Socioeconomic status (SES) is a profound and consistent predictor of health status at both the population and individual level 1, between and within countries across the globe 2. Despite regularity in patterns of health inequality across SES strata, the means by which disadvantage incurs risk of illness remain inadequately understood. This is not to suggest that little can be done to alleviate health inequalities without a thorough understanding of the biopsychosocial pathways linking social adversity to health disadvantage. Evidence of widening socioeconomic inequalities 3 mandates consideration of collective failures to apply what is already known, or deficits in awareness of more fundamental drivers of disparity and an ability to influence these. This seems particularly poignant when considering the health disadvantage of Indigenous and other marginalised populations, where ill health can be seen as a function of larger forces that impinge on specific sub-populations, rather than as a problem of these sub-populations.

Indigenous peoples in Australia are profoundly disadvantaged on virtually all markers of health and socioeconomic status across the life course 45. Much of the epidemiological and policy landscape of Australia has supported the "social determinants" perspective of health as a framework through which to understand such disparity. The cultural and gender-specific determinants of behaviour, treatment, health awareness, prevention and causation have yet to be determined, however. Prevention efforts from the individual level to the societal level have not thus far incorporated into practice anything beyond a general understanding of the basis of Indigenous health disadvantage, or attempted to make explicit the filters of such effects.

Psychosocial factors are relevant to understanding how risk conditions affect cognition and thus health through behavioural and biological responses 6. Such measures are receiving growing acceptance as contributors to disease aetiology 7, and as modifiers and mediators of social health disparity 8. Support for a psychosocial explanation for health inequalities is most often seen for chronic illness and outcomes including cardiovascular reactivity 9 and mortality 10, survival after myocardial infarction 11, perceived health 12, all-cause mortality 13 and various psychopathological and emotional states 14. Control, operationalised as mastery 15, self-efficacy 16, or locus of control 17, has emerged as a central integrating concept in health research, and as a psychosocial resource in the aetiology of health and disease 18. A growing literature attests to a connection between social disadvantage, nominal control, risk behaviour and morbidity 1920. Chronic low hope for influencing conditions of living is a strong predictor of morbidity and mortality 21, whereas chronic instability and disorder predict psychological distress and poor health behaviour 2223. Used as an explanation for how ethnicity, race, or SES relate to health through the quality of experience in one's environ, psychosocial factors have the potential to capture feelings of oppression and alienation relevant to marginalised peoples.

Minimal empirical research has been published on psychosocial factors in Indigenous Australians, and quantitative evidence of a link between psychosocial factors and health has not been reported. Studies amongst Canadian First Nations have documented associations between mastery, health behaviour, and cardiovascular risk 2425, but such groups have had a longer period of westernisation than Indigenous peoples in Australia. It is important to know if relationships observed for disadvantaged, and primarily non-Indigenous populations elsewhere, apply also to Indigenous Australians who have undergone a recent, rapid transition to westernised living. This study sought to undertake culturally-sensitive measurement of psychosocial factors and to test relationships between mastery, perceived stress, and health behaviours including recreational physical activity and vegetable and fruit consumption, in a remote Indigenous settlement in Australia.

Persons surveyed were volunteers for a community-based chronic disease risk factor screening initiative among Yolngu people residing in and around Galiwin'ku settlement, Elcho Island, of northeast Arnhem Land in the Northern Territory of Australia. This population is isolated and, as recently as 1985, people in this region lived a traditionally-oriented lifestyle with no risk factors for diabetes or cardiovascular disease 26.

Persons 15 years of age and older were targeted for screening. Participants gave their written, informed consent. The protocol was reviewed and clearance provided by the Top End Human Research Ethics Committee with input from an Indigenous Sub-Committee. Menzies School of Health Research, in Darwin, Northern Territory, was responsible for the study.

