Research has shown that maternal mental illness can affect mother–infant interactions with implications for infant outcomes. Severe and chronic mental illness (SMI), particularly schizophrenia, is associated with the greatest risk. Schizophrenia is also associated with impairments in attribution of mental states, ‘theory of mind’ (ToM). Recent attachment research has suggested that maternal mentalizing skills are strongly associated with attachment outcome in infants. To date, no research has explored the relationship between ToM and maternal sensitivity in mothers with SMI using standard tests of ToM. The present study was designed as an exploratory study in order to investigate this.

A total of 40 women with SMI in the postpartum period were administered a battery of ToM tasks and general neuropsychological tasks. The women were also filmed in an unstructured play session with their infants, which was coded for maternal sensitivity using the Crittenden CARE-Index.

One ToM task, the Frith–Happé Animations, predicted maternal sensitivity across all diagnoses. There was also an effect of diagnosis, with lower sensitivity observed in women with schizophrenia. ToM impairments did not fully explain the effect of diagnosis on sensitivity. Mothers of girls were rated as being more sensitive than mothers of boys.

The results suggest that ToM is a significant predictor of maternal sensitivity across all mental health diagnoses, extending the results of studies focusing on healthy populations. Clinical interventions emphasizing the importance of understanding the perspective of the infant may enhance maternal sensitivity.

Depression in mothers during pregnancy and in the postnatal period has been recognized to have wide-ranging adverse impacts on offspring. Our study examines some of the outcomes and long-term economic implications experienced by offspring who have been exposed to perinatal depression.

We analysed the effects of perinatal depression on child development outcomes of children at ages 11 and 16 years from the community-based South London Child Development Study. Economic consequences were attached to those outcomes through simple decision-analytic techniques, building on evidence from studies of epidemiology, health-related quality of life, public sector costs and employment. The economic analysis takes a life-course perspective from the viewpoints of the public sector, individual and society.

Additional risks that children exposed to perinatal depression develop emotional, behavioural or cognitive problems ranged from 5% to 21%. In addition, there was a high risk (24%) that children would have special educational needs. We present results in the form of cost consequences attached to adverse child outcomes. For each child exposed to perinatal depression, public sector costs exceeded £3030, costs due to reduced earnings were £1400 and health-related quality of life loss was valued at £3760.

Action to prevent or treat mothers' depression during pregnancy and after birth is likely to reduce public sector costs, increase earnings and improve quality of life for children who were exposed to the condition.

Maternal experience of childhood maltreatment and maternal antenatal depressionare both associated with offspring childhood maltreatment and offspring adjustmentproblems. We have investigated the relative impact of maternal childhoodmaltreatment and exposure to depression in utero on offspringmaltreatment and psychopathology.

The sample included 125 families from the South London Child Development Study. Aprospective longitudinal design was used. Data on maternal childhood maltreatment,maternal antenatal depression (36 weeks of pregnancy), offspring childhoodmaltreatment (age 11 years) and offspring adolescent antisocial behaviour anddepression (ages 11 and 16 years) were obtained from parents and offspring throughclinical interview.

Mothers who experienced childhood maltreatment were significantly more likely tobe depressed during pregnancy [odds ratio (OR) 10.00]. Offspring of mothers whoexperienced only childhood maltreatment or only antenatal depression were no moreat risk of being maltreated or having psychopathology; however, offspring ofmothers who experienced both maternal childhood maltreatment and antenataldepression were exposed to significantly greater levels of childhood maltreatmentand exhibited significantly higher levels of adolescent antisocial behaviourcompared with offspring not so exposed. Furthermore, maternal childhoodmaltreatment accounted for a significant proportion of the variance in offspringchildhood maltreatment in only those offspring exposed to depression in utero.

Maternal childhood maltreatment and maternal antenatal depression are highlyassociated. The co-occurrence of both insults significantly increases the risk ofoffspring adversity. The antenatal period is an optimum period to identifyvulnerable women and to provide interventions.

Antenatal depression and childhood maltreatment have each been associated with offspring psychopathology, but have never been examined in the same sample.

To determine whether childhood maltreatment influences the association between antenatal depression and offspring psychopathology.

Prospectively collected data on antenatal depression, offspring maltreatment (age 11) and offspring psychopathology (age 11 and 16) were analysed in 120 mother–offspring dyads from the community-based South London Child Development Study.

Antenatal depression increased the risk of maltreatment in the offspring by almost four times. Children exposed only to antenatal depression or only to childhood maltreatment were no more at risk of developing psychopathology; however, children exposed to both antenatal depression and childhood maltreatment were at almost 12 times greater risk of developing psychopathology than offspring not so exposed.

Research investigating exposure to adverse events inutero and offspring psychopathology should take account of postnatal adverse events such as maltreatment.

Previous cross-diagnosis studies of interaction between mothers with severe mental illness and their babies have concluded that mothers with schizophrenia have deficits in interaction, but these studies have not included healthy controls.

In-patients on a mother and baby unit, with diagnoses of schizophrenia (n=15), depressive mood disorders with or without psychosis (n=23), or mood disorders where mania was the predominant feature, with or without psychosis (n=12), were observed interacting with their infants on admission and discharge. Mothers' mind-mindedness and other measures of the quality of maternal and infant behaviour were coded. Findings from this sample were compared with those from healthy mothers and their infants (n=49).

Compared with healthy controls, on admission depressed mothers were marginally less likely to comment appropriately on their infants' mental states. Both the depressed and mania groups were more likely to touch their babies and engage in attention-seeking behaviours. Interactional behaviours of mothers in the schizophrenia group were not markedly different from healthy controls. On discharge there were fewer differences between the clinical and healthy groups, although the depressed group continued to engage in more attention-seeking and touching behaviour and the mania group continued to touch their infants more. Only mothers in the schizophrenia group showed changes in interactional behaviours between admission and discharge, talking more to their infants.

The findings challenge previous conclusions that mothers with schizophrenia have deficits in their interactions with their babies, and demonstrate that mothers with severe mental illness are able to respond appropriately to their infants' cues.