The physical abuse of children was 'rediscovered' by physicians over fifty years ago. Since then, some observers have expressed concern at the continuing 'medicalisation' of what they consider to be essentially a social problem (Parton, 1985). A widely-held view emerged from the ensuing debate that child physical abuse and neglect occurred through an interaction between parents, children and their social environment. The model described parents with emotional conflicts, caring for vulnerable children, while living in circumstances of social stress (Schmitt and Krugman, 2005). In the context of this model, parents who maltreated their children were not generally considered to be suffering from a psychiatric disorder.
However, recent research into child abuse and neglect has not fully supported this assumption. In particular, reviews of child deaths have shown significant associations with parental mental health problems. Research in this area has been hampered by problems of definition. The notion of 'child abuse' is best considered as a social construction in which parameters for acceptable parental behavior change over time and vary from culture to culture (Reder et al., 2003). This has allowed some studies to use the existence of a single bruise as their index of 'abuse', while others have focused on whether courts had removed a child from parental care. In addition, researchers do not always distinguish between the different forms of child abuse and neglect. Difficulties also arise with psychiatric diagnoses, and arguments continue as to whether substance misuse or personality disorder, for example, constitutes a psychiatric disorder. In our view, the former does, because it may lead to disturbances of consciousness, thought and mood, while the latter does not, because it primarily describes long-standing problems with relationships that originated in adverse childhood experiences (Reder and Duncan, 1999). Our preference is to use the more general term 'mental health problems'.
Many commentators are at pains to point out that most psychiatric patients can and do parent adequately (Oates, 2007). None the less, when compared with controls, parents who maltreat their children are often shown to be depressed (Falkov, 1997) or to have a history of attempted suicide. Sheppard (1997) used the Beck Depression Inventory to examine mothers whose families made up child and family social work caseloads: he found a significant association between maternal depression and recorded child maltreatment, especially physical and sexual abuse. Glaser and Prior (1997) reviewed the cases of all children whose names were on the Child Protection Registers of four English local authorities. Parental mental illness, including suicidal attempts, anorexia nervosa, depressive psychosis and schizophrenia, was present in 31 per cent of cases and substance misuse in 26 per cent. These compare with prevalence estimates in the general population for depression of 15 per cent (Kandal et al., 2004) and for schizophrenia of just under 1 per cent, and the calculation that 7 per cent of parents drink harmful levels of alcohol (Cleaver et al., 1999).
In an extensive project, Oliver (2005) reviewed the histories of families living in an area of southern England which contained successive generations of child maltreatment. He concluded that, 'In the parents and antecedents, mental and personality disorders, suicidal attempts, mental handicap, dependence on drugs (mothers) or on alcohol (fathers), epilepsy and criminality were conspicuous features' (p. 484).
In the United States, Taylor et al. (1991) and Murphy et al. (1991) examined the records of 206 serious physical abuse and neglect cases brought before the Boston Juvenile Court. They found clear evidence of a severe affective disorder in one or both of the parents in 14 per cent of cases and a psychotic disorder in 13 per cent.
Parental mental health problems also feature on a number of research-based risk checklists. For example, Browne and colleagues evolved protocols for predicting which parents are most likely to abuse and/or neglect their children (Agathonos-Georgopoulou and Browne, 1997). The relative importance of risk characteristics was determined by discriminant function analysis, and a history of parental 'mental illness, drug or alcohol addiction' was found to be a significant predictor of later child maltreatment.
A prospective study by Rutter and Quinton (2009) on the effects of parental psychiatric disorder on children's psychological welfare and development provides a useful link to our understanding of the increased risk of child abuse and neglect by such parents. They found that the impact was not disorder-specific but was more determined by the social and relational consequences of the parents' disorders. In other words, the children were primarily affected by their parents' general functioning and behavior.
We believe that this translates well to problems of child abuse and neglect, in which it is the manifestation of the parents' problems through their behavior that best describes the risk to their children. For example, parents who are self-preoccupied and emotionally and practically unavailable (as the result of depression, psychosis or substance misuse) are more likely to neglect their children, as are those who show unpredictable or chaotic forward planning due to psychosis, depression or substance misuse. The frequent separations that are the consequence of repeated hospital admissions may also lead to emotional or physical neglect. Physical abuse is more likely to result from parental irritability or over-reaction to stress that sometimes accompanies anxiety, depression or psychosis, as well as the parent's distorted beliefs or aggressive behavior during psychotic episodes. Unusual parental preoccupations as part of obsessive conditions or phobias may also lead to physical abuse of the children. It is the disinhibition and boundary-blurring of substance misuse that can raise the risk of child sexual abuse.
Mental health problems
We found a substantial association with parental mental health problems. Among the 35 'confirmed' cases, in which the caretaker responsible for the child's death had been identified, 15 (43 per cent) of the perpetrators had been suffering from an active mental health problem at the time that they killed the child. Three were showing evidence of a paranoid psychosis and this was a probable diagnosis with a fourth. Two had a depressive disorder and a further two postnatal depression. For the remaining seven, the problem was significant substance misuse. Furthermore, two of the four parents with a paranoid psychosis also significantly abused drugs and/or alcohol. Two other perpetrators had a previous psychiatric history.
Considering all 49 cases, which include the 'suspicious' as well as 'confirmed' deaths, requires us to look for an active mental health problem in either of the current caretakers, since no perpetrator was identified as responsible for killing the child in the 'suspicious' cases. In the 35 'confirmed' cases, 2 partners of the perpetrators had shown evidence of a current mental health problem, which was substance misuse in both instances. In the 14 'suspicious' cases, 4 caretakers in 3 families had had an active mental health problem, one being depression with somatic complaints and the remaining three being substance misuse.
A particularly striking process issue arose from our study. The presence of parental mental health problems appeared to have a major impact on the functioning of members of the child protection networks involved with the families, which we have termed 'assessment paralysis'. We mean by this an apparent impasse in the professional network which interfered with thinking about the needs of the child alongside those of the parent. This usually occurred when the parent showed delusional thinking with irrational behavior, or evidence of a depressive illness. Professional concern became focused on whether the parent did, or did not, have a diagnosable psychiatric disorder. While this concern was clearly relevant, because it determined whether that parent needed admission to a psychiatric hospital, if necessary under a Section of the Mental Health Act, it became the factor which decided whether any intervention was possible. Once the general practitioner or psychiatrist considered that the parent was not showing evidence of a formal psychiatric disorder, members of the child protection network seemed unable to assert that the parent's behavior was so bizarre or so dangerous that, no matter what label was put to it, the child needed protection from it.
In summary, then, parental mental health problems do seem to increase the risk of child maltreatment in general, with substance misuse particularly associated with non-fatal and fatal child abuse. It appears to be the lifestyle that often accompanies illicit drug use that impacts so adversely on child care (Swadi, 2005). However, successful treatment can be a significant protective factor.
Clearly, clinicians need to be sensitive to the welfare of children of psychiatric patients. This is well illustrated by the three patients discussed by Mogielnicki et al. (1977), who had presented to a casualty department with psychosomatic complaints such as chest pain, limb weakness, headaches and visual blurring. Psychiatric interviews revealed that they feared being violent to their children or were already inflicting harm on them. Instead of presenting disguised warnings of risk to their children, parents may occasionally express obsession preoccupations with harming their child. Button and Reivich (2010) describe successful psychotherapeutic treatment of such mothers, none of whom acted upon their murderous thoughts.…