Some of the greatest improvements in health in recent years have come from the recognition of 'non-disease' aspects of health and the discovery of strategies to address these issues especially through 'health promotion'. The consequent fall in levels of tobacco smoking, in drink-driving and in careless driving, in levels of saturated fats in the diet and increased levels of exercise and activity have impacted significantly on the health of various populations. However, there is considerable evidence that a group of factors have been overlooked in health promotion and health strategy literature. A group of family related variables including parenting styles, spouse relationships, family stability, family composition and marital status impact surprisingly strongly on health outcomes. To date they are given virtually no attention in national health strategies. If recognised and addressed, these family related variables offer a new area for health promotion with strong possibilities for further success in reducing the burden of ill-health in the community.

Family related variables and mental health

The Western Australian Child Health Survey states that 'Child mental health problems are a lifestyle associated disease of epidemic proportions' (Zubrick 1995) and finds a number of family-related factors strongly linked to child mental health problems such as parenting styles, family functioning (the quality of adult relationships), the life skills competency of parents and family composition.

Parenting styles

Many studies have noted links between parenting styles and children's mental health. One of the most recent to do so, The Western Australian Child Health Survey identified four common parental disciplinary styles which were linked to the prevalence of mental health problems (Silburn et al. 1996c p. 43). The styles (in order of frequency) were: encouraging, inconsistent, neutral and coercive. Children whose parents used either the coercive or inconsistent disciplinary styles were more likely (29 per cent and 24 per cent respectively) to have a mental health problem compared with only around 11 per cent of children whose parents used the encouraging parenting style or 18 per cent of all children. These children were also more likely to exhibit a behavioural syndrome such as delinquency, attention or social problems and somatic complaints.

Family functioning

The way the family functions, especially the presence or absence of serious discord and conflict also impacts on the mental health of parents and children (Eastman 1996; Eastman 1989; Lewis et al. 1976; Scientific Advisory Committee on Families and Mental Health 1995; Walsh 1982). There is increasing evidence it also impacts on their physical health (Crockett and Tripp 1994; Lewis et al. 1976). One of the most important aspects of family functioning is the way the spouses or partners relate to each other, as this sets the tone of the whole family's way of relating. (Eastman 1989). The Western Australian Child Health Survey collected data on the quality of adult relationships and related this data to children's mental health. For each family form studied (original--that is intact, step and blended, and single parent) 'good quality adult relationships are associated with lower rates of mental health morbidity' (Taskforce on Families in Western Australia 1995 p. 62).

Family stability/mobility

One aspect of family stability is the stability of the family composition--for example whether children are with their original parents or not. Another aspect of stability is geographic--the number of dwellings lived in and the number of schools attended. Geographic mobility tends to be linked with the stability of the family composition. The Western Australian Child Health Survey found that mobility seems to have a strong impact on children's mental health. Thirty-three per cent of children in families that had moved within the past two years had a mental health problem, compared with 18 per cent of those who had resided in their current dwelling for 11 years or more.

Family composition and marital status

For some time now there has been a focus on the processes within the family (rather than the family composition) as relevent to the mental and physical health of the members. For example, in 1989 I wrote that the family processes were more related to health and well-being of family members than the composition of the family--whether biologically intact, two parent; or step or blended family; or single parent family (Eastman 1989). Increasingly evidence is emerging that the composition of the family is also important. See, for example the Western Australian Child Health Survey's data on marriage/relationship breakdown and child mental health problems. It was reported above that for all family forms, the quality of adult relationships was related to children's mental health outcomes. But the children's mental health outcomes were also related to the different family forms.

Percentage of children with mental health problems

Intact	Step/Blended	One parent
14%	26%	29%

'In this and other studies where the effects of low household income or other measures of socio-economic adversity are controlled for, children and adolescents from one parent families are consistently more likely than their peers in couple families to have mental health problems or to engage in health compromising behaviours, such as drug and alcohol use, unprotected sex, cigarette smoking and dropping out of school.' (Silburn et al. 1996c pp. 61-62)

There is little doubt that marital status is a health factor comparable to smoking in significance. In fact, it is likely that it outweighs smoking as a lifestyle factor affecting health. US data shows that divorced men who do not smoke have a death rate from lung cancer only slightly less than married men who smoke a packet or more of cigarettes a day (Morowitz 1975). Australian data (as reported, for example, in the National Health Strategy's 1992 report Enough to Make You Sick) are consistent with Morowitz' data.) However, this is by no means the whole story as Australian data shows that divorced women have mortality rates from heart disease 50 percent higher than married women, and divorced men's mortality rate from heart disease is even higher--twice that of married men. Divorced men and women suicide three times more often than their married counterparts (and the never married rates are not far behind) and divorced men have more than double the rate of deaths from motor vehicle accidents than married men with divorced women's rates are almost double the rate of married women. In what follows the focus is on one of the family related variables--one of the easiest on which to collect data, that is marital status. Data that indicates the importance of marital status as a factor impacting on health outcomes and on the overlooking of marital status (and other family related variables) in health strategy documents is reviewed. Finally some of the policy implications of the recognition of these family related variables are briefly discussed.

The contribution of marital status and family to health

Marital status is a significant factor impacting on health, outweighing in impact the factor of smoking or not smoking, yet its importance has been overlooked in health strategy documents (Dominian 1991; Eastman 1996; Eastman 1989; Larson et al. 1995).

In a classic paper Berkman and Syme studied four types of social connectedness: marital status, contact with extended family and friends, church membership, and other group affiliations. They found that each type of social connection and all of them together predicted how likely a subject was to die within the next nine years. Those least socially connected were twice as likely to die as those with the strongest social ties. Overall, while all types of social connection were important to survival, whether or not an adult was married was the most important predictor of survival (Berkman and Syme 1979).

Hu and Goldman (Hu and Goldman 1990) analysed marital status specific death rates for 16 developed countries. In all countries, death rates of single, widowed and divorced men greatly exceeded those for married men. The differences were greater in the younger age groups and greater for the divorced. For the widowed and divorced in their twenties and early thirties mortality risks were sometimes 10 times those of the married.

Morowitz re-examined data from the 1963 Hammond report which had documented the health risk of smoking and found that nonsmokers who were divorced had only a slightly lower risk of dying from cancer than married men who smoked a pack or more of cigarettes a day (Morowitz 1975). The age-specific death rate for divorced people in the US is 1.84 or 84% higher than for married people (Larson et al. 1995) and Larson declares that overall that translates to a loss of ten years life per divorced man. In terms of men's health, divorce equals smoking a packet of cigarettes a day for the rest of one's life.

