Copyright Sven Silburn and Stephen Zubrick, 1996. One copy of this paper can be made for the purpose of personal, non-commercial use, subject to proper attribution to the authors .

The WA Child Health Survey: Methodology and Policy Implications

Sven R Silburn and Stephen R Zubrick

Division of Psychosocial Research
TVW Telethon Institute for Child Health Research
(University of WA & Princess Margaret Hospital for Children)
PO Box 855, West Perth, WA 6872
Telephone: (09) 340-8474
Fax: (09) 388-3414

Paper presented at

Fifth Australian Family Research Conference
'Family Research: Pathways to Policy'

27-29 November 1996

As we set about preparing children and adolescents for the next century we face new challenges which demand serious questioning of how families, schools and the other key socialising institutions of our society are currently functioning. Increases of common behaviour problems, such as conduct disorder & ADHD which have their onset early in childhood, have been reported in most western industrialised nations over the past few decades. Similar increases have been reported for other problems that rise in frequency or tend to peak in the teenage years including adolescent depression and suicidal behaviours, eating disorders, abuse of alcohol and drugs and juvenile crime (Rutter & Smith, 1995). These trends highlight the rapidly changing context of child-rearing and re-focus attention on the importance of adolescence as a critical period when behavioural patterns, once established, can significantly influence long term health status.

While adolescence is (and probably has always been) a challenging period of major biological and social change, it is not usually problematic unless societal conditions prompt it to be so. Most societies, including Australia, have been slow to recognise the increased challenge which our complex societies pose to healthy adolescent development and this has further increased the likelihood that adolescence will be a difficult period. The higher rates of psycho-social problems among adolescents deserve urgent attention as these behaviours appear to be compromising the future life-chances of an increasing proportion of our youth. They should also be of concern if we are to learn anything from the recent experience of the United States where similar trends to those we are now seeing in Australia were evident several years ago.

So how do we account for the rise of such problems among our young? All of the aforementioned difficulties are known to be associated with social disadvantage. We might therefore expect that with the improvements in material well-being which have occurred since World War II, the frequency of mental health problems would have decreased as living conditions have improved. After all, physical health has improved in step with better living conditions, as seen in steadily falling infant mortality and a steadily increasing life expectancy. However, mental health problems affecting children, adolescents and young people have shown no such fall in frequency, and indeed have mostly become substantially more prevalent.

The increased proportion of children exposed to early life stresses such as child abuse and neglect has been suggested as a possible reason for the rise in adolescent mental health problems. Children who enter their teenage years already vulnerable psychologically or socially are likely to experience a more difficult adolescence and this may well account for some of the increase in adolescent disorders. However, the influence of broader social stresses which impact differentially on adolescents and their families is suggested by the fact that many of the difficulties and alienation of young people are not confined to the disadvantaged and marginalised. Until quite recently our public and political debate on youth issues has largely focused on disadvantaged youth. While this is understandable, it risks encouraging the misconception that these issues do not reflect problems within mainstream society.

Within our own lifetime we have seen major changes in the role and function of the family, of neighbourhoods, organised religions and the media and in how they each contribute to the transmission of culture, the development of competencies, and the shaping of young people's expectations, attitudes and values. In the course of a single generation, the relative influence of each of these institutions has changed quite radically. Parents of many of today's adolescents feel beleaguered by these changes and often blame themselves for what are all too often cultural problems. At the same time, the pace of technological, economic and social change has been so rapid that has been difficult for society to pause for long enough to take stock and evaluate what these changes may mean for the health and development of children and adolescents.

The need to benchmark the nature and extent of the problems affecting youth today and to obtain a better understand the changing context of child rearing in the 1990s were the primary motivations for the Western Australian Child Health Survey which was conducted by the TVW Telethon Institute for Child Health Research in conjunction with the Australian Bureau of Statistics in late 1993 and early 1994.

