Stress, allostatic load and mental health in Indigenous Australians^*

* This article is dedicated to the memory of Professor Bruce S. McEwen.

Abstract

The aim of this narrative review was to demonstrate how the notion of allostatic load (AL) relates directly to the mental health disparities observed between Indigenous and non-Indigenous Australians. We also endeavored to synthesize the results of the limited number of studies examining stress and AL in Indigenous Australians in order to explore the potential public health benefits of the AL concept. A range of literature examining health inequalities, psychosocial determinants of mental illness and AL was explored to demonstrate the applicability of stress biology to the significant mental health burden faced by Indigenous Australians. Furthermore, all original studies indexed in MEDLINE that provided quantitative data on primary stress biomarkers in Indigenous Australians were selected for review. Evidence of hypothalamic-pituitary-adrenal axis dysregulation and increased AL is apparent even in the handful of studies examining stress biomarkers in Indigenous Australians. Urinary, salivary, hair and fingernail cortisol, hair cortisone, urinary epinephrine, heart rate variability and the cortisol awakening response are all AL parameters which have been shown to be dysregulated in Indigenous Australian cohorts. Furthermore, associations between some of these biomarkers, self-perceived discrimination, exposure to stressful life events and symptoms of psychiatric disorders in Indigenous Australians have also been demonstrated. The continued assessment of AL biomarkers and their relationship with past traumas, lifetime stressors and socio-economic factors amongst Indigenous Australians is important to addressing the mental health this population. Measurement of AL biomarkers in a culturally appropriate manner may lead to more targeted preventative measures, interventions and policies, which mitigate the effects of stress at both the individual and societal level.

Keywords:

• Stress
• allostatic load
• Indigenous Australians
• aboriginal
• torres strait Islander

Introduction

The considerable health disparities between Aboriginal and Torres Strait Islander Australians (hereafter respectfully referred to collectively as Indigenous Australians) and non-Indigenous Australians are widespread and well-documented (AIHW, 2015; AIHW, 2016; Hunter et al., 2012). The life expectancy of Indigenous Australians is between 9.5–10.5 years lower than that of non-Indigenous people (AIHW, 2015). The mental health burden of Indigenous Australians contributes significantly to this disparity and may account for 19% of the overall health gap between Indigenous and non-Indigenous Australians (AIHW, 2015).

Attempts to quantify the burden of mental illness in Indigenous Australians have shown hospitalization rates for psychiatric disorders in this demographic group are significantly higher than in the non-Indigenous population (AIHW, 2015). It is also sobering to note that the age-adjusted Indigenous suicide rate in certain regions of Australia, such as the Kimberly region, may be seven times that of the age-adjusted suicide rate in the general Australian population (Black, Ranmuthugala, & Kondalsamy-Chennakesavan, 2015; Campbell, Balaratnasingam, McHugh, Janca, & Chapman, 2016). Despite these harrowing statistics, according to the Burden of Disease Study for Aboriginal and Torres Strait Islander Australians, 37% of the overall excess disease burden seen in Indigenous Australians is avoidable (AIHW, 2016). The implications of this finding is that gaps in health status, in terms of physical but also mental health, can be targeted through effective public health measures (AIHW, 2016).

There are many psychosocial determinants which underpin mental health disparities between Indigenous and non-Indigenous Australians. A number of the psychosocial factors affecting this subset of Australians, described later in this review, are common but still uniquely distinct to that of other First Nations populations and ethnic minorities across the globe (Boksa, Joober, & Kirmayer, 2015; Kozlov, Vershubsky, & Kozlova, 2003; Riva et al., 2014). A significant biological mediator between such psychological and social factors and the manifestation of chronic disease is the stress response (Berger & Sarnyai, 2015; Marmot, 2011). Considerable evidence has been accumulated to link the chronic activation of the biological stress response to not only metabolic syndrome and cardiovascular diseases but also substance abuse and mental illness, which are all conditions associated with higher disease burden and excess mortality among Indigenous Australian people (Bjornthörp, 1991; Brunner, 2007; McEwen & Seeman, 1999; Sarnyai, Shaham, & Heinrichs, 2001). Chronic stress and the notion of allostatic load (AL) describes the cumulative deleterious effects that prolonged stress exerts on the body (McEwen & Stellar, 1993). Sarnyai and Berger have proposed that AL is a potential predictor of mental health outcomes in Indigenous Australians (Berger, Juster, & Sarnyai, 2015; Sarnyai, Berger, & Jawan, 2016). Such a predictor, if properly quantified, may potentially serve as an effective public health tool to assist in identifying “at-risk” individuals and closing the gap in the burden of psychiatric disorders between Indigenous and non-Indigenous Australians.