Ladbrook found that in Wisconsin, professional women (employment category 'Technical and Professional') were dying at higher rates than were men. The main factor accounting for this reversal of the usual pattern (males in the US usually die six years earlier than females) was the higher ratio of women who were never married, widowed, separated or divorced compared with the married than was the case with men. A considerably higher percentage of men than women in this category were married and the physical health protection offered by marriage was so great that men in this category were living longer than women. In other words, marriage is such a powerful protector of physical health, that the higher ratio of never married, separated and divorced to the married among women than men in the category of Technical and Professional resulted in death rates for all women in that category being higher than death rates for men. (Ladbrook 1990)

Marriage even protects from contracting cancer and also offers better chance of survival after cancer diagnosis. Lilienfield et al found that nearly every type of terminal cancer struck divorced individuals of either sexes, both white and non-white, more frequently than it did the married (Lilienfield et al. 1972). Divorced males had double the rate of respiratory cancer, a four-fold increase in buccal cavity and pharynx (throat) cancer, and more than a 50 percent increase in cancer of the digestive system and peritoneum and urinary tract. James Goodwin studied the relationship between marital status and survival among several thousand cancer patients and found that married cancer patients did better medically than unmarried cancer patients (Goodwin et al. 1987).

Stressful marriage contributes to ill-health

On the other hand it is well known that stressful and dysfunctional marital and family relationships contribute to ill-health (Walker 1996) and are implicated in contributing to intractable physical pain(Towers 1995). Hence marital and family stress, conflict, and dysfunction are probably a contributing factor to many forms of poor health. This is another reason why programs to increase the number of happily married couples should be a health strategy priority.

National health strategies ignore the role of marital status and family stability and processes in creating or undermining health

Despite evidence that marital status, marital stability, parenting styles and spouse relationships are clearly linked to health outcomes, there is little sign of this being recognised in national health strategy documents. The following outlines how family variables have been overlooked in some of the major recent health documents relating to national health strategies.

Lee, S.H., Smith, L., d'Espaignet, E. and Thomson, N. 1987. Health Differentials Among Working Age Australians. Canberra: Australian Institute of Health.

This report was produced under the auspices of the Australian Institute of Health as part of its Health Differentials Project. The first objective of this project was

'to collect and integrate the existing empirical data on the differences in health status and risk factors between various socio-demographic groups in the Australian population, in order to document the position both of high risk groups at whom investigation and prevention programs should be directed, and of those groups with better health who can be regarded as providing target levels for the rest of the population.

In Health Differentials Among Working Age Australians the health risks of the never married and divorced/widowed were clearly identified. :

There are very large differences in mortality between married/separated men on the one hand, and never married and divorced/widowed men on the other. The latter groups have standardised rates over twice the former's. . . . separated/divorced/widowed men have more acute symptoms and mental health problems and smoke and drink more, although only the smoking and mental health differences are of comparable magnitude to the mortality differences. (Lee et al. 1987 p. 25)

The differences between women in different marital status groups are not quite as extreme as those for men, but the mortality of never married women is still 80 percent higher than that of married women, and that of divorced/widowed women is over 60 percent higher. The separated/divorced/widowed women in the surveys also report mental health problems, and smoke, at levels 80 percent above married women, and they report 20 percent more acute and chronic symptoms, the latter in contrast to men in the same group who show no excess. (Lee et al. 1987 p. 27)

It is surprising, therefore, to see how this strong finding appears to be overlooked in reports since that time.

Enough to Make you Sick: How Income and Environment Affect Health: the National Health Strategy's Research Paper No 1, 1992

The National Health Strategy was set up in 1990 to identify ways of improving the effectiveness and efficiency of the health system. Enough to Make you Sick is the 'first research paper by the National Health Strategy (p. 3). Staff from the AIHW and the Commonwealth Department of Health, Housing and Community Services helped in its compilation but it is not clear from the document what is the exact relationship between these various bodies. In Enough to Make you Sick the focus has shifted slightly from the AIHW research cited above. Lee's research focused on 'differences in health status and risk factors between various socio-demographic groups'. Enough to Make you Sick focuses on 'the challenge of reducing health inequalities in Australia' and especially how those in the lowest socio-economic bracket suffer much more ill-health than those in the highest bracket. That is, the assumption seems to be that inequality and disadvantage is socio-economic only (though differences related to ethnicity, aboriginality, metropolitan/non-metropolitan location and marital status are collated and reported) The result is that although inequalities between groups of different marital status are clearly identified, only the inequalities relating to low socioeconomic status are seen as significant. This leads to the following paradox.

Enough to Make you Sick confirms the strong correlations between marital status and health outcomes reported in the research cited above, especially the AIH 1987 Health Differentials report. It summarises its data on correlations between marital status, sickness and mortality in the following words:

'With the exception of stomach cancer, brain cancer, pancreatic cancer (in women) and prostate cancer (in men), married individuals aged 25-64 are at less risk of dying from all selected causes of death than never married individuals, widowed/divorced individuals or both (of the same age).' (National Health Strategy 1992 p. 46) ( See Figure 25)

Enough to Make you Sick gives the eight causes of death for which correlations are strongest between low socio-economic status and cause of death. For seven of these causes of death, correlations are even stronger with marital status. That is, the never married, widowed and divorced have higher death rates on seven of these eight causes of death compared with the married than do the lowest socio-economic bracket compared with the highest socio-economic bracket. (See the charts on the following pages.) Amazingly, apart from the one sentence quoted above, this striking correlation between marital status and mortality rates is completely ignored in the report.

Chapter 5, 'Summary of major health inequalities for all Australians' does not even mention marital status, nor does it refer to the increasing rates of divorce, the tendency to defer or not marry and the implications of this for the health of Australians. A short section on 'behavioural or lifestyle influences' fails to mention marital status or family composition as health factors. It could, of course, be argued that Australians will increasingly avoid formal marriage and opt for informal or de facto marriage and that as this becomes more common and acceptable, the health outcomes will equate with formal marriage. But surely the scientific attitude is to monitor and evaluate such changes, not to assume they will be benign.