The Child Health Survey is unique for a number of reasons. Set in the context of a general health survey, it aimed to delineate the nature and extent of mental health problems in a state-wide representative sample of over 2,700 children aged 4 - 16 years. The survey was based on an ecological view of child development in which the family, the school and the local community are seen as the three key spheres of influence shaping children's development. For this reason information on comparable measures was collected separately from parents, teachers and teens. To our knowledge this is the first large scale Australian population study to have used this type of multiple informant methodology.

Now while the overlapping spheres of influence of family, school and community are the most proximal influences in children's development, their functioning needs to be seen within the broader context of the larger social, economic, political and cultural environment. Our collaboration with the Australian Bureau of Statistics has been particularly useful in this respect because it has enabled us to relate the Child Health Survey data on influences within the family, school and local community to more general population trends such as those reported from the census and the ABS monthly population surveys.

The Child Health Survey sought to identify the risk and protective factors which help to explain why it is that not all children exposed to adverse family and social circumstances develop problems. Just what is it that enables some children to overcome difficulties in situations where one would generally expect an adverse outcome? What is it about these children and their particular situation which enables them to 'overcome the odds'?

In examining these questions we were concerned with examining spectral outcomes (such as mental health problems, academic competency, drug and alcohol use and juvenile offending) to help define their shared risk settings and to examine the degree to which they were inter-related. Establishing the relative strength and prevalence of specific risk factors in the general population is an essential first step in the systematic development of prevention strategies whose efficacy can be shown to be effective.

In contrast to other areas of child and adolescent health, preventive approaches to reduce mental health problems have been slow to develop. Until fairly recently, preventive interventions have mostly been 'one-off', politically initiated responses to deal with the latest problem which is receiving media attention. Such preventive interventions have all too often been motivated by a 'problem of the month' kind of thinking and have seldom been implemented for long enough to be effective or be evaluated. Whilst well intended, many of these programs have often been poorly targeted, inadequately resourced or simply unable to justify their continuation once the media attention has died down.

Systematic approaches to prevention have been highly effective in tackling complex social and health problems such as tobacco smoking and containment of the HIV/AIDS epidemic. Public health prevention strategies generally operate in two main ways. They either aim to reduce the occurrence or exposure to particular risk factors or they seek to interrupt the pathways or mechanisms whereby risk factors lead to adverse outcomes. In tackling large scale and complex social problems, multi-level and multiple-impact approaches have generally been found to be more effective. These may range from highly targeted programs for individuals at known increased risk or they may target the general population or large sectors of the population and aim maintain health and to prevent problems arising in the first place.

The Child Health Survey was designed specifically to provide an epidemiological knowledge-base which could assist in deciding which models of prevention are likely to most effective in fostering children's growth and development - and at what cost. I would like to share with you this morning some of the insights we believe we have gained from a close look at Australian family life in the 1990's and to reflect on some of the implications of the survey findings for the parenting of adolescents and how this can be supported by public policy and strategies for prevention.

The first volume of findings from the Child Health Survey described health issues and concerns affecting young Western Australians and drew attention to mental health problems as a increasing cause of suffering to children, young people and their families and our community. The findings pointed to the significant discrepancy between the number of children with serious mental health problems and the number actually receiving assistance. They provided one of the first overviews of the proportion of troubled children coming to attention in settings such as schools, health services and the juvenile justice system.

This volume also documented the nature and extent of a range of adolescent health-risk behaviours and the factors associated with their occurrence. These included dietary behaviours, alcohol consumption, tobacco smoking, marijuana use, sexual behaviour and, deliberate self-harm. As has been shown in other surveys of this kind such as the recent Victorian adolescent health survey, those adolescents who display one health-risk behaviour are much more likely to display other health risk behaviours. For example, adolescents who drank alcohol regularly were more likely to have used marijuana, to have had sexual intercourse, to have had police contact, or been to the children's court. This has important implications for prevention. It suggests that far more effective targeting of these programs can be achieved if the whole range if risk behaviours is considered, rather than simply focusing on a single behaviour. However this requires identifying modifiable causal risk factors which are common to all or most of them.