Aims and review methodology

In this narrative review, we outline mechanisms of chronic stress and AL and discuss how these likely contribute to the mental health disparities observed between Indigenous and non-Indigenous Australians. We also synthesize the results from the limited amount of studies that have directly investigated stress biology in Indigenous Australians. Furthermore, we explore the public health benefits of measuring stress biomarkers in this heterogeneous demographic group. Peer-reviewed studies which provide original quantitative data on biomarkers and stress in the Australian Indigenous population were identified by entering the terms Indigenous Australian, Aboriginal, Torres Strait, Stress, HPA axis and Allostatic load into MEDLINE. Articles were then selected for review based on relevance to the aforementioned aims.

Social determinants of mental health in Indigenous Australians

Broadly speaking, health inequalities are highly determined by social inequalities rather than just differences in lifestyle and access to healthcare (Marmot, 2004). This all begins even before conception, continues during gestation into infancy, childhood and adolescence in a unidirectional lifestream in which true “reversal” is impossible but rather redirection through resilience can improve quality of life and “healthspan” (Halfon, Larson, Lu, Tullis, & Russ, 2014; McEwen & McEwen, 2017). Universal factors that are thought to contribute to social disadvantage include childhood development, education, employment, income, lack of community connectedness and the absence of targeted approaches to prevent inequality (Krieg, 2009; Marmot, 2011). Marmot has proposed that all of these factors are applicable in the context of Indigenous Australians (Marmot, 2011).

Moreover, an additional unique factor which underpins social, and therefore mental health, inequalities is the way in which Indigenous people are marginalized and disempowered within mainstream Australian society (Marmot, 2011). Such marginalization is grounded in the imposition of a modern, Western lifestyle upon that of the vastly different pre-colonial Indigenous hunter-gatherer way of life, over a relatively short period of time (Hunter, 2010). Division and inequality between Indigenous and non-Indigenous Australians has in turn been perpetuated through transgenerational trauma, discrimination, neglect and racism (Berger et al., 2015; Larson, Gillies, Howard, & Coffin, 2007; Herring, Spangaro, Lauw, & McNamara, 2013). It is also pertinent to note that the mental health of the large proportion of Indigenous Australians dwelling in rural and remote regions is impacted significantly by resource maldistribution and poor access to services which would otherwise be available in metropolitan areas (Hunter, 2007).

All of the aforementioned factors which underpin social inequalities have negative psychological effects, which may drastically influence health outcomes (Brunner, 2007). It has been demonstrated that psychological distress in the Indigenous Australian population is associated with low annual household income, unemployment, poor educational attainment, lack of home ownership, financial stress and food insecurity (Markwick, Ansari, & Sullivan, 2015). An important biological conduit between these psychosocial factors and the development of pathological mental states is the chronic activation of the stress response.

Biological stress response and disease

The acute stress response–the biological process which underpins acute psychological distress–is initiated by perception of a stressful stimulus which triggers a cascade of biochemical events. Stress hormones are released via activation of the hypothalamic-pituitary-adrenal (HPA) and sympathetic-adrenal-medullary (SAM) axes. These hormones, together with pro- and anti-inflammatory cytokines, exert a number of effects on cellular activities, which seek to maintain homeostasis (McEwen, 2017). Such effects, caused by chemical mediators such as epinephrine, norepinephrine, cortisol, interleukin-6 (IL6) and tumor necrosis factor alpha (TNF-alpha) are elicited at the level of enzymes, receptors, ion channels and genes (Sarnyai et al., 2016).

Over time, biological systems recalibrate their own activities to compensate for the over and/or under production of primary mediators (McCaffery, Marsland, Strohacker, Muldoon, & Manuck, 2012; Sarnyai et al., 2016). This may eventually lead to metabolic, cardiovascular, and immune dysregulation, which may manifest in alterations in insulin and glucose levels, changes in levels of triglycerides, cholesterol and lipoproteins and increased levels of acute phase proteins respectively. Such dysregulation may lead to the development of physical and mental disorders (Mauss, Li, Schmidt, Angerer, & Jarczok, 2015). As a result of sustained chronic stress and biological wear and tear, bodily systems become overloaded and clinical end points and pathological states such as increased mortality, metabolic and cardiovascular disease, and psychiatric disorders emerge (Berger et al., 2015; Mauss et al., 2015).