Similarly, Chapter 7, 'Reducing the differences in health in Australia', does not mention marital status or family composition other than to note under 'social support' 'isolated parents rearing young children' and 'sole parents who may have few social contacts' among those whose health risks are magnified. (P 105) There is absolutely no discussion of the pathways by which marital status and health may be causally connected. Nor, on the other hand, is it argued that the correlations are misleading or irrelevant. They are simply ignored. There is no discussion of how these correlations may be further studied. There is no discussion of the multitude of ways in which marriage may be supported and marital breakdown reduced. There is no awareness shown of research showing the effectiveness of pre-marriage education programs (Harris et al. 1992) or of marital counselling (Wolcott, AIFS).

The final paragraph of Enough to Make You Sick states, 'The association between poor health and poverty, between poor health and inadequate housing, between poor health and unemployment is unequivocal.' But while the mortality rate for males who are unemployed or not in the workforce is 17% higher than for employed males, for never married men it is 124% higher than for married men and for divorced/widowed men 102% higher--yet marital status is not even mentioned in the summing up of the report's findings.

Appendix 1 reviews Australian and overseas literature on inequalities in health. The section on mortality focuses on low socio-economic status, ethnicity and metropolitan/non-metropolitan location. No mention is made of marital status though a large literature (and the report's own data) shows its significance far outweighs the factors mentioned.

Figure 1: Mortality differentials - women

Figure 2: Mortality differentials - men

Under morbidity socioeconomic status, income, education, and employment status are mentioned, but again, no mention is made of marital status. The AIH's 1987 findings that low income males report more mental health problems, chronic symptoms and acute symptoms is mentioned but not the correlations reported between mental health problems and chronic symptoms and marital status. The literature on morbidity and marital status is not even mentioned.

A major section of the appendix reviews the literature on the causal link between socio-economic status and health status. But although the report found much stronger links between marital status and health outcomes, for marital status there is absolutely no discussion of causal links.

Despite the fact that Enough to Make You Sick asserts that 'health policies should not rely solely on overall rates of health outcomes, risk factors and access to services, but that the distribution of these indicators within a population should always be examined' (National Health Strategy 1992 p.108). the health disadvantages of the single, widowed and divorced have been neither highlighted nor kept in mind by the report. This avoidance of the data on links between marital status and health is to be found in other health strategy documents.

The National Mental Health Policy: 1992, Canberra:

The National Mental Health Policy (National Mental Health Policy 1992) was sponsored by the Australian Health Ministers Advisory Council, the National Working Group on Mental Health Policy and the Mental Health Branch, Commonwealth Department of Human Services and Health.

The National Mental Health Policy does not mention the role of parents in children's mental health despite the fact that it has been known for many years that children's mental health is closely related to the kind of parenting they receive. More recently, work such as that of Dr Matt Sanders shows that the mental health of children can be markedly improved following educational programs for parents on parenting practices (Scott 1995). A group of scientific experts recently stated that if the government's policy objectives in the field of mental health are to be realised:

much greater attention to the needs of families is required. Coordinated strategies are required that link mental health and family policy. The well-being and stability of the family unit must become a priority area in the provision of mental health services' (Scientific Advisory Committee on Families and Mental Health 1995 p. 2).

Work at the Western Australian Institute for Children's Health also indicates the need for support of marriage if children's mental health problems are to be reduced (Silburn and al 1996; Silburn et al. 1996a; Silburn et al. 1996b; Silburn et al. 1996c; Taskforce on Families in Western Australia 1995; Zubrick 1995)

Review of National Health Goals and Targets: 1992

'[The goals and targets] report is one of a series conducted by the Commonwealth Department of Health, Housing and Community Services (DHHCS) to assist in setting directions for the organisation and funding of health services, and for improving health in Australia. . . These include a comprehensive evaluation of the National Better Health Program and the wide-ranging work of the National Health Strategy (NHS)'. (Nutbeam et al. 1993)

In referring to priorities for the future the progress report of the Review of National Health Goals and Targets (Nutbeam et al. 1992) discusses setting targets in six broad areas: the natural environment; housing, homes and community infrastructure; transport and communication; schools and other educational institutions; work and workplaces; hospitals and health services. The specific contribution of marital status, marriage relationships, family composition and structure and family dynamics to health goals and targets is not mentioned and families are not areas for which health goals and targets should be established.

Goals and Targets for Australia's Health in the Year 2000 and Beyond, Nutbeam, et al, 1993.. Sydney: Department of Health, University of Sydney for the Commonwealth Department of Health, Housing and Community Services.

In the final report of the review (Nutbeam et al. 1993) the specific contribution of marital status, marriage relationships, family composition and structure and family dynamics to health goals and targets is not mentioned, not even in Chapter 2.4 'Healthy Environments' which includes section 2.4.3 on 'Housing, Home and Community Infrastructure'. Under 'safe housing' the issue of domestic violence is mentioned, specifically mentioning Burdekin's report into homeless children, but not mentioning his finding that marital breakup is a major cause of homelessness and that children in single parent families are at six times the risk of abuse of children in two parent families. (Other studies find that children in step-families are at higher risk of abuse than children in intact two-parent families.).

Section 2.1.10 on Mental Health Problems and Disorders does not mention the role of different parenting styles in contributing to or preventing mental health problems in children. The higher rate of mental health problems and of suicide among the never married and divorced (in the case of suicide 300 per cent higher) is not mentioned but this was clearly evident in earlier publications on health differentials by the National Health Strategy (Lee et al. 1987; National Health Strategy 1992).

Section 2.3.5 on Social Support nowhere mentions the pre-eminent role of family in creating social support, nor the health implications of high rates of marriage break-up. One paragraph which does report on the role of family and marriage as an aspect of social support (on p. 168) states:

'Social support, especially as expressed through close personal relationships found in the family, has been shown to influence the outcomes of health and wellbeing. (Here two 1985 and 86 US books are cited--but not the earlier publications by the AIH and the National Health Strategy itself (Lee et al. 1987; National Health Strategy 1992)). Several studies have associated lack of social support with increased mortality in the old, (here a 1987 article on the Alameda County study is cited--with no mention that the Alameda County research is the classic research which found marriage to be the form of social support most connected with health outcomes, nor that marriage, powerfully impacts on mortality in all age groups not just for the old, nor is there any mention of the AIH and the National Health Strategy's own data showing that the never-married and the divorced have considerably higher mortality rates than the married) increased incidence of coronary disease, (here a citation of a 1984 US article on psychosocial influences on mortality after myocardial infarction, but no reference to National Health Strategy's own data showing mortality from ischaemic heart disease and from cerebrovascular disease is 50 to 150 per cent higher for the never-married or divorced) and susceptibility to chronic illness.' (Another 1984 US article cited but no mention of National Health Strategy's own data on marital status and chronic illness.)