The second volume of findings extended this description into the domain of families and communities. It documents how parents and caregivers are faced with an increasing range of choices about how they organise their work and family responsibilities. It describes how these decisions are made in a social context that includes the composition of the family; the resources available in terms of employment, income and time; the parents knowledge and skills in parenting; the supports available from family, friends and neighbours; and the extent to which these are backed up by broader community and environmental factors such as 'family friendly' industrial relations policies, and access and entitlement to child care and other benefits.

The third volume of survey findings will be launched in February next year. This volume focuses on aspects of children's health, mental health and academic competency that may be directly affected by schools, their staffing and pastoral care arrangements and how these relate to the survey data about families and communities. It has a particular emphasis on the educational implications of mental health problems and the important protective value which school serves for many students.

What do the survey findings reveal about the context in which families are managing the task of parenting adolescents in the 1990s? Three major themes immediately stand out. These are the increased diversity of family living arrangements, the difficulties which many families experience in balancing their work and family responsibilities and the extent to which families are able to draw upon neighbourhood and broader community supports.

Family diversity.

Describing family life in the 1990s is a surprisingly difficult task. It's almost impossible to capture the diversity and complexity of today's families with children and adolescents. Because our beliefs, attitudes and values are generally conditioned by our own experiences it is useful to reflect how many of the expectations which are placed on the families of adolescents derive from our own family experience in a different era. We need to acknowledge just how much things have changed if we are to understand the range of issues which today's families must contend with daily. The demands, expectations and temptations which adolescents now encounter are more numerous and often carry larger risks than those experienced by adolescents only a decade or two ago. These include growth of the media's influence, reduced expectations for employment and opportunities for exposure to risks and potentially hazardous situations unknown to previous generations.

Central to the changing context of parenting is the increased proportion of children growing up in non-traditional family forms. Despite the extent of this change over the past few decades we need to keep in mind that the majority of Australian adolescents (i.e. around three quarters) live with both parents in their original families. We also need to remember that almost 60% of one-parent families result from separation or divorce and that the proportion of persons remarrying following divorce is continuing to rise. The effects of this in creating complex new patters of family relationship are well summed up by Judith Stacey (1991) in her book Brave New Families where she quotes Delia Ephron's modern variation on the theme of the extended family as:

....your ex-husband or (ex)-wife, your ex's new partner, your new partner, possibly your new partner's ex, and any new partner that your new partner's ex has acquired. It may consist entirely of people who are not related by blood, many of who can't stand each other. (p.61)

It used to be that one would, upon meeting another adult ask how many children they had. Today, it may be more appropriate to ask children how many parents they have. This is a real consideration for family professionals who need to know whether the child has one or two residences, whether or not they have siblings and whether or not those siblings have the same surnames. These facts are far from trivial for those that experience them. They are also something of a nightmare for a researcher trying to meaningfully categorising the various permutations which can arise.

Work and family life

More parents are now working and they are working longer hours. The average Australian family is now a dual earner family. About a third of principal caregivers (mostly mothers) in both couple and one parent families are not in the labour force but these proportions are decreasing over time. In practical terms this means that since 1980 increasing numbers of women are participating in earning the family income. Secondary caregivers (mostly fathers) are now working 48 hours per week on average - but this varies considerably by occupational group. Managers, administrators and professionals average between 50 and 60 hour per week while clerks average around 40 hours per week. These are just the average figure - the range of hours worked reveal that some parents are working even longer - especially those working in the service industries. When primary caregivers (i.e. mothers) are employed, they work an average of 25 hours per week although this also varies from 21 to 39 hours for individual occupational groups.

Family and community life

Children in families with higher levels of community and civic participation e.g. with greater access to social supports from family, friends and neighbourhoods and those who were more actively involved with school, recreational and other community activities were found to be clearly advantaged in comparison with children in families with less community involvement. The ability of parents to successfully raise their children is supported by the facilities and resources which are available within their own community. The human and financial investment made by schools, local government, the business community, service clubs, church groups, youth clubs, sporting, recreational and other community organisations all contribute substantially in promoting the social cohesion and the quality of life of all members of the community. They are also critical in promoting two key protective factors which foster the development of positive mental health. These are a sense of being cared about; and a sense of connectedness in their lives, particularly a sense of connectedness to family and schools (Resnick et al, 1993).

