INTRODUCTION

In Australia and elsewhere there is an assumption thatthere are certain health benefits to living in the country:clean air, fresher food, better housing, less stress and agreater sense of community and support.1,2 Recent Aus-tralian statistics show otherwise. People in rural andremote areas suffer from a health differential that isskewed toward higher mortality and morbidity rates forsome diseases and increased rates of hospitalisation.1

‘Living and working in the country, especially the mostremote parts of Australia, is a health hazard. The air maybe cleaner than in the cities, the roads emptier, the noiselevels lower, but the living is hazardous, especially foryoung men’ (p. 1).3 Other research shows that the healthgains by indigenous Australians, the majority of whom

live in rural and remote areas, are inconsistent and thereappears little improvement in death rates.4

While rural–metropolitan health variations are beingdocumented, there is very little explanation of the differ-ences. Given the explosion in interest in what are beingcalled the social determinants of health, it seems timely toexplore the range of factors that interact to produce ruralhealth status. Very briefly, the social determinants ofhealth refer to the economic, social and cultural factorsthat influence individual and population health bothdirectly and indirectly, through their impact on psychoso-cial factors and biophysiological responses. This field ofresearch departs from explanations of genetic inheritance,biomedical processes and behaviours to address the con-text in which individuals live, work and play.5

The present paper uses insights from the social deter-minants of health literature to offer tentative explanationsfor the major health differentials between rural and metro-politan Australia. We conclude with several priorities forresearching the rural–metropolitan health differential.

Aust. J. Rural Health (2000) 8, 254–260

Correspondence: Jane Dixon, National Centre for Epidemi-ology and Population Health, Australian National University,ACT 0200, Australia. Email: jane.dixon@anu.edu.au

Accepted for publication June 2000.

RESEARCHING THERURAL–METROPOLITAN HEALTHDIFFERENTIAL USING THE ‘SOCIALDETERMINANTS OF HEALTH’

1Health Inequalities Research Collaboration, National Centre for Epidemiology and PopulationHealth, Australian National University, Canberra, Australian Capital Territory, Australia and2Department of Psychology, The University of Waikato, Hamilton, New Zealand

ABSTRACT: Recent research indicates that the health status of rural people is inferior to that of people living inmetropolitan Australia. This paper summarises the rural–metropolitan health differential and turns to the field ofresearch being called the social determinants of health for explanations of rural health inequalities. The paperexplores the ways in which psychosocial factors can interact with material, behavioural and sociocultural factors tocontribute to health outcomes. It suggests that the concepts of place and rurality may be useful in future research onthe determinants of population health. Further research issues are identified that need to be addressed if we are tounderstand the complexities of rural health disadvantage.

KEY WORDS: research, rural health inequalities, social determinants.

Jane Dixon1 and Nicky Welch2

Review Article

RURAL–METROPOLITAN HEALTH DIFFERENTIAL: J . DIXON AND N. WELCH 255

DESCRIBING RURAL–METROPOLITANHEALTH DIFFERENTIALS

The National Rural Health Policy Forum argues that forsome conditions and outcomes, the health of rural Aus-tralians is ‘substantially worse than for other Australians’(p. ii).6 In particular, rates of death by suicide and injury,road vehicle accidents, asthma, diabetes and infantmortality are notably higher than those experienced inmetropolitan areas.1 Higher rates of death are recorded fornon-indigenous men in rural and remote areas than forthose in metropolitan Australia.1 Cancer rates are also dif-ferentially distributed according to place of residence,with lower rates recorded for rural Australians. This sec-tion provides more detail for the major health statusdifferences.