The data reported here indicate that for Yolngu people in north-eastern Arnhem Land, mastery (i) has age-specific relationships with health behaviour and (ii) is negatively related to perceived recent stress, chronic stress and youth stress. A causal relationship would support action on factors underlying perceived stress as potentially modifiable determinants of mastery. Some such exposures are most likely heterogeneous at the level of the individual ("random stressors") with still others homogenous within (but not necessarily across) disadvantaging environments ("systemic stressors") 29. The living conditions that define the population surveyed – limited access to transportation, communications and means of food storage, and a low prevalence of even basic education – indicate unequivocally the extent of environmental disadvantage that applies in this setting. While some persons because of their unique experiences may perceive greater stress or lower mastery, there is a need to view the distribution of stressors and stress responses as a function of forces acting on the population as a whole 30.

Additional factors besides perceived stress no doubt influence mastery, and need to be elucidated. Nonetheless, mastery would seem to have cross-cultural utility in linking untoward perceptions that may reflect the disadvantaging experiences of Indigenous people, to actual health behaviour. Little empirical research has targeted mastery or "control" in seeking to understand Indigenous health in a context of westernisation and disadvantage. Our findings that mastery was positively related to health behaviour for individuals ≥25 years of age, while being negatively related to health behaviour for those <25 years of age, indicate that the health correlates of mastery are not invariably positive. Such results suggest that, depending on age or maturity, health salience, or perceived health, mastery may be expressed in different ways. A broader, more positive influence of mastery consistent with findings amongst individuals ≥25 years of age is indirectly supported by a study showing improved health status for Indigenous Australians who relocated to traditional lands relative to counterparts in regional centres 31. Collectively, homelands dwellers have autonomy, an elevated capacity for self-determination, and less environmental stress. Therefore, at the population level, mastery, conceptually linked to Indigenous people's connection to traditional lands, offers explanatory as well as preventive salience within the context of this analysis.

Used to explain how health is affected by the quality of experience in one's environment, mastery has the potential to contextualise the legacy of colonisation experienced by Indigenous communities and to capture feelings of alienation relevant to marginalised peoples. Alienation in this sense can be understood as a dissociation of people from meaningful work, their social collectivities, or their own identities, or of being distanced from power and resources that may enable self-determination in political, economic, and social settings 32. Such cognitions may be more common or salient among individuals aged 25 years or more. This broader application of control to health and health-related behaviour reaches beyond past research on job demands and job decision latitude in work environs 3334. It is consistent, however, with the growing body of empirical research attesting to important connections between social disadvantage, risk behaviour and morbidity 35363738.

Evidence is accumulating that life experiences and social arrangements, one's place in the social context, or perceived stresses of living, embed themselves in cognitions reflected by psychosocial characteristics and biology over the life cycle 3940. This study extends such knowledge to Indigenous Australians who have undergone a recent shift to westernised living in a geographically isolated area. The finding among persons ≥25 years of age that mastery is associated with physical activity and vegetable (but not fruit) consumption has not previously been reported for an Indigenous population in Australia. A lack of association with fruit may reflect its greater popularity and availability in the community. Such relations are consistent with research indicating that mastery is related to behavioural risk factors for First Nations in Canada 2425. As depression and hopelessness predict for non-Indigenous populations the development of atherosclerosis 7, type 2 diabetes 41 and coronary heart disease 42, mastery may well be related through behaviour to the development of disease in Indigenous peoples. Few longitudinal studies have been mounted, none featuring psychosocial measures.

That mastery was inversely associated with age for women of all age groups and for men 25 years and older (Figure 1), implies age-dependant decreases in mastery, a phenomenon that for which the literature offers little insight. Most studies on mastery have simply controlled for age. For Indigenous Australians, mastery may reflect different facets of psychological wellbeing at different stages of life, and the apparent decrease may not be pathogenic. Other explanations beyond age include period or cohort effects, but our cross-sectional design precludes assessing the presence of these. Certainly, the finding that higher mastery at younger ages coincides with negative, rather than positive, health behaviour supports longitudinal research on psychosocial status, its evolution and correlates. Given the social and cultural construction of identity, role and place across the life-course, the process of aging may be an important contributor to and manifestation of psychological wellbeing, or of one's sense that they are in control of the factors affecting their lives, opportunity, health, and community wellbeing.