Recapping briefly, increased family diversity, changing household patterns of employment and the reduced time available to spend with family members or participate in recreation and community life each have important implications for the upbringing of children. What are their individual and combined effects on the child and adolescent well-being? Before discussing this I need to clarify our use of the term mental health problems as it we use it as a general indicator of child and adolescent well-being.

Mental health problems refer to emotional or behaviour problems identified by parent or teacher reports on the Achenbach Child Behaviour Checklist to be within the clinical range. They involve disturbances of feelings, behaviours and thoughts which disturb the child, distress others or impede development. They include diagnostic categories of disorder such anxiety and depressive disorders, conduct disorder, attention deficit-hyperactivity disorder (ADHD) and somatisation disorders.

The Child Health Survey found that around 17% of 4 - 16 year old children and adolescents have significant mental health problems - a prevalence rate which is comparable to the rates reported by similar recent surveys in Canada, the USA, and Europe. Over half of the parents of these children reported that they had been seen or consulted schools, police or the WA Family & Children's Services with regard to concerns about their child's behaviour during the past six months but only 2% of these children had attended a mental health service within the same period.

Figure 1: Parents employment status and child mental health

Now, returning to the question of what the implications of parents work arrangements might be for the mental health of children it is clear that in both couple and one-parent families, some level of parental employment is protective of child mental health. Practically speaking, there is only so much time in a week and only so many hands that can supply the care a family needs. Employment erodes the energy available for the care of all family members and disrupts the continuity in family caring. In couple families, the traditional roles where of a father who worked and a mother who stayed at home. In effect, this arrangement tied the father's time and energy in securing family income and the mother's time and energy to the supply of care to the family. This arrangement is a choice now made by just over a quarter of families with children within this age group. The most common pattern is for one parent to work full-time and the other to undertake part--time work.

The data shown here indicate that in couple families there is little difference in the likelihood of children aged 4-16 years having a mental health problem whether both parents are employed or if one stayed at home. In one-parent families parental employment is clearly associated with better child mental health. However, other data in the survey show that the income related benefits from the parent's work outside the home can be at the expense of that parent's physical and mental health - particularly where the parent is alone in managing the responsibilities of child rearing and work.

Figure 2: Household income and percentage of children with problems

There is now a broad literature on the relationships among various family variables and child and adolescent health and psycho-social outcomes. From this literature, family income is frequently identified as a foremost 'predictor' of outcomes., Not surprisingly, the combined household income of families within the Child Health Survey found to be predictive of whether or not children experienced mental health problems or had low academic achievement. However, this association does not, in and of itself, further an understanding of the underlying mechanisms whereby poverty produces these adverse outcomes for children. In exploring this further, it emerges that household income is of lesser importance than other more proximal relationships affecting the ability of parents to provide children the stability and security they need.

Figure 3: Family type and percentage of children with problems

There has been a lot of interest in what the Child Health Survey has shown about the mental health and well-being of children living in different family structures. However, it must be kept in mind that the survey is not a study of divorce - its merely provides a 'snap-shot' of the average mental health status of the children who are living in different family living arrangements at a particular point in time. These findings tell us nothing about the longer term outcomes for individual children. Our findings of higher rates of mental health problems experienced by children one-parent and step/blended families are similar to those of recent studies in the UK and north America. However the mechanism whereby children in non-traditional family forms come to have higher risks for adverse mental health and poorer educational outcomes requires a consideration of other factors.

Figure 4: Family discord and percentage of children with problems

Dysfunctional family relationships and parental disciplinary style have long been recognised as generic risk factors which can make a critical difference for the health and development of children. About 12% of families report a level of family discord which seriously impairs family functioning. Compared with other families, these troubled families are more likely to avoid discussing concerns (71% vs 13%), they have difficulties planning activities (53% vs 11%), confiding in one another (51% vs 4%) and expressing feelings (47% vs 3%). These families are 12 times more likely than other families to report marital difficulties. Children living within such discordant families also experience significantly higher rates of mental health problems and low educational performance.