The statistics for diabetes show significant differencesbetween metropolitan and rural areas.1 Hospitalisationdata show that both males and females living in remotezones and males living in rural zones are hospitalised fordiabetes twice as often as their metropolitan counter-parts.1 Females in rural zones have a rate of hospital-isation for diabetes that is 25% higher than females inmetropolitan areas. What is disturbing is the fact that thediabetes differential is growing.4 Deaths due to asthma aresimilarly skewed, with deaths occurring at higher rates inremote zones than in rural zones, which are in turn alsohigher than those in metropolitan areas.1

Rates of injury are shown in Fig. 1 to be significantlydifferent in metropolitan areas and rural and remoteregions, with deaths per 100 000 population rising from

53 in capital cities to death rates at 77.5 deaths per

100 000 population in rural areas.1 Female rates are

markedly lower overall but still follow the trend of more

deaths due to injury as the geographical spread of popula-

tion gets wider.1 ‘Overall, the rates increase with increas-

ing remoteness, suggesting that those living in rural and

remote zones are at greater risk of death from injury than

are those living in the metropolitan zone’ (p. 17).1

Suicide rates in Australia have remained relatively

constant over the last 100 years, but the highest rates of

male suicide and self-inflicted injuries are found in large

rural centres and ‘other remote’ areas.1 Fig. 2 reveals a

relatively uneven urban–remote gradient for suicide.

An absence of national time series data precludes

analyses of whether rural–metropolitan health differen-

tials are increasing. However, in one state-based study,

Burnley showed that while mortality rates from ischaemic

heart disease (IHD) in NSW declined between 1969 and

1994, spatial variations were maintained and had

increased among middle-aged men in some areas; specifi-

cally, the incidence of IHD had increased for men aged

40–64 years in inland small towns/rural areas but not in

the coastal urban areas of NSW.7 Burnley could not

explain this particular trend and surmised that lifestyle

differences and a shortage of health services and health

workers were responsible. He concluded, however, with a

call for research to investigate the interaction between

cultural/behavioural, social causation and structural/

material factors.7

FIGURE 1: Death rates for all causes of injury, 1992–1996.(�), Male; (�), female.

FIGURE 2: Death rates for suicide and self-inflicted injury,1992–1996. (�), Male; (�), female.

256 AUSTRALIAN JOURNAL OF RURAL HEALTH

EXPLAINING THE HEALTH INEQUALITIESBETWEEN RURAL AND METROPOLITANAUSTRALIA

Considerable consensus exists among epidemiologistsabout the reasons for health differentials. In particular,lifestyle risk factors, physical environmental factors andhealth service access and utilisation are repeatedly cited.The social determinants of health literature nominatesother relevant factors, namely, socioeconomic status, raceand ethnicity, gender, sociocultural and psychosocialfactors. The present paper uses an expanded view of thesocial determinants of health to suggest explanations forthe poorer health status of rural Australians.

Socioeconomic statusIn a recent Australian overview of the research on socio-economic status (SES) and health, the authors concludedthat ‘the evidence on socioeconomic status and health inAustralia is unequivocal: those who occupy positions atlower levels of the socioeconomic hierarchy fare signifi-cantly worse in terms of their health’ (p. 33).8 Socio-economic status is generally measured in one of threeways: (i) income; (ii) education; and (iii) occupation.

When SES is measured by income, clear associationsare apparent with health: Australians living on lowincomes are more likely to suffer disability, chronic ill-nesses or report recent illness.9,10 While it is difficult tocompare rates of income across geographical boundariesbecause of the different make-up of income between ruraland metropolitan areas, it is possible to argue that the dif-ferent distribution of incomes in urban and rural Australiacontributes to the health differential on the basis of:(i) mean annual taxable income; (ii) the proportion offamilies with dependent children receiving governmentpensions and benefits; (iii) the proportion of children liv-ing in low-income working families and adults aged olderthan 25 receiving labour market benefits.11 Non-metropol-itan Australia contains a greater percentage of people whoare significantly poorer than does metropolitan Australia.The exceptions are mining towns and wine producingregions.11

Similar findings arise when education and occupationare used as indicators of SES. While the number of peoplewith tertiary qualifications in rural Australia has grownsignificantly, educational levels are still lower than formetropolitan Australia. Other research has shown thatunemployment increases the risk of premature death fromsuicide, cardiovascular disease and respiratory disease,particularly among men.12 Longitudinal studies have alsoshown that unemployment causes mental ill health and

the greater use of health services.12 Higher rates of un-employment in many rural and regional centres mayexacerbate the poorer health status in those areas.

