Growing Up In Scotland: Maternal mental health and its impact on child behaviour and development

This document reports how many mothers in Scotland experience poor mental health in the first four years of their child’s life, as well as the characteristics of these women. It further shows the impacts of maternal mental health on child development and behaviour.


4.1 The key findings

The longitudinal Growing Up in Scotland ( GUS) survey was used to explore whether being exposed to a mother with poor mental health affected children's development at age four. The mental health of a random sample of over 3000 mothers in Scotland was assessed at four points in time between 10 and 46 months after the study child was born. We distinguished between mothers who were emotionally well throughout the survey period (that is, were emotionally well at every interview), those who had scores suggesting mental health problems on one occasion only and, finally, those who appeared to have repeated mental health problems.

At any one point, between 12 to 16% of these mothers were deemed to be experiencing poor mental health. However, overall, almost a third were defined as having poor mental health at some point in the first four years of the cohort child's life: 17% had a brief episode of poor mental health, but a similar proportion (14%) were found to have been in poorer mental health at more than one survey sweep.

Maternal mental health was closely associated with women's socio-economic conditions and the quality of their inter-personal relationships. Although children's social, emotional, behavioural and cognitive development was associated with social conditions, it was also linked with maternal characteristics such as age, occupational status and educational attainment. Importantly however, maternal mental health was found to exert a significant independent effect on key child outcomes. For all of the child outcomes included in the analyses, it was clear that children with mothers who were free from mental health problems were more likely to have more positive outcomes and that those children whose mothers had persistent mental health problems had the poorest outcomes. Children who had more prolonged exposure to a mother with mental health problems were more likely to have an adverse developmental outcome. This possible "dose response" relationship (the higher the "dose" the greater the impact) between maternal mental health and child outcomes indicates that the relationship may be causal; that is the mother's mental health problems are a causal factor in the child's poorer outcomes.

4.2 The findings in context

It has been suggested that there is a complex causal association between the quality of mothering and a range of emotional, behavioural, cognitive and social outcomes for children. Murray (1992, 1996) has postulated that mothers suffering from depression may have weaker attachment to the baby and may, as a result, be less attuned and emotionally responsive to the baby and that this may explain deficits in their child's development.

Much of the evidence linking maternal depression to longer-term outcomes for children has focused on relatively early and often severe postpartum depressive disturbance and rather less on depression beyond the first few weeks or months following a birth. While there is a body of observational research, much of the research evidence exploring links between maternal mental health and child outcomes has relied on retrospective analyses. The longitudinal Growing Up in Scotland ( GUS) study provides an opportunity to explore links over time, based on measures obtained at discrete points in a child's life. Thus, measures of maternal mental health when the baby is almost a year old can be linked to later measures of child development which are not contaminated by the mother's mental health at an earlier phase of the survey. They may, of course, be influenced by her current mental health status and the effect that might have on our findings is considered below.

4.3 Measurement issues: reliability and validity

In the absence of standardised clinical assessment, the definition and identification of mental ill-health is problematic. There is a range of measures and instruments which can be used by non-clinicians in non-clinical settings. The choice of assessment tool has to be valid and reliable, but also has to be appropriate for the population and the mode of administration. The instruments used within the GUS survey interview had to compete with a large number of questions in a time-limited context and had to be acceptable to the respondents. Two different instruments were used and, in hindsight, it might have been preferable for one consistent measure to have been used throughout. However, both instruments are well-validated and there was a high correlation within our sample between the measures.

As we noted (see Appendix A), neither of the instruments used (the SF12 ( MCS) nor DASS) have threshold scores which define whether an individual's score is indicative of a psychiatric disorder. Such stringent criteria are not necessarily required. This analysis used within-cohort scores to define those with poorer mental health in relation to other mothers rather than defining these women as suffering from a mental illness which meets The Diagnostic and Statistical Manual of Mental Disorders ( DSM) criteria.

We have emphasised on several occasions that we cannot assume that women with repeated poorer mental health were necessarily suffering from a chronic mental health problem. Although it is quite likely that some of the women who had scores indicative
of poorer mental health on more than one occasion were experiencing long standing difficulties, we would urge caution in over-extrapolation. However, it was clear that children exposed to repeated maternal mental health difficulties had poorer outcomes than those exposed briefly who, in turn, had poorer outcomes than those whose mothers were emotionally well at each survey sweep.

4.4 Artefact and attribution

Depressed or anxious mothers may have a negative perception of themselves, their relationships and their children and this could be reflected in their responses to the survey questions about their child. We cannot preclude that this may have occurred in some cases, particularly in relation to reports of behavioural, social and emotional outcomes. However, the longitudinal research design reduces the chance that the information provided by mothers about child outcomes was contaminated by her responses to mental health questions at an earlier time. The use of well-established, standardised and validated instruments, suitable for administration by interviewers, to assess child development also ameliorated potential biases associated with negative maternal affect.

There was suggestive evidence that women in conflictual relationships where the couple either did not cope well in the early months or showed signs of difficulty at a later point were more likely to be experiencing poorer mental health and for there to be an independent effect of these relationship tensions on children. Again, we cannot be confident that the poor relationships are not themselves a consequence rather than a cause of maternal mental state. However, there is a plethora of evidence suggesting that women who have unsupportive relationships are more vulnerable to depression, particularly in the context of stressful life events or financial hardship (Brown and Harris, 1978).

4.5 Conclusions and implications

Maternal mental health was associated with socio-economic disadvantage, impoverished interpersonal relationships and with poor social support. It was clear that maternal mental health problems at Sweep 1 when the cohort baby was 10 months old and therefore beyond the immediate postnatal period usually associated with depression, were common. Previous research has also observed the high incidence of depression beyond the early postnatal months (Martin et al., 1989). Moreover, it was apparent that mental health problems among the GUS cohort mothers at Sweep 1 were often associated with further episodes of poor mental health. Indeed, it was unusual for a mother to have a repeated mental health problem which began (or was first recorded by the survey interview) after the first year. Exposure to a mother with mental health problems, in turn, was found to be significantly associated with negative impacts on their children with prolonged or repeated exposure having a greater impact than brief exposure.

The study children were almost four years old at the time of Sweep 4 and most were attending pre-school. At this early stage, there was evidence already of clear deficits in relation to their emotional, social and behavioural development linked to their mothers' emotional well-being. These deficits will undoubtedly shape their pre-school and subsequently their early school experiences. It may influence how well they integrate with other children and how they may be perceived by the adults they come into contact with at these establishments.

There were significant associations between brief and repeated exposure to maternal mental ill-health and child outcomes, but the relationship was most marked for children who experienced repeated exposure. The evidence that repeated exposure to maternal mental ill-health was associated with a greater likelihood of an adverse outcome suggests that the effects may be causal. It was not possible to explore how or why maternal mental health impacted on child outcomes, but we might postulate that deficits in attachment may play a role by disrupting the mother-child relationship and diminishing the quality of their interactions (Murray et al., 1996).

Monitoring maternal mental health beyond the first months after a birth could facilitate early interventions to support mothers and prevent or ameliorate further mental health problems. This may also have a direct impact on children's development and enhance their social and educational experiences.

Back to top