Figure 5: Parental disciplinary style and percentage of children with problems

Parental disciplinary style is an another important and potentially modifiable family risk factor. The Child Health Survey obtained information about the frequency of two contrasting discipline approaches - those based on reasoning and encouragement those which rely on power assertion. This showed that children reared with predominantly encouraging styles of parenting (i.e. where parents put greater effort into rewarding desirable child behaviour than punishing undesirable behaviours) have substantially lower rates of mental health problems and low academic achievement. It comes as no surprise that harsh and inconsistent practices of parenting are harmful to the mental health of children. However the current focus on preventing physical and sexual abuse has tended to ignore the effect which more prevalent and negative practices of parenting make to the mental health outcomes of children.

The recent developmental literature on discipline effectiveness has drawn attention to . the way in which discipline facilitates or impedes children internalising the values and attitudes of society and the development of socially acceptable behaviour which is motivated not by anticipation of external consequences but by intrinsic or internal factors. Parental disciplinary style also plays an important role in the development of self-esteem and the maintenance of the parent-child relationships. Over-reliance of power assertion is generally detrimental to the socialisation process, because it arouses hostility and anger in the child with accompanying opposition or unwillingness to comply with the parent's wishes. Power assertion also provides models of aggression that increases the likelihood of anti-social conduct.

There is now clear evidence from Australian and overseas studies that when parents are taught how to modify damaging styles of parenting, the effects on children's mental health are direct and positive - particularly when this is applied at critical stages in a child's development -such as in the pre-school years and in early adolescence i.e around the age of 11 -14 years.,,

Most of the above family risk factors are not independent of one another. One-parent families are clearly over-represented in the lower ranges of family income and knowing the income or structure of a family tells us nothing about what goes on within the family. To understand how these risks interact and off-set one another it is possible to calculate the relative contribution of each of these factors. This was done using a statistical technique which measures the strength of association of each risk-factor whilst holding the others constant.* When fitted to the Child Health Survey data , this model had an explanatory power to correctly classify 83% of children with mental health problems solely on the knowledge of three questions: What is the child's family type; what is the parental disciplinary style, and is there a significant level of family discord present? Interestingly, family income did not reach significance in predicting mental health status when these other family variables were simultaneously taken into account.

Table 1

/TR> >
Family risk factors associated with child mental health problems (a)
P levelRisk
odds ratio)
Interval (95%)
Parental Disciplinary Style:
- Coercive <0.00013.31.9 - 5.6
- Detached0.00042.21.3 - 3.7
- Inconsistent<0.00012.21.7 - 3.0
Family Type:
- Step-blended <0.00012.41.6 - 3.6
- One parent<0.00012.51.8 - 3.5
Level of Family Discord:
- High0.00041.71.2 - 2.4

(a) Risks are calculated relative to the least adverse combination of family risk factor i.e. encouraging parenting style, original family type and low family discord

In this table the risk (adjusted odds ratio) associated with each risk factor is shown relative to the first category of each variable (e.g. for family type, the odds are shown relative to original families). These findings demonstrate that regardless of family income, children in one parent and step/blended families have a similar increased risk of mental health problems in comparison with children in original families when similar conditions of family discord and parental disciplinary style apply. Thus in predicting or preventing child mental health problems, what happens between people within families is just as important as the family structure the child happens to be living in.

Table 2.
Estimate of preventable fraction for child mental health problems

p (%)N for State  O.R. (Adjusted odds ratio) A.R. (Attrib -utable risk)Estimated no.of cases attributable to risk factorCases theoretically preventable by 20% reduction of risk exposure
16.6 %50,2812.519.4%10,3792,075
Family discord11.2 %33,9241.78.0%4,280 856
Adverse parent-
50.9 %154,1762.519.1%10,2182,043

(a) Includes coercive, neutral, & inconsistent parenting styles

Having established the magnitude of the risk associated with each of these factors and knowing its prevalence (i.e. how commonly it occurs in the general population) it is then possible to estimate the proportion of cases which may theoretically be attributed to each factor. Reading along the top line of this table you can see that 9.6% or about 29,000 Western Australian children aged 4 -16 years live in step/blended families. Their average risk for mental health problems is 2.4 times that of children in original families. Thus 11.9% or 6,366 cases of mental health problems in WA children are theoretically attributable to the contribution of this factor. This means that if we were somehow able to reduce by 20% the effects of this risk-factor then the maximum reduction in the number of cases we could hope for would be 1,273.