Race and the indigenous health differentialThe health status of indigenous Australians is signifi-cantly worse than that of the non-indigenous populationon every health indicator: life expectancy, maternal mor-tality, infant mortality, child mortality and childhood andadult morbidity.1,13 It is conceivable that the entirerural–metropolitan health differential could be attribut-able to the fact that indigenous Australians make up agreater proportion of rural populations. However, this isnot the correct explanation. While in remote areas, theindigenous population makes up a greater percentage ofthe total population and therefore influences the lowerhealth status of remote areas, the indigenous population isnot large enough in metropolitan and rural zones to affectthe health differential.1

Demographic composition aside, research conductedin the US does not support claims that race as a biologicalcharacteristic leads to health inequalities.14 Nor does thefact that racial minorities are generally poorer, becausepoverty is a predictor of mortality for all racial groups.Instead, those who have investigated the interplaybetween income, race and other factors show that per-ceived discrimination and race-related stress play a role.In other words, racism as a sociocultural characteristic isthe health-damaging factor. Furthermore, contrary to pop-ular opinion, behavioural risk accounts for only a smallproportion of income-related health disparities across age,sex and race categories.14–16 Thus it is not possible toassert that the behaviours of particular races and ethnicgroups are responsible for differential health status.

Environmental factorsBoth the greater exposure to occupational injury and thepoorer road quality in rural and remote areas possibly addto the rural health differential. Agriculture, as a singlyrural industry, is one of the highest risk groups for occupa-tional injury and disease. Both the farming environmentitself and the diversity of production processes that arecarried out daily on farms, contribute to the high rate ofinjury.17 Poor road quality combines with greater periodsspent on the road to increase the potential for road acci-dents. A diminished police presence may compound theproblem.3

Risk-taking behavioursWhile the high rate of injury in rural and remote areas ispartly due to environmental factors, behavioural factors

RURAL–METROPOLITAN HEALTH DIFFERENTIAL: J . DIXON AND N. WELCH 257

question of whether this is solely because of a lack ofquality services or because of behavioural or socioculturalfactors that inhibit utilisation is yet to be answered. Somelight is shed on what may be considered sociocultural fac-tors by research which reveals the different understand-ings that rural people have of health. It appears that ‘thepotential consumer’s willingness to seek care depends, inpart, on an individual’s attitudes towards health, knowl-edge about health care, learned definitions of illness(social and cultural) and perceptions of need for healthcare service’ (p. 61).26 Rather than concerns over pain orcosmetic attractiveness, ‘maintaining performance or pro-ductivity, despite adversity, is an important concept forwell-being amongst rural dwellers’ (p. 63).18

Definitions of health and wellness subsequently affectthe utilisation of health-care services. People in ruralareas commonly describe health in the negative, as anabsence of disease.26 If one understands health to be anabsence of disease, the main concern becomes the cure ofillness as opposed to the maintenance of good health.Therefore, curative treatment becomes the focus of ahealth-care system20 and demand is made for acute andchronic disease management as opposed to primary careand health promotion.27 Moreover, a demand for preven-tive services is negligible and this is evidenced in muchdental and oral health research.28–30

Psychosocial factorsA psychosocial factor has been defined as ‘a measurementthat potentially relates psychological phenomena to thesocial environment and to pathophysiological changes’(p. 1460).31 In relation to coronary heart disease (CHD),psychosocial factors have been argued to operate alongthree pathways. First, they may catalyse health-relatedbehaviours such as smoking, diet, alcohol consumptionand physical activity. Second, they may act directly onpathophysiological pathways. Third, they may mediateresponses to assistance (as illustrated previously in rela-tion to consumer understandings of health and wellbeing).

Because they have been implicated in CHD, one dis-ease that may be increasing in rural Australia,6 we offersome speculative comment about the potential for thepsychosocial factors of social support and social status toassist in the explanation of health inequalities.