Although in one place (p. 169) Enough to Make You Sick is specifically mentioned, the health differentials related to marital status were not mentioned. The only data reported related to socioeconomic disadvantage.

'Better Health Outcomes for Australians' Commonwealth Department of Human Services and Health 1994.

The opening pages note that the Goals and Targets for Australia's Health in the Year 2000 and Beyond report sought to expand the sectors traditionally involved in health promotion. It identified the need to create 'healthy environments' in which people are able to maintain good health and access health services when necessary. '

The 'environment' includes: the physical environment, such as clean air and water; safe and accessible transport; safe, clean and secure housing; safe and supportive workplaces; supportive and educative schools; and effective, responsive, accessible and reliable health services. (Commonwealth Department of Human Services and Health 1994 p. 17)

It is surprising that family is not one of those environments identified as essential for the maintenance of good health. Research has clearly identified parental disciplinary styles, spouse relationship, marital stability and marital status as strongly correlated with health outcomes. No comparable research based health differentials have been noted, for example, for schools. As far as workplaces go an earlier National Health Strategy document reported few morbidity differentials within the group of employed people or according to occupational group (National Health Strategy 1992 p. 119). Overlooking family as one of the key environments related to health outcomes is as serious an oversight as it would be to leave out tobacco smoking from a health strategy document.

The national goals and targets were developed within a context of already existing policies and programs which address the focus areas directly or indirectly. The report lists thirty-one of these national policies and projects such as: National Aboriginal Health Strategy, National Drug Strategic Plan, National Food and Nutrition Policy, National Health Policy on Alcohol, National Health Policy on Methadone and National Health Policy on Tobacco. (Commonwealth Department of Human Services and Health 1994 p. 281-286) In view of the impact of family variables on health outcomes perhaps there should be National Health Policies on Positive Parenting, a National Family Life Strategic Plan, a National Health Policy for Pre-Marriage Education and a National Marriage for Life Strategy. In the absence of such projects and programs, it is almost inevitable that the essential role of families and the task of resourcing that role will not be kept in the mind of the health promotion bodies and the national health strategy.

What impact would it have on those bodies who do provide, say, parent education or marriage or pre-marriage education to be included in those bodies listed above and to be formally linked to the National Health Strategy? It would be energising, empowering, revitalising. The failure to recognise their area of work as vital to Australia's health strategies sends an unintended demoralising and disempowering message of lack of support.

That Better Health Outcomes for Australians has not given any significant attention to family variables in health outcomes or to the possible contribution of families to improved health for all Australians is evident from the fact that the index has no entry for 'family' or family characteristics, nor for marital status, marriage education or pre-marriage education, nor for parent education, parent styles, parent skills or parent practices. But the data is very strong that these factors all impact very strongly on health outcomes and that in some areas such as mental health problems in children and suicide and motor vehicle accidents there is little hope of significantly improving the outcomes with out addressing the family characteristics involved. A closer look at one cause of death for which strong links have been found with both family composition and family functioning indicates that there are good reasons to include family variables in health strategies.

Suicide

Suicide (2294 deaths in 1992) and motor vehicle accidents ( 2066) are the leading cause of death by injury and account for over half the total deaths by injury, (7489 in 1992) (p. 179).

The link between suicide and depression is stressed

'Suicide represents one of the major forms of mortality from mental illness. The vast number of suicides are associated with mental illness, especially major depression and schizophrenia.' p. 22

but the strong links between suicide and marital status are not mentioned. The link with 'family dysfunction' for young people is briefly noted.

The discussion that follows notes correlations with two variables not amenable to intervention: age and gender. The 'male mortality rate from injury is over two-and-a-half times the female rate' (p. 180). The higher rates of injury in Aboriginal and Torres Strait Islander communities (a threefold higher rate) is discussed (p. 189) but there is no discussion of family variables in suicide. Yet marital status alone is a factor stronger than any of those that did rate a mention: For men and women the divorced /widowed have suicide rates over three times that of the married and the never married rates are almost three times the rate of the married. Both divorced/widowed men and never married men die of motor vehicle traffic accidents at over double the rate of married men, and divorced/widowed women and never married women at almost double the rates of their married sisters. UK research reports that those who are separated but not divorced have suicide rates 20 times that of the married (cited in (Dominian 1991). But as well as marital status there is family function and the links between aspects of family functioning such as destructive conflict and depression and suicide are well-established.

The main strategy proposed by which suicide may be addressed is to reduce the opportunities for suicide through

'other measures to reduce access to means of suicide (for example, anti-suicide barriers at well known jump sites, tighter prescription measures for selected drugs, restrictions on the availability of firearms and the use of available devices to prevent carbon monoxide emission from car exhausts). . . there is evidence to suggest that when a common and culturally acceptable method is made unavailable, the reduction in suicide that results is only partly offset by an increase in the use of alternative means (NCIP 1989). (Commonwealth Department of Human Services and Health 1994 p. 175)

Nowhere in the strategies discussion is the concept of further researching the nature of the family's ability to protect from suicide and from the preceding mental health problems. See for example Sanders' publications on the very positive results possible from parent education programs on children's mental health (Sanders and Dadds 1993; Sanders and Markie-Dadds 1992)

Better Health's 'focus areas for action'

Better Health Outcomes selects four 'focus areas for action'. In each of these areas--cardiovascular disease, cancer, injury, and mental health--family related variables powerfully affect health outcomes. The data and arguments that the family related variables should be recognised especially in the strategies to address these areas are very strong. It is difficult to see how health strategies in these focus areas will be successful without addressing family related variables.

Healthy Victorians 2000--Towards Victorian Health Promotion Goals and Targets: 1995

Healthy Victorians 2000--Towards Victorian Health Promotion Goals and Targets discusses smoking, high alcohol consumption, poor diet and lack of exercise in creating health risks. It does not mention the health risks of certain marital status groups or of the links between certain family lifestyles and smoking and high alcohol consumption. The report is arranged under the headings of 'Populations through the lifespan', 'Key health promotion challenges' and 'Health promotion settings' but nowhere is the data on the impact of family on health outcomes reported. Under 'Health promoting communities' the family gets one third of a page. Its role in health promotion 'via the alteration of attitudes, behaviours, support and readiness to access health services' is acknowledged but the strong contribution of marital status and other family variables to health outcomes is nowhere acknowledged. (This is equivalent to health strategy documents and health promotion campaigns stating that 'health messages can be printed on cigarette packages' and ignoring the statistical links demonstrated between smoking and disease and failing to take measures to inform people of these links and to discourage smoking.) This is a serious failure of the report to follow one of its own key criteria:

The lifestyles which people adopt are a key determinant of good health. It is therefore important for health programs and health education initiatives to focus on the provision of information to assist people in making informed choices about their lifestyle behaviour patterns and habits. (Vichealth and Health and Community Services 1995 p. 15)

NHMRC Review and Revision of the National Health Goals and Targets; 1995

The NHMRC's review of the infrastructure supports for national health advancement (National Health and Medical Research Council 1995) focuses on 'learning from success', particularly the health improvements gained by reductions in smoking, road deaths, premature mortality from cardiovascular disease, reduction in the transmission of AIDS, and improvement in aboriginal health indicators. No mention of family factors or marital status appears in the whole document, yet the data suggests that this is an area with at least as much potential for health advancement as in any of the areas that are mentioned.