A noteworthy finding of this study was low levels of mastery among men between 15–24 years. Young Indigenous males are amongst the most vulnerable in Australian society 434445. Indigenous males have long recognised the significance of the loss of their authority and self-esteem through alienation, loss of culture, and recurrent challenges to individual and collective spiritual well being and a sense of control over ones life 46. These have been considered as important antecedents to violence, self harm, drug dependence, depression, and anxiety, with suicide among young Indigenous males being an important indicator of the social, emotional and psychological well-being of Indigenous communities 45. This particular vulnerability of Indigenous male youth resonates with the pattern of mastery exposed in this analysis.

Despite the conceptual congruence of mastery within this population, limitations are worth noting. This report is based on small numbers, but primary objectives were to explore within-population relations between psychosocial and behavioural characteristics, rather than assessment of the prevalence of such factors. Despite a potential for information bias through adapted instruments, we contend this analysis provides a reasonable assessment of the cross-cultural utility of mastery within an Indigenous settlement. The cross-sectional nature of the study precludes determination of causality and the temporal direction of the relations assessed. Reliance on volunteers could affect results through selection bias. Reporting bias is possible for self-reported data, although the extent of heterogeneity observed for all measures, and the lack of differences between the broader sample and persons completing psychosocial surveys, suggests no underlying systematic bias, and adequate representativeness. The acceptability of questionnaire items might bear improvement, however, given that 16% of participants did not fully complete the survey. While the overall sample was representative of the target population, the generalisability of the results may be limited given that probability-based sampling was not viable for the target population and hence not utilised. Few data have been reported, however, on psychosocial status in Indigenous populations, and this report provides a basis for initiating more comprehensive longitudinal surveys with representative samples of Indigenous peoples to assess relationships between perceived control, mental health and health-related behaviour across the life course.

Understanding the psychosocial correlates of health and behaviour in Indigenous peoples involves significant challenges, but offers tremendous opportunity. Assessment of mastery was deemed relevant and important by the community under study, demonstrated adequate semantic and construct equivalence and, on the whole, correlated with measures of behaviour and acute and chronic stressors in the anticipated direction. Mastery is a positive construct, which in this Indigenous Australian population seems to have utility in linking the perception of adversity to health-related behaviour. Making explicit the relationship between disadvantage and poor health in marginalised populations is requisite to the identification of factors amenable to intervention. Assessment and cross-cultural adaptation of mastery and related constructs offer the methods and opportunity to explore deeper understandings of the nature, interactions, and consequences of an individual's historical and social context, and the relationship between disadvantage and poor health status in marginalised populations. Little research has focused on the role of such factors in attempting to explain the unacceptable level of health disparity of Indigenous peoples. Ultimately, however, this work points to the start of a process of unpacking the social determinants of Indigenous peoples' health, which embraces the symbiotic relationship between the body, mind and spirit, individual and the collective, the person and place, and the past, present and future.

The author(s) declare that they have no competing interests.

MD conceived and designed the study, undertook the analyses, interpreted the results, and drafted the manuscript. AB contributed to analysis and interpretation of the data and drafting of the manuscript with attention to Indigenous cultural and scientific perspectives. GD made substantial intellectual contributions through cultural adaptation of psychosocial measures, design and implementation of culturally appropriate data collection protocols, acquisition of data, and critical revision of the manuscript for important intellectual content. MDC provided substantial contributions to analysis and interpretation of the data, and critical revision of the manuscript for important intellectual content. KOD made substantial contributions through conception and design of the study, and critical revision of the manuscript for key intellectual content. All authors read and approved the final manuscript.