Social support can be defined as a social resourceprovided by another person, or the degree to which thecomfort and esteem needs of a person are met.32,33 Suchsupport can come from a variety of sources, includingfamily, friends, workmates, the family doctor and commu-nity nurse or community organisations.33 A well-knownstudy demonstrates the benefits of social support to health

also play a role. A contributing factor in the high rate ofinjuries suffered by rural males is acknowledged as risk-taking behaviour, which in turn affects their driving. ‘Sen-sation seeking and aggression have been found to be themain reasons adolescent men drive recklessly’ (p. 2).3

Anecdotal evidence suggests that risk-taking behav-iour has another dimension. An attitude is prevalent inrural areas that may not encourage preventive healthbehaviours and may in fact tolerate smoking and exces-sive drinking.5 Attitudes that emphasise the need to main-tain the ability to perform one’s role and stoicism towardadversity, are common in rural communities.18 Regardlessof involvement in the agricultural industry, attitudes suchas self-reliance, independence and a reluctance to seekhelp, are displayed by residents of rural communities.19

Like the type A personality, the much admired rural per-sonality may be a health risk factor.

Physical and cultural access to servicesThe reality of living in rural and remote areas of Australiais that there are fewer health-care services.1 However, itshould be noted that geographical distance to a practi-tioner has been found to be neither the sole nor the mostimportant determinant of choice of general practicecare.20

Geographic isolation and problems with access to andshortages of providers and services are multidimensionalproblems. For instance, the quality of roads not only con-tributes directly to higher incidences of injury it also com-promises access to health services. Bond noted thedifficulties in utilising services in rural and remote areasbecause of poor quality roads and the added expense oftravel because of the high cost of petrol.21 A lack of trans-port is a further barrier to services22 and while varioustypes of government assistance in the form of reimbursingcosts of utilising a service do help, the reality is thatallowances fall short of the true cost of taking time offwork and associated social costs.1 Moreover, difficult eco-nomic circumstances indisputably impact upon access toand demand for health services, particularly rehabilitationservices.17 This is especially relevant during the economicdownturn that rural areas have been subject to lately.10,23

A difficulty in meeting the health-care service needsof people in rural areas is their high expectations of whata health service should comprise. These expectationsappear to be media-driven24 and have, in some cases,been fuelled by rural health policies that promote doctorshortages and hospital closures as the only concern forrural health.25

Medicare data indicate that rural people utilise healthservices less than people in metropolitan areas do.1 The

258 AUSTRALIAN JOURNAL OF RURAL HEALTH

‘The most pressing aspect of relative deprivation andlow relative income is less the shortage of the materialgoods which others have, as the low social status and thedesperate lack of sources of self-esteem which usuallygoes with it. If social cohesion matters to health, then per-haps the component of it which matters most is thatpeople have positions and roles in society which accordthem dignity and respect … Respect affects how we aretreated, what help from others is likely, what economicarrangements others are willing to engage in with us, whenreciprocity is to be expected’ (p. 34).39

In a provocative treatment of whether political or eco-nomic equality matters, Anne Phillips argues that respectincreasingly arises from definitions of economic loca-tion.40 Phillips finds that, as well as pursuing access toresources and influence over political decisions, groupsare increasingly pursuing claims for recognition. Whatgroups want recognised is the equal worth of their ways ofdoing things, their value systems and contributions to thesociety. Phillips describes status injuries ‘that arise out ofbeing denied the status of full partner in social interac-tions or being prevented from participating as a full equalin social life’ (p. 88).40 Status injuries have most oftenbeen associated with sexism, heterosexism and racism.Perhaps status injuries need to be extended to places.Rather than classifying people according to high and lowstatus occupations, researchers might consider rankingplaces according to media coverage, in-migration andbusiness relocation and popular perceptions of the reputa-tion of the town or area. High and low status places couldthen be correlated with differential health status.

IMPORTANCE OF PLACE AND OFRURALITY

While place of residence has been used as a proxy forSES, particularly in Australia, it has been itselfresearched to a lesser degree than the other factors. Thishas obscured the rural–urban dimension to healthinequalities. A place may be thought of as a location inwhich social relations are constituted.2 Indeed, instead ofthinking about the rural personality as a risk factor, on thebasis of differential health status it is more plausible toopt for rural place as a risk factor.

