Epidemiologists and psychoanalysts have been equally concerned about the intergenera‐tional concordance of disturbed patterns of attachment. Mary Main's introduction of the Adult Attachment Interview (AAI) has provided the field with an empirical tool for examining the concordance of parental and infant attachment patterns. In the context of a prospective study of the influence of parental patterns of attachment assessed before the birth of the first child upon the child's pattern of attachment to that parent at 1 year and at 18 months, the Anna Freud Centre—University College London Parent‐Child Project reported a significant level of concordance between parental security and the infant's security with that parent. In the context of this study, a new measure, aiming to assess the parent's capacity for understanding mental states, was developed and is reported on in this paper. The rating of Reflective‐Self Function, based upon AAI transcripts, correlated significantly with infant security classification based on Strange Situation assessments. The philosophical background and clinical importance of the measure are discussed.

This paper presents a summary of the Anna Freud Centre-University College London, Parent-Child Project. Its most important finding was that the security of the infants' relationship with both parents at 12 and 18 months could be predicted on the basis of qualitative aspects of the parents' accounts of their own childhoods collected before the birth of the child. This confirmed Selma Fraiberg's observations concerning the reemergence of childhood conflicts at early stages of childbearing. Possible mechanisms mediating this link are explored with particular reference to the role of the parents' accurate mental representations of the infants' mental world.

In this paper we describe two models of the psychoanalytic treatment of mental disturbance. The first describes the mechanism by which the patient is helped to recover threatening ideas and feelings which have been repudiated or distorted as a result of conflict and defense. The second points to the therapeutic effects of engaging previously inhibited mental processes within the psychoanalytic encounter. The two forms of therapeutic action imply two distinct means available to the individual to deal with psychological conflict. They highlight different aspects of the psychoanalytic process and technique in child and adult psychoanalysis.

SUMMARY Glycosylated haemoglobin concentrations, C peptide secretion, insulin dose, psychiatric state, intellectual functioning, and the extent to which the implementation of the diabetic regimen was shared between parent and child were studied in a cross sectional study of 50 children with diabetes aged 6-16. Indications of psychological disturbance in the children and their parents predicted low glycosylated haemoglobin concentrations in the children, and accounted for 44% of variance in blood glucose control. The child's early and independent participation in the implementation of the diabetic regimen was associated with poor control. The association between psychiatric disorder and poor blood glucose control has been explained by poor adherence to the diabetic regimen among psychiatrically disturbed patients.8 9 We have examined the association between blood glucose control (measured by HbA1 assay), psychological adjustment, and patterns of diabetic management-;in a sample of children for whom C peptide measureents and other clinical variables were available. Subjects and methodsThe study comprised all 71 diabetic children (aged 6 to 16 years) attending the outpatient department of a rural district general hospital. Families of patients who met the selection criteria were approached individually. The criteria were duration of diabetes of at least six months, absence of major medical problems unrelated to diabetes, and absence of major intellectual impairment. Of the 61 patients meeting the criteria, 50 (85%) agreed to participate in response to a letter from the consultant. Table 1 gives details of the children. There were slightly more over the age of 12, and the median duration of illness was 3*5 years. Social class was determined using the Registrar General's classification; the sample showed a slight bias towards middle class families. Fifteen per cent of the children came from broken homes. The 11 families who did not participate were comparable in sex ratio, social class, and age with the children studied, and 18%were single parent families.The questionnaires were completed and psycholo-

No abstract