The AIHW Health Monitoring Series 1994-6

Four reports focusing on health differentials for different age groups have been published in the AIHW Health Monitoring Series (Mathers 1994a; Mathers 1994b; Mathers 1995; Mathers 1996). The health monitoring series like the National Health Strategy Report No. 1, Enough to Make You Sick aimed to extend and update the earlier report by Lee, (Lee et al. 1987). The series confirms the 'very large differences in age standardised mortality between married and unmarried people. (Mathers 1994a pp. 33, 38)

In the final report--on young Australian adults aged 15-24--the family composition analysis reappears but with a sharply contrasting finding to the patterns noted in Enough to Make You Sick. For the 15-24 age group the married have the second worse health statistics of any of the groups (dependents still living at home, singles, single parents or married). Single mothers were the only group to record poorer health than the marrieds. The discussion of this startling finding (it reverses the patterns found in all the above cited research and all the previous AIHW reports) is extremely brief (one paragraph). No mention is made of the fact that for this age group the balance between de facto and formal marriage is quite different from the older age groups. 'Younger people are more likely to live together in a de facto relationship than older people'(ABS 1992 p. 52). In 1992 about 8 per cent of all couple families were de facto unions and in 1985-7 (the mortality data under analysis) the rate would have been less than that. The rate of de facto unions among couples in the 15-24 age group however is far higher than 8 per cent and must be around 30 per cent (ABS 1992 p. 52).

There are many possible explanations of the poor health of the marrieds in the 15-24 age group but, in view of the research reviewed above, one possible explanation that needs to be explored is that formal and de facto marriage differ markedly in their ability to protect health. In view of the rapidly increasing rate of de facto unions and the decline in formal marriage this should be a priority research area.

The overlooking of marital status and other family variables in health strategy documents

Reasons for the overlooking of the family related variables in health strategy data

It is not clear why the health strategy documents have avoided the clear evidence of marital status as a variable in health research. No doubt one reason is that there are very controversial and sensitive aspects to certain family issues. There may be a reluctance about inquiring into what are increasingly seen to be private matters, of no relevance to anyone but the individuals involved. Distinguishing between all the different family forms can be very difficult. Publishing data such as the increased mortality rate of the never-married, separated, widowed and divorced or increased rates of mental illness and suicide of certain marital status groups may seem to be stigmatising those groups. Many people would question the appropriateness of publishing such data. It is likely to be upsetting to people in the high risk categories and, as one single person said on reading some of the data above, 'Well, you can't just get married, can you?' Most people have no desire to return to a now rejected pattern of stigmatising those who do not marry or who separate or divorce, and publishing the data reviewed above may be seen to have that inevitable result.

On the other hand, it seems essential that individuals need to know the implications of the decisions they make in relation to marrying and divorcing and that policy makers make policy on the basis of the full rather than partial information. Linda Waite (Waite 1995), in her recent presidential address to the Population Association of America, summarised much of the research about marriage and well-being and concluded with the following statement urging social scientists to make such data available:

I think social scientists have an obligation to point out the benefits of marriage beyond the mostly emotional ones, which tend to puxh people toward marriage but may not sustain them when the honeymoon is over. We have an equally strong obligation to make policy makers aware of the stakes when they pull the policy levers that discourage marriage. (p. 500)

And as well as the reasons cited above, some less defensible reasons also impact on decisions to recognise family related variables, in health and also in other areas.

The status of marriage: ambivalence among policy makers

Marriage is a social institution that is most highly valued by the general population but marginalised by policy makers. One US scholar of marriage and family has commented on the strange case of marriage counsellors who avoid discussing marriage (Blankenhorn 1995) and another on publishers of family life textbooks who find the word marriage too 'preachy' and 'old-fashioned' and so leave it out, certainly out of titles and increasingly out of the books (Glenn 1996). Blankenhorn's report of one example of how the word 'marriage' is being excised from social science language is particularly striking:

Consider the case of the American Association of Marriage and Family Therapy (AAMFT). Founded in 1942, AAMFT, with over 16,000 members, is the nation's premier professional association of marriage and family counselors. Yet at its annual conferences, the topic of marriage is literally nowhere to be found. For example, at AAMFT's fifty-first annual conference, held in 1993, the conference's 'Subject Guide' listed forty-three major topics to be explored. These topics included: Abuse, AIDS, Couples, Custody, Divorce, Gender, Models, Sexual Issues, Stepfamilies/Single Parents, and Teaching Family Therapy. These major topics were further divided into 234 subtopics, including Communications, Gender Construction, Psychoeducation, and Substance Abuse. Yet in this total of 277 subject listings, the word marriage did not appear once. The closest we get is the subtopic called Remarriage Ceremony, listed under Stepfamilies/Single Parents. As odd as it may seem, the American Association for Marriage and Family Therapy has excised the word marriage from its basic vocabulary. If these professionals were naming their organization today, instead of living with a name chosen by others in 1942, it seems highly unlikely that marriage would appear in its name. (Blankenhorn 1995 p.314)

This paragraph of Blankenhorn's raises fundamental issues. Surely something strange is going on when at a conference of marriage and family therapists out of 277 topic areas, not even one is on marriage. In view of the great impact marriage has on health outcomes (and similar case could have been mounted for the impact of the marital relationship on children's learning achievement in school (Eastman 1994), combined with the need to monitor and evaluate the great changes taking place in marriage and family in the USA over the last thirty years, one could reasonably expect that a conference of marriage and family therapists would place marriage at the centre of its deliberations.