Macintyre (quoted in Curtis) has summarised how shesees the influences of place on health.2 The componentsof place are the physical environment, availability ofhealthy environments, services provided, socioculturalfactors of the locale, representation of locale, lay systemsof beliefs and behaviours and labour markets.2 They are amix of economic, physical, social, environmental and

status. Based in Alameda County, California, Berkmanand Syme randomly sampled 6928 adults and followed upthe study 9 years later. Their findings indicated thatpeople who lack ties to the community were more likely todie than those with support networks.34

Figures 1 and 2, which were used earlier to illustrateunequal health outcomes in rural and urban Australia,offer indirect credibility for the importance of social sup-port. The data indeed revealed a more stark differential,namely a gender divide. Part of the explanation for thegendered injury gradient could be the more hazardousoccupations undertaken by men. The glaring suicide dif-ferential cannot be so explained but nor can the fact thatthe rural–urban gradient is reversed for women, with citywomen being more at risk than women in rural and remotezones. It is tempting to invoke social support mechanismsto explain women’s better health outcomes and to suggestthat female appropriate social support is stronger in ruralAustralia. In short, ‘what has value as support can varyaccording to situations, categories of people, contexts andcultural values’ (p. 290).35

However, while rural communities are often known fortheir relatively high levels of cohesion and support, thiscan be reflected in a negative way. If social support offersa protective advantage, the question remains as to whysuicide rates in large rural towns and the remotest parts ofAustralia are so high. Stevens asks whether it is possiblethat social rules operate to limit people’s behaviour and todiminish opportunities for support.19 In illustration of thispoint, Brown and co-authors describe difficulties withhealth service providers being personally known by con-sumers,36 while Warr and Hillier identify privacy issuesover accessing adolescent sexual information in a smalltown.37

Social status is another psychosocial factor that isreceiving epidemiological attention. In cross-nationalcomparative work, Wilkinson concluded that over acertain threshold, income levels are not related to healthstatus in OECD countries. Instead, income as a proxy forsocial status is the important factor.38 Much interest insocial status derives from research on the physiologicalconsequences of social hierarchy among monkeys. Atleast two different research teams have demonstrated alink between low social status and raised levels of basalcortisol leading to artherosclerosis, the precursor to CHD.Subordinate monkeys also suffered obesity, depressionand poorer immune function. Being low in the status hier-archy of the monkey colony is a health hazard; and so it isin humans. In research on homicide in US cities, a clearlink was made to the relative deprivation and poor socialstatus of the assailants. The researchers concluded that:

RURAL–METROPOLITAN HEALTH DIFFERENTIAL: J . DIXON AND N. WELCH 259

sociocultural factors. The experience of these factors isincreasingly shown to have psychosocial repercussionsand to mediate how different groups define health andsubsequently engage in health-promoting behaviours.41

What few have done is to tease out the various dimen-sions to place, whether rural or urban. Does, for example,place refer to area effects, community effects or the effectsof social practices engaged in by an area’s residents?What is it about rural places or the rural experience thatcontributes to differential health outcomes? How is rural-ity embodied, resulting in various physical and mentalstates of health and wellbeing? These questions need tobe answered before we can explain the rural–metropolitandifferences that have been described in this paper.

CONCLUSION

This paper raises a number of issues with regard to thecurrent status of health experienced by people in ruraland remote areas. With researchers asserting that shame,respect and self-esteem impact on individual healthstatus, it is possible that the health status of rural com-munities is not improving as fast as that of metropolitanAustralia because of relativities in community socialstatus. Is it not plausible that whole rural communitiesfeel shamed and low in self-esteem because they are rep-resented as having value only as contributors to the GDP?

Such questions highlight the importance of adequatelyresourced research and development to inform ruralhealth policy. The call for establishing a set of bench-marks against which urban–rural differentials in healthstatus can be monitored is imperative.24 Other researcherssee the need for more research into definitions of health,specifically how different groups define health. In order totease out the important pathways to health, we alsorequire multilevel analytic techniques, using models thatcan incorporate qualitative as well as quantitative data.Furthermore, if place is a determinant of health, we needto identify how individuals embody aspects of place,including rural places. Answering this question wouldrequire interdisciplinary teamwork that has, to date, beenmissing in much health research.

