~9%).Only the normative parents used Type B; parents of child patients were biased toward Type A3+ (including A/C), and parents who were psychiatric patients were biased toward Type C.

The parents of child psychiatric patients appear to be very similar in psychological functioning and attachment to parents who themselves are in psychiatric treatment and both groups seem very different from normative parents. Figure (2) shows that the distribution curves for the two clinical groups are the inverse of that for the normative group.

Only the use of a Type A strategy (sometimes within an A/C strategy) and unresolved loss stand out as typifying parents of child psychiatric patients more than parents who are in psychiatric treatment themselves. We note that A1-2 appears to be very infrequent in the Italian population (whereas C7-8 was not expected in any of these three samples because the criminal population was excluded.) The A/C strategy deserves an explanation; it can vary from A1/C1 to A8/C8. A look at these data showed that normative parents who used an A/C strategy tended to use A1-3/C1-4 strategies whereas the parents in the two clinical groups used A3-7/C3-6 strategies, i.e., there was a difference within A/C that is not displayed in Figure (2).

Implications for children

For children, coping with a parent who changed strategies frequently (A/C) and who was psychologically engaged with a deceased person must be very difficult; this challenge might explain some of the maladaptive behavior of child patients. The strong effect sizes indicate that, if replicated on a larger sample, these findings might apply to the great majority of parents of child psychiatric patients. Moreover, the use of strategies with substantial distortion of information (the ones used by the two clinical groups) implies that, when there was a conflict of interest between parent and child, the parents might use a self-protective, rather than a child-protective, strategy. In cases where the strategy number was 7-8, the parents might even treat the child as a source of threat.

Our findings strongly suggest that the parents of child patients who were considered normative by psychotherapists functioned in ways that were not likely to foster children’s normative development. This would imply that both children in psychiatric treatment and their parents would benefit from a family psychological assessment, psychiatric diagnosis of both parents and children, and treatment directed toward family needs. Notably, the inhibitory Type A strategies of the parents of child psychiatric patients (including those with an A/C strategy) are likely to pass unnoticed by professionals as compared to the exhibitionistic Type C strategies of parents who have sought psychiatric treatment for themselves. We would encourage professionals dealing with children in treatment to find ways to explore the parents’ self-protective strategies, psychological trauma, and unresolved losses – because these may affect the children’s well-being.

We are appreciative of the willingness of the parents to be interviewed and contribute to this research. We thank the coders of the AAIs who volunteered their efforts and the Family Relations Institute for allowing us to use their archive of data.

1. Kelly JF, Zuckerman T, Rosenblatt S. Promoting first relationships: A relationship-focused early intervention approach. Infants & Young Children. 2008; 21: 285-295.

2. Letourneau N, Tryphonopoulos P, Giesbrecht G, Dennis C, Bhogal S, Watson B. Narrative and Meta-Analytic Review of Interventions Aiming to Improve Maternal-Child Attachment Security. Infant Ment Health J. 2015; 36: 366-387.

3. Division of Clinical Psychology, British Psychological Society. Division of Clinical Psychology position statement on the classification of behaviour and experience in relation to functional psychiatric diagnoses: time for a paradigm shift. 2013.

4. Bowlby, J. Attachment and Loss (2nd ed). New York: Basic. 1982.

5. Crittenden PM. Raising parents: Attachment, representation, and treatment, 2nd edition. London: Routledge. 2015.

6. Crittenden PM, Ainsworth. Child maltreatment and attachment theory, in D. Cicchetti & V. Carlson (eds), Handbook of Child Maltreatment. New York: Cambridge University Press. 1989.

7. George C, Kaplan N, Main M. The Adult Attachment Interview: Interview Protocol. Unpublished manuscript, University of California, Berkeley. 1996.

8. Crittenden P, Kozlowska K, Landini A. Assessing attachment in school-age children. Clin Child Psychol Psychiatry. 2010; 15: 185-208.

9. Kozlowska K, Elliott B. Don ‘t forget the siblings: School-aged siblings of children presenting to mental health services show at risk patterns of attachment. Clin Child PsycholPsychi. 2016.

10. Ma N, Roberts R, Winefield H, Furber G. The prevalence of psychopathology in siblings of children with mental health problems: A 20-year systematic review. Child Psychi Human Devel, 2015; 46: 130-149.

11. Jean Gilles M, Crittenden PM. Maltreating families: A look at siblings. Family Relations. 1990; 39: 323-329.

12. Hamilton-Giachritsis CE, Browne KD. A retrospective study of risk to siblings in abusing families. J Fam Psychol. 2005; 19: 619-624.

13. Crittenden PM, Landini A, Kozlowska K. The effect of parents psychiatric disorder on children’s attachment: Theory and cases. Reupert D. Maybery, J. Nicholson M. Seeman M, Gpfert (Eds.). Parental Psychiatric Disorder: Distressed Parents and their Families, 3rd edition. 2015; 29-41.

14. Farnfield S, Hautamäki A, Nørbech P, Sahhar N. DMM assessments of attachment and adaptation: Procedures, validity and utility. Clin Child Psychol Psychiatry. 2010; 15: 313-328.

15. Ainsworth MDS, Blehar MS, Waters E, Wall S. Patterns of Attachment: A Psychological Study of the Strange Situation. 1978.

16. Crittenden PM. Quality of attachment in the preschool years. Development and Psychopathology. 1992; 4: 209-241.

17. Strathearn L, Fonagy P, Amico J, Montague PR. Adult attachment predicts maternal brain and oxytocin response to infant cues. Neuropsychopharmacology. 2009; 34: 2655-2666.

18. Ringer F, Crittenden PM. Eating disorders and attachment: the effects of hidden family processes on eating disorders. Eur Eat Disord Rev. 2007; 15: 119-130.

19. Zachrisson HD, Kulbotton G. Attachment in anorexia nervosa: An exploration of associations with eating disorder psychopathology and psychiatric symptoms. Eating & weight disorders. 2006; 11: 163-170.

20. Hughes J, Hardy G, Kendrick D. Assessing adult attachment status with clinically-orientated interviews: a brief report. Br J Med Psychol. 2000; 73: 279-283.

21. Crittenden PM, Newman L. Comparing models of borderline personality disorder: Mothers ‘ experience, self-protective strategies, and dispositional representations. Clin Child Psychol Psychiatry. 2010; 15: 433-451.

22. Crittenden PM, Heller MB. The roots of chronic PTSD: Childhood trauma, information processing, and self-protective strategies. Chronic Stress. In press.

23. Crittenden PM, Dallos R. All in the family. Clin Child Psychol Psychi. 2009; 14: 387-407.

24. Crittenden PM, Landini A. Assessing Adult Attachment: A Dynamic-Maturational Approach to Discourse Analysis. 2011.