The question of why marriage was so avoided at the conference must also be explored. It stretches the imagination a little to speculate that it was just overlooked. It seems that a quite powerful dynamic must be operating to banish marriage from a conference of marriage and family therapists and that ambivalence and even hostility to marriage have to be considered. I have speculated elsewhere on what are the powerful dynamic(s) that cause marriage and family to be so avoided (Eastman 1989) but here it necessary to consider the possibility that an anti-marriage culture has developed in some sectors. And it must also be noted that negative attitudes to marriage are restricted to certain elites. Social science surveys consistently find that marriage is a major source of happiness and contentment, and is the highest value for most Australians.

There is evidence that this overlooking, ambivalence or even hostility to marriage is not restricted to the USA but is an issue also in Australia. For example in 1988 a director of one of the state marriage guidance councils reported that a senior government bureaucrat had said to him, 'The problem with the Marriage Guidance Council is the word marriage in your title' (Crawley 1988). Since then, of course, the offending word has been removed and the Council renamed Relationships Australia.

Further evidence of ambivalence or hostility to marriage is that very few universities have departments or courses on family studies (compared with, say, youth studies or women's studies) and even fewer train professionals in marriage education theory and skills or do research in this area. This lack is evident in the relative scarcity of papers on marriage education or other ways to promote marriage satisfaction and stability at AIFS conferences. Despite the importance of marriage (and especially happy marriage) in children's mental and physical health, in the mental and physical health of adults, in children's learning achievement in school, in preventing violence and crime and in community stability and peace, there is not one university in Australia that has a Professorial Chair in Marriage Studies or even in Family Studies.

But the cultural ambivalence toward marriage mentioned earlier (marriage not appearing as a topic at a conference of marriage therapists) must also be considered as one of the reasons that the family related variables, so clearly recognised in the data, have not been addressed in health strategy documents. The cultural ambivalence about marriage is fuelled by a number of widely accepted negative beliefs about marriage, beliefs that while most common among intellectual elites are in fact not founded on research and run counter to research and evidence. While they are commonly thought of as facts they are more accurately described as myths.

Myths of marriage involved in ambivalence towards, overlooking of marriage and family variables

One reason for marriage's marginal status is that a number of negative views of marriage are widely held. Some examples of these negative views are, for example, that marriage is a recent invention; that 'marriage was originally designed to facilitate both the maintenance of class inequality and the oppression of women' (Aulette 1994 p. 278); that the 'traditional' family has almost disappeared; that the two-parent family is a thing of the past; that marriage is good for men and bad for women; that marriage contributes to health and happiness for men, but has the opposite effect on women; that marriage makes women sick; that marriage typically involves violence and abuse; that marriage is a 'hitting licence'; that research finds absolutely no difference between children of divorce and children in intact families and that all family forms are equally beneficial and to be supported and that to propose to reduce the amount of family breakdown is actually to attack, demean and stigmatise those who have experienced marriage break-down. These views culminate in some overarching beliefs: that current trends towards high levels of marriage/relationship breakdown cannot and should not be reversed, that 'support' of marriage is of concern only to those of the extreme right--especially Christian fundamentalists.

Evidence to support the above views has been strongly challenged and evidence to refute them is extremely strong and constantly growing (Eastman 1996).

Some policy implications of the recognition of the impact of family related variables on health outcomes

Pre-marriage education should be further developed and researched and made more available

In most societies marriage is a rite of passage, and therefore a time and ritual that provides the opportunity to provide educative experiences. Evaluation has found very positive outcomes in lowered rates of marriage break-up and in greater satisfaction and happiness in marriage following premarriage programs. These should be researched and developed as much as possible.

Establish parent education and marriage and relationship education in alternative settings

Perhaps marriage counselling agencies should be required to provide a certain ratio of preventive services (parent education, pre-marriage education and counselling and marriage education and enrichment) to remedial or mediation services (counselling for persons already in crisis or serious conflict).

Pilot projects should be attempted to introduce such programs in TAFE colleges and universities--for their own students and staff and for community outreach programs. See the work at the University of South Australia where a short course of the impact of PhD studies on marriage is provided to students entering higher degree studies.

Workplaces are obvious places where marriage support programs should be introduced, perhaps as part of healthy workplace initiatives. Also, data on the impact of marriage and family on work productivity and how to create a workplace that is family friendly needs to be even more available to businesses and as part of business education.

Encourage Adult Education bodies to promote and offer education for marriage and parenthood as part of their regular stable of programs

When the impact of family dynamics on the outcomes of children's learning in school is considered, the argument for marriage and parenthood education being a major focus of adult education seems very strong (Eastman 1994).

Parent education, marriage and pre-marriage education should be part of the National Health Strategy

Marital status, family stability, parental disciplinary styles, the relationship between spouses (or partners) and parental efficacy have a powerful impact on the mental and physical health of both adults and children (Berkman and Syme 1979; Cassileth et al. 1985; Dominian 1991; Eastman 1996; Goodwin et al. 1987; House et al. 1988; Ladbrook 1990; Larson et al. 1995; Lee et al. 1987; Lilienfield et al. 1972; Mathers 1995; Morowitz 1975; National Health Strategy 1992; Reynolds and Rob 1988; Scientific Advisory Committee on Families and Mental Health 1995; Silburn et al. 1996c; Taskforce on Families in Western Australia 1995; Towers 1995; Walker 1996; Zubrick 1995). When this is taken into account, and also evidence of the effectiveness of preventive programs, the argument for parent education and pre-marriage and marriage education to be promoted as part of the National Health Strategy becomes very strong. It is recommended that such programs be funded and promoted in a way similar to anti-smoking and safe driving campaigns.

Preventive programs to support marriage and family should be a priority of health strategies

They are much more cost effective and also effective in preventing untold human pain and anguish. (See Dr Matt Sanders' work for an example of the effectiveness of prevention in the child mental health field (Scientific Advisory Committee on Families and Mental Health 1995; Scott 1995).)

This means relationship education in schools is important. But at least as important, and probably more important, is pre-marriage, early marriage and parenthood education. See Markman's evaluations of pre-marriage education's effectiveness (Markman and Hahlweg 1993) and Bader and Bader's earlier research on the same theme (Bader et al. 1980; Bader et al. 1981).

Establish marriage and family support programs and strategies as a research priority

For example, the ARC Grants Scheme sets priorities for research funding and it would be relatively easy to make research in areas related to marriage breakdown and family support a priority. Once consideration is given to the size and value of the domestic or household economy, to the importance of marital status and family dynamics as public health factors and to the contribution of parents to children's learning achievements in school, the case that these areas should be priority areas appears extremely strong. In fact it appears surprising that none of the words 'marriage', 'family' nor 'domestic or household economy' even appear on the ABS Field of Research Classification, or the ARC Socio-economic Objective Classification, let alone any of them ever having been a priority area.