REFERENCES

1 Australian Institute of Health and Welfare. Health in Ruraland Remote Australia. Canberra: Australian Institute ofHealth and Welfare, 1998.

2 Curtis S, Rees Jones I. Is there a place for geography in theanalysis of health inequality? In: Bartley M, Blane D,

Davey Smith G. (eds). The Sociology of Health Inequalities.Oxford: Blackwell Publishers, 1998; 85–112.

3 National Rural Health Alliance. Drugs and Alcohol inRural Australia. Rural Health Information Paper No. 4,August 1998. Canberra: National Rural Health Alliance,1998.

4 Australian Bureau of Statistics. The Health and Welfare ofAustralia’s Aboriginal and Torres Strait Islander Peoples.Cat no. 4704. Canberra: Australian Bureau of Statistics,1997.

5 Marmot M, Wilkinson R (eds). Social Determinants ofHealth. Oxford: Oxford University Press, 1999.

6 National Rural Health Policy Forum. Healthy Horizons:A Framework for Improving the Health of Rural, Regionaland Remote Australians. Canberra: National Rural HealthPolicy Forum and National Rural Health Alliance for theAustralian Health Ministers’ Conference, 1999.

7 Burnley IH. Inequalities in the transition of ischaemicheart disease mortality in New South Wales, Australia.1969–94. Social Science and Medicine 1998; 47:1209–1222.

8 Turrell G, Oldenburg B, McGuffog I, Dent R. Socioeconomicdeterminants of health: Towards a national research pro-gram and a policy and intervention agenda. Brisbane:Queensland University of Technology School of PublicHealth, 1999.

9 National Health Strategy. Enough to Make You Sick. HowIncome and Environment Affect Health. Melbourne:National Health Strategy, 1992.

10 Hupalo P, Herden K. Health Policy and Inequality. Depart-ment of Health and Aged Care Occasional Papers. NewSeries no. 5, 1999. Canberra: Department of Health andAged Care, 1999.

11 Haberkorn G, Hugo G, Fisher M, Aylward R. CountryMatters: Social Atlas of Rural and Regional Australia.Canberra: Commonwealth of Australia, Bureau of RuralSciences, 1999.

12 Mathers C, Schofield D. The health consequences of unem-ployment: The evidence. Medical Journal of Australia1998; 168: 178–182.

13 National Health and Medical Research Council. Promotingthe Health of Indigenous Australians. A Review of Infra-structure Support for Aboriginal and Torres Strait IslanderHealth Advancement. Canberra: National Health andMedical Research Council, 1997.

14 Williams DR, Yu Y, Jackson JS, Anderson NB. Racialdifferences in physical and mental health. Journal ofHealth Psychology 1997; 2: 335–351.

15 McDonough P, Duncan GJ, Williams D, House J. Incomedynamics and adult mortality. American Journal of PublicHealth 1997; 87: 1476–1483.

16 Lantz PM, House JS, Lepowski JM, Williams DR, Mero RP,Chen J. Socioeconomic factors, health behaviors, andmortality. Journal of the American Medical Association1997; 279: 1703–1708.

260 AUSTRALIAN JOURNAL OF RURAL HEALTH

Australian private general practice. Australian Journal ofRural Health 1998; 6: 12–17.

30 Short L, Patterson A. Strategies to improve equity in oralhealth care for the rural aged. In: Second BiennialAustralian Rural and Remote Health Scientific ConferenceProceedings. Toowoomba, August 1994. Toowoomba:Darling Downs Health Services Foundation and RegionalHealth Authority.

31 Hemingway H, Marmot M. Psychosocial factors in theaetiology and prognosis of coronary heart disease: System-atic review of prospective cohort studies. British MedicalJournal 1999; 318: 1460–1467.

32 Rosenfeld E. Social support and health status: A literaturereview. Adelaide: South Australian Community HealthResearch Unit, 1997.

33 Sarafino EP. Health Psychology: Biopsychosocial Inter-actions. New York: John Wiley and Sons, 1998.

34 Berkman L, Syme L. Social networks, host resistance andmortality: A nine year follow-up of Alameda County resi-dents. American Journal of Epidemiology 1979; 109:186–204.