In the same way (reading along the bottom line of the table) it can be seen that if the population prevalence of adverse parenting (i.e. coercive, detached, and inconsistent parenting) was reduced by 20% we could theoretically decrease the number of significant mental health problems by about 2,000. This is close to the total number of cases seen by Perth child and adolescent mental health services each year. (Incidentally, the cost to the State for treating this number of cases in the public system in 1992 was in the region of $4.5 million.) These data highlight the futility of providing more and more treatment services without also setting into motion programs of prevention. Improving parenting skills and strengthening family functioning is clearly an area of community need and one which has significant potential for prevention. I would like to give you two examples of how the survey findings on parenting have now been translated into policy and practice for prevention.

The first is an example of a preventive intervention targeted to the specific needs of high risk families. The WA Health Department has over the past 18 months spent half a million dollars in developing, implementing and evaluating an adaptation of the Matt Sanders Triple P program (i.e. levels 4 & 5) delivered as a generally available public health intervention for all parents with pre-school aged children. By next month, 800 Perth families living in disadvantaged areas with the highest child abuse notification rates will have completed the program. Pre- and post intervention results for the first 400 families to have completed the program are encouraging. Three-in-five eligible families in these areas volunteered to participate and 85% of these families completed at least 7 of the 8 weekly program sessions. In the short term the program has been effective in reducing the incidence of adverse or seriously dysfunctional parenting from twice the population average prior to the program down to the general population level. It has also significantly reduced disruptive behaviour disorder among the children of the participating families. We are now following them up on a longer term basis and comparing their progress with another group of 800 families recruited from similar but separate high-risk areas. We would like to think that over and above prevention of child mental health morbidity, the program has potential for reducing child abuse by shifting population norms in the use and 'acceptability' of coercive child management practices.

The second example is a universal preventive intervention targeting the general community. The survey data were used by the WA Ministerial Task-Force on Families to back its recommendation for a high-profile, community-wide parenting campaign to raise community awareness to the importance of parenting and to encourage parents access to parent training opportunities and to seek help where this is needed. The WA Family and Children' Services launched its parenting campaign earlier this year as part of an overall organisational change towards a stronger preventative emphasis in all of its services. The following 5 minute video clip shows two of the television commercials and vignettes from three of parenting training videos which illustrate how some of the key concepts and information about parenting can be made more generally available through the mass media.

To conclude, I would like to return to some of the themes touched on earlier. Families in advanced industrial countries such as Australia do not see themselves as self-sufficient in meeting all of their families needs. Parents tell us they believe their role is important but that they feel this not properly acknowledged nor supported by the broader community. The capacity to successfully rear children is influenced by the extent to which they are supported through 'family friendly' public policy and the availability of informal community supports and good information .

In communities where there are few family supports, schools, child-care and other community agencies are taking on an increasingly important role in supporting families and children in the complex tasks of socialisation. The global village has converted the pastoral capacity of families to an economic expectation. It is ironic therefore that in this time when private and public-sector reform is placing emphasis on productivity and a return to core-business, schools, mental health and other agencies of social support are increasingly being looked to by parents and communities to fulfil roles which were formerly the primary responsibility of families.

While the pace of change has been formidable we do not have to be at the mercy of these changes. The challenge is to tackle these issues in systematic and appropriate ways. Just as families need to be supported in their roles, so do teachers, schools, employers and community organisations. We need to keep raising and talking about these issues so that ideas can become shared knowledge which informs pro-active policies to ensure the health, well-being and competence of the next generation and the kind of future we all hope to share.

Return to Conference papers / AIFS Home Page