Create incentives for universities to establish chairs in marriage or family studies, and courses with a family focus

There is a need for psychology courses based not on notions of human persons as 'abstracted individuals' but as members of social and cultural groups, with families as one of the most important of these groups. See for example the four-year BSS Family Studies course at Australian Catholic University, Christ Campus, Chadstone.

There is a need for Youth Studies courses with a family dimension.

There is a need for economics courses that include an awareness of the household economy and that focus on research on the household/domestic economy.

There is a need for business courses which focus on family business (as 80 per cent of Australian businesses are family businesses and 50 per cent of workers are employed in family businesses. Such courses should contain a recognition of the family variables that contribute or detract from the business' success.

Courses on public health and health promotion should give due consideration to data showing that marital status and family composition and family dynamics are at least as important in public health as issues such as smoking.

A critical perspective needs to be taken towards what is taught in schools about marriage

It is often suggrested that marriage or relationship education should be part of the school curriculum. But also schools need to be critical of what they teach about marriage. Recent research on College level text-books on marriage and the family in the USA finds serious flaws are common (Glenn 1996). A stated aim of some books is to 'dispel the myth' that the traditional nuclear family of husband, wife and their biological children is superior to one-parent families, stepfamilies or gay and lesbian families. (Empirical studies in the USA, UK and Australia all indicate that on many measures, such as children's mental health and adults' mental and physical health, the traditional nuclear family is superior.) Glenn finds that many books have a distinct bias against marriage and the traditional nuclear family and present an overly sanguine view of family change in modern societies, especially the increase in divorce rates and out-of-wedlock births. Data on the positive contribution made by marriage and family is almost entirely absent from these texts, despite considerable research data emerging on this topic in recent years (Glenn 1996).

Take seriously the cultural forces that undermine marriage and family and set out to create a culture that supports marriage

The USA has a much higher rate of marriage breakdown than does Australia. Up to now Australia has followed the US trends. The now-evident negative results of the high marriage breakdown rate has led many US social scientists to review their formerly sanguine view of the effect of high marriage breakdown rates. In fact now most US social scientists view current trends as negative. Australia has the opportunity of learning from US experience and putting in place policies that will prevent our breakdown rates ever rising to US levels.

Some ideas can be garnered from the USA where a number of groups are trying to reverse the very high rates of marriage breakdown in that country. See for example the following two bi-partisan congressional reports from the USA:

The Beyond Rhetoric Report: Beyond Rhetoric is the final report of the National Commission on Children, chaired by Senator John Rockefeller of West Virginia. (Senator Rockefeller is a prominent Democrat and the National Commission on Children was a bi-partisan commission with 24 members appointed by Democratic Congressional leaders and 12 by the Republican White House.) Beyond Rhetoric, Report of the National Commission on Children, USA was issued in June 1991.

Families First: Report of the National Commission on America's Urban Families. In January 1993 another US bi-partisan committee agreed in significant ways with the Beyond Rhetoric findings. The National Commission on America's Urban Families was established by President George Bush following a January 1992 meeting with a group of US mayors. In his State of the Union Address the President reported that the mayors said that everyone of them, Republican and Democrat, agreed on one thing: 'that the major cause of the problems of the cities is the dissolution of the family'. Responding to their concern, the President announced in his address his decision to establish a National Commission on America's Urban Families. Almost two years later the final report was presented to the President.

Report on Marriage in America: A Report to the Nation, The Council on Families in America, March 1995. The Report offers recommendations of strategies that various organisations could implement in order that young Americans can achieve strong and healthy family relationships, and in the Report's view that will mean making marriage stronger. They offer strategies to religious leaders and organisations; to civic leaders and community organisations; to employers, to social work, health care and other human service professionals; to marriage counselors, family therapists, and family life educators; to pregnancy health care providers and counselors; to family law attorneys and judges; to children's advocates; to teachers, principals, and leaders in education; to foundation executives and philanthropic leaders; to family scholars; to print and broadcast media journalists and editors; to entertainment industry writers, producers, and executives; to local, state, and federal legislators; to the general public.

Conclusion

Addressing the overlooking of family related variables in national health strategies may involve some re-thinking of the place of marriage in society. But it also offers the potential for marked improvements in the mental and physical health of both adults and children, and a consequent reduction in the burden of ill-health and premature mortality on individuals, families and the entire society.

References

ABS 1992. 'Social Indicators Number 5, Cat. No. 4101.0' . Canberra.

ABS 1994. 'Focus on Families: Demographics and Family Formation' .

Aulette, J.R. 1994. Changing Families. Belmont, CA: Wadsworth.

Bader, E., Microys, G., Sinclair, C., Willett, E. and Conway, B. 1980. 'Do marriage preparation programs really work? A Canadian experiment'. Journal of Marital and Family Therapy 6: 171-179.

Bader, E., Riddle, R. and Sinclair, C. 1981. 'Do marriage preparation programs really work? A five year study'. Family Therapy News.

Berkman, L.F. and Syme, S.L. 1979. 'Social networks, host resistance, and mortality: A nine-year follow-up study of Alameda County residents'. American Journal of Epidemiology 109: 186-204.

Blankenhorn, D. 1995. Fatherless America: Confronting Our Most Urgent Social Problem. New York: Basic Books (A Division of HarperCollinsPublishers).

Cassileth, B., Lusk, E.J., Miller, D.S., Brown, L.L. and Miller, C. 1985. 'Psychosocial correlates of survival in advanced malignant disease?'. New England Journal of Medicine 312: 1551-55.

Commonwealth Department of Human Services and Health 1994. 'Better Health Outcomes for Australians' . Canberra: Commonwealth Department of Human Services and Health.

Crawley, J. 1988. 'Helping families stay together' Family chaos and social welfare. Perth: Australian Institute of Public Policy.

Crockett, M. and Tripp, J. 1994. 'Children living in re-ordered families' . York: Joseph Rowntree Foundation.

Dominian, J. 1991. 'Marital Breakdown and the Health of the Nation' . London: One Plus One.

Eastman, M. 1994. 'The role of family education in Adult Education'. Australian Journal of Adult and Community Education 34: 195-205.

Eastman, M. 1996. 'Myths of marriage and family' in Popenoe, D., Elshtain, J.B. and Blankenhorn, D. (eds.) Promises to Keep : Decline and Renewal of Marriage in America. Lanham, MD: Rowman and Littlefield.

