Depression in children has only recently become a focus of mental health research. There has been relatively little previous research on the origins and causes of depression in young children and youth. Much of the existing research has questioned whether depression is a real condition experienced by children. It has been argued that childhood depression might be an aspect of childhood development.
This study is concerned with the social and structural bases of depression in children and examines why some children are depressed and others not. In particular, what contribution do poverty, parental marital breakdown, maternal alcohol abuse and peer relationships make to childhood depression? Is the age of onset a factor in the severity of the condition? We have examined the symptoms as reported throughout the child's developing years and into adolescence.
This study is part of the Mater-University Study of Pregnancy (MUSP) which began in 1981 (Keeping et al., 1989). Some 8,556 mothers were enrolled at their first prenatal visit at a large public hospital and followed up at the birth of their child, and again six months later. Women giving up their child for adoption and multiple births were excluded (7,421 mothers at the commencement of follow-ups). The mother and child were again assessed when the child was 5 and then at 14 years. Sixty-nine per cent (5,227 mothers and 5,262 children) of the birth cohort completed the 14-year follow-up. The mother completed a questionnaire about herself, her partner and her child at each follow-up. At the 14-year follow-up the child completed a questionnaire about himself/herself and was administered a cognitive and physical assessment. The measures used were the Child Behaviour Check List (CBCL) completed by the mothers and the Youth Self-Report form (YSR) completed by the child at 14 years.
The frequency of the child's symptoms reported by the mother varied significantly in many cases from those reported by the child. This raises fundamental questions about the meaning of surrogate (maternal) reports on childhood depression.
The predictor variables were aggregated under six categories: sociodemographic, prenatal and perinatal; maternal mental health; maternal lifestyle; mother's attitude to child and child-rearing; and cognitive factors. The variance in depression levels attributed to the child and statistically associated with each of these six classes was compared.
Demographic and perinatal variables as well as maternal mental health were found to be significant predictors of childhood depression. Where the mother reported a level of mental health impairment, her child was more than three times as likely to be depressed at age 5. Variables from the different classes were found to be intercorrelated. Logistic regression analyses were used to allow the independent effects of the individual variables to be evaluated.
Children are observed to experience depression, even from the age of 5 years. Mothers are consistent in attributing (or not) depression to their children. Socio-structural variables are only modest predictors of depression attributed to children. These results are consistent with the view that depression may have an early onset and high level of continuity for perhaps half of those children who experience depression. For the other half of children experiencing depression, late onset appears to be the more characteristic pattern. These findings must raise the question of whether, for a substantial minority of the population experiencing depression, it is a biological process with early onset and a lifelong but intermittent course.