35 Corin E. The cultural frame: Context and meaning in theconstruction of health. In: Amick BC, Levine S, Tarlov AR,Chapman Walsh D (eds). Society and Health. Oxford:Oxford University Press, 1995; 272–304.

36 Brown WJ, Young AF, Byles JE. Tyranny of distance? Thehealth of mid-age women living in five geographical areasof Australia. Australian Journal of Rural Health 1999; 7:148–154.

37 Warr D, Hillier L. That’s the problem with living in a smalltown: Privacy and sexual health issues for young ruralpeople. Ausralian Journal of Rural Health 1997; 5:132–139.

38 Wilkinson R. Putting the picture together: Prosperity, redis-tribution, health, and welfare. In: Marmot M, Wilkinson R(eds). Social Determinants of Health. Oxford: Oxford Uni-versity Press, 1999; 256–274.

39 Wilkinson R, Kawachi I, Kennedy B. Mortality, the socialenvironment, crime and violence. In: Bartley M, Blane D,Davey Smith G (eds). The Sociology of Health Inequalities.Oxford: Blackwell Publishers, 1998; 19–37.

40 Phillips A. Which Equalities Matter? Cambridge, UK:Polity Press, 1999.

41 Wyn J, Stokes H, Stafford J. Challenges and resilience fac-tors for young people living in rural Australia in the 1990s.In: Gregory G, Murray D (eds). National Rural PublicHealth Forum Proceedings. 12–15 October 1997. Canberra:National Rural Health Alliance, 1998; 234–243.

17 Broughton K, Fragar L, Coleman R. Rehabilitation servicesfor injured and disabled farmers. In: Fourth National RuralHealth Conference Papers. Perth, Western Australia 9–12February 1997. Perth: Promaco Conventions, 1997;327–333.

18 Elliot-Schmidt R, Strong J. The concept of well-being in arural setting: Understanding health and illness. AustralianJournal of Rural Health 1997; 5: 59–63.

19 Stevens S. Nursing culture, bush culture: Change and sur-vival. In: Fourth Biennial Australian Rural and RemoteHealth Scientific Conference Proceedings. Toowoomba,27–28 August 1998. Toowoomba: Cunningham Centre andToowoomba Hospital Foundations, 1998; 2.117–2.124.

20 Humphreys JS, Mathews-Cowey S, Weinand HC. Factors inaccessibility of general practice in rural Australia. MedicalJournal of Australia 1997; 166: 577–580.

21 Bond A. Consumer concerns. In: Malko KN (ed.). SecondNational Rural Health Conference Proceedings. Armidale,12–14 February 1993. Armidale: National Rural HealthAlliance and School of Health, University of New England,1993; 21–24.

22 Humphreys JS. Rural health and the health of rural com-munities. Worner Lecture Series, 1998 (on-line).http://www.bendigo.latrobe.edu.au/rahdo/research/worner98.html

23 Humphreys JS. Past conferences, future strategies. Addressgiven to Fifth National Rural Health Conference. Adelaide.14–17 March 1999. http://ruralhealth.org.au/fifthconf/summaries.htm

24 Humphreys JS. Delimiting ‘rural’: Implications of an agreed‘rurality’ index for health care planning and resource allo-cation. Australian Journal of Rural Health 1998; 6:212–216.

25 Keleher H. Rural public health matters. Australian andNew Zealand Journal of Public Health 1999; 23: 342.

26 Humphreys J, Rolley F. Health and health care in ruralAustralia: A literature review. Armidale: University of NewEngland, 1991.

27 Coster G, Gribben B. Primary care models for deliveringpopulation based health outcomes. Discussion paper for theNational Health Committee, August 1999 (on-line).http://www.nhc.govt.nz/pub/phc/phcmodels.html

28 Stewart JF, Carter KD, Brennan DS. Adult Access to DentalCare: Rural and Remote Dwellers. Canberra: AustralianInstitute of Health and Welfare, catalogue no. DEN 41,1998.

29 Brennan D, Spencer AJ, Szuster F. Rates of dental serviceprovision between capital city and non-capital locations in