Eastman, M.B. 1989. Family the Vital Factor. Melbourne: CollinsDove.

Glenn, N. 1996. 'A critique of twenty family and marriage and the family textbooks' Council on Families in America. The Brandeis House, 12 East 77th Street.

Goodwin, J.S., Hunt, W.C., Key, C.R. and Samet, J.M. 1987. 'The effect of marital status on stage, treatment, and survival of cancer patients'. JAMA 258: 3125-30.

Harris, R., Willis, P. and Barrie, A. 1992. Love, Sex and Waterskiing: The Experience of Pre-marriage Education in Australia. Adelaide: University of South Australia Centre for Human Resource Studies.

House, J.S., Landis, K.R. and Umberson, D. 1988. 'Social relationships and health'. Science 241: 540-44.

Hu, Y. and Goldman, N. 1990. 'Mortality differentials by marital status: an international comparison'. Demography 27: 233-250.

Ladbrook, D. 1990. 'Sex differentials in premature death among professionals (Parts 1 and 2)'. Journal of the Australian Population Association 7: 1-26, 89-115.

Larson, D.B., Swyers, J.P. and Larson, S.S. 1995. 'The Costly Consequences of Divorce: Assessing the Clinical, Economic, and Public Health Impact of Marital Disruption in the United States' . Rockville, Maryland: National Institute for Healthcare Research.

Lee, S.H., Smith, L., d'Espaignet, E. and Thomson, N. 1987. Health Differentials Among Working Age Australians. Canberra: Australian Institute of Health.

Lewis, J.M., Beavers, W.R., Gossett, J.T. and Phillips, V.A. 1976. No Single Thread: Psychological Health in Family Systems. New York: Brunner/Mazel.

Lilienfield, A.M., Levin, M.L. and Kessler, M.J. 1972. Cancer in the United States. Cambridge: Harvard University Press.

Markman, H.J. and Hahlweg, K. 1993. 'The prediction and prevention of marital distress: an international perspective'. Clinical Psychology Review 13: 29-43.

Mathers, C. 1994a. Health Differentials Among Australian Adults Aged 25-64 years: Health Monitoring Series No. 1. Canberra: Australian Institute of Health and Welfare: AGPS.

Mathers, C. 1994b. Health Differentials Among Older Australians: Health Monitoring Series No. 2. Canberra: Australian Institute of Health and Welfare & AGPS.

Mathers, C. 1995. Health Differentials Among Australian Children: Health Monitoring Series No. 3. Canberra: Australian Institute of Health and Welfare: AGPS.

Mathers, C. 1996. Health Differentials Among Young Australian Adults: Health Monitoring Series No. 4. Canberra: Australian Institute of Health and Welfare & AGPS.

Morowitz, H.J. 1975. 'Hiding in the Hammond Report'. Hospital Practice.

National Health and Medical Research Council 1995. 'Health Australia: Promoting Health in Australia' . Canberra: National Health and Medical Research Council.

National Health Strategy 1992. 'Enough to Make you Sick: How Income and Environment Affect Health' . Melbourne: National Health Strategy.

National Mental Health Policy 1992. 'National Mental Health Policy' . Canberra: National Mental Health Project Funding: Innovative Grants Program/Australian Health Ministers Advisory Council, National Working Group on Mental Health Policy [and] Mental Health Branch, Commonwealth Department of Human Services and Health.

Nutbeam, D., Wise, M., Bauman, A., Harris, E. and Leeder, S. 1993. 'Goals and Targets for Australia's Health in the Year 2000 and Beyond' . Sydney: Department of Health, University of Sydney for the Commonwealth Department of Health, Housing and Community Services.

Nutbeam, D., Wise, M., Bauman, A. and Leeder, S. 1992. 'Review and Revision of the National Health Goals and Targets' . Sydney: University of Sydney.

Reynolds, I. and Rob, M.I. 1988. 'The role of family difficulties in adolescent depression, drug-taking and other health problem behaviours'. The Medical Journal of Australia 149: 250-256.

Sanders, M. and Dadds, M. 1993. Behavioural Family Intervention. Boston: Allyn and Bacon.

Sanders, M. and Markie-Dadds, C. 1992. 'Towards a technology of prevention of disruptive behaviour disorders: the role of behavioural family intervention'. Behaviour Change 9: 186-200.

Scientific Advisory Committee on Families and Mental Health 1995. Healthy Families: Healthy Nation. Brown Hills Qld: Australian Academic Press.

Scott, L. 1995. 'Family mental health targeted' The .Australian. Sydney.

Silburn, S. and al, e. 1996. 'Family functioning' Western Australian Child Health Survey: Family and Community Health. Perth: ABS and TVW Telethon Institute for Child Health Research.

Silburn, S., Zubrick, S. and Garton, A. 1996a. 'Family and community health' Western Australian Child Health Survey: Family and Community Health. Perth: ABS and TVW Telethon Institute for Child Health Research.

Silburn, S., Zubrick, S., Garton, A., Gurrin, L., Burton, P., Dalby, R., Carlton, J., Shepherd, C. and Lawrence, D. 1996b. 'Family health and mental health' Western Australian Child Health Survey: Family and Community Health. Perth: ABS and TVW Telethon Institute for Child Health Research.

Silburn, S.R., Zubrick, S.R., Garton, A., Gurrin, L., Burton, P., Dalby, R., Carlton, J., Shepherd, C. and Lawrence, D. 1996c. 'Western Australian Child Health Survey: Family and Community Health' in P.C. Kelly, D.C.S. (ed.) . Perth, Western Australia: Australian Bureau of Statistics.

Taskforce on Families in Western Australia 1995. 'W.A. Families our Future' . Perth: Ministry for Community y Development.

Towers, K. 1995. 'Chronic pain linked to unhappy childhood' The Weekend Australian. Sydney.

Vichealth and Health and Community Services 1995. 'Healthy Victorians 2000: Towards Victorian Health Promotion Goals and Targets' . Melbourne: Vichealth: Victorian Health Promotion Foundation.

Waite, L. 1995. 'Does marriage matter?'. Demography 32: 483-507.

Walker, M. 1996. 'Marital discord 'damages health'' The Age. Melbourne.

Walsh, F. 1982. Normal Family Processes. New York: Guildford.

Zubrick, S.R.et.al. 1995. Western Australian Child Health Survey: Developing Health and Well-being in the Nineties. Perth: Australian Bureau of Statistics and the Institute for Child Health Research.