Stress and the concept of allostatic load

The term allostasis refers to the normal, adaptive physiological responses organisms activate when homeostasis is disrupted (McEwen & Stellar, 1993). When chronically activated, allostatic mechanisms become physiologically taxing—referred to as AL—that consequently increase one’s susceptibility to disease (McEwen, 2017). AL therefore represents the physiological strain organisms experience when allostasis is repeatedly activated. Multiple allostatic mediators function as part of a non-linear network that contributes to the development of allostatic load. During mammalian evolution, this biological response has developed to deal with acute emergencies, such as threat from predators or the mobilization of energy resources during a hunt, with a proper shutting-off of the biological processes which would be taxing to bodily systems in the long term (Korte, Koolhaas, & Wingfield, 2005). Modern psychosocial stressors faced by Indigenous Australians such as poverty, trauma and discrimination are different in that they lead to frequent and protracted activation of the HPA axis (Berger et al., 2017; Berger et al., 2019; Sarnyai et al., 2016).

Frequent activation of the HPA axis leads to glucocorticoid secretion which is not effectively terminated by usual feedback mechanisms or, alternatively, is not effectively stimulated in times of acute stress (McEwen, 1998). This results in allostatic overload with downstream effects resulting in adverse secondary and tertiary outcomes and poor physical health (McEwen & Wingfield, 2003). Brain networks involving connections between the prefrontal cortex (PFC), anterior cingulate cortex (ACC), the amgygdala, insula and thalamus are also modulated by cortisol release and the release of the neurotransmitters serotonin, norepinephrine and dopamine (Sarnyai et al., 2016). Disruption of these regions through chronic stress and increased HPA activity results in sustained hypervigilance, and loss of specificity in the perception of stressors, which are hallmark features of a number of neuropsychiatric disorders (Berger et al., 2015). In the context of Indigenous Australians–as detailed later in this review–it has been shown by both Berger and Davison that repeated HPA axis activation leads to its exhaustion and hypofunction. This finding is consistent with studies in populations also subjected to ongoing chronic stressors (Berger et al., 2017; Davison, Singh, & McFarlane, 2019a; Davison, Singh, & Oguoma, 2019b). Suppression of the HPA axis, as a result of chronic stress and increased AL, is associated with the emergence of depression and psychotic disorders (Berger, Kraeuter, & Romanik, 2016; Dedovic & Ngiam, 2015).

Given that such disorders manifest after chronic dysregulation of the HPA axis, it is pertinent to consider the impacts of stress on neurodevelopment and the development of chronic disease throughout the lifespan of Indigenous Australians. During gestation, maternal stress can lead to fetal programing of chronic disease; a concept that was developed and described by Barker and colleagues (Barker, 1995). There is evidence to suggest that maternal stressors during pregnancy lead to an increased lifetime risk of cardiovascular and metabolic disease in the fetus, as well an increased risk of developing psychiatric disorders such as schizophrenia later in life (Babenko, Kovalchuk, & Metz, 2015). Epigenetic mechanisms are postulated to underpin the link between maternal stress and the deleterious biological consequences in offspring, which may be perpetuated across generations (Babenko et al., 2015). This is a factor that is amplified in Indigenous Australians due to transgenerational trauma as a result of colonization, and, in particular, the removal of offspring from parents, which has resulted in a large, historically significant subset of forcibly removed Indigenous Australians otherwise known as the Stolen Generations (Atkinson, Nelson, & Brooks, 2014).

It is also important to consider the health behaviors that can result from chronic stress, including the stressors often experienced by Indigenous Australians at high rates across the lifespan. These stressors include adversity in early life, such as exposure to poverty and maltreatment, which has been correlated with an increased likelihood of smoking, alcohol consumption and increased body mass index (BMI), which can all subsequently lead to poorer mental and physical health (Goldstein, Flett, & Wekerle, 2010; Midei & Matthews, 2011; Najman, Toloo, & Siskind, 2006). Alcohol and substance abuse, in particular, has been deemed by Hunter to play a significant role in the onset of psychotic disorders in Indigenous Australians (Hunter et al., 2011; Hunter et al., 2012). Another pertinent health behavior related to chronic stress, leading to poorer long-term mental health outcomes, is sleep deprivation and circadian rhythm disturbance which results in increased inflammatory tone, dysregulated metabolism, cognitive rigidity and changes in neural connectivity in the PFC and hippocampus (Bowles, McEwen, & Boutin-Foster, 2017; Cho, 2001; Karatsoreos, Bhagat, Bloss, Morrison, & McEwen, 2011; Karatsoreos & McEwen, 2014). These health behaviors, as well as other health risk behaviors such as overeating and obesity, have recently been shown to contribute significantly to elevated AL (Suvarna, Suvarna, & Phillips, 2020). The proposed interplay of factors leading to increased stress and AL in Indigenous Australians and their downstream effects is summarized in Figure 1.

Figure 1. Stress and allostatic load in Indigenous health and disease (Image and text reproduced with permission from Sarnyai et al., 2016). Transgenerational trauma (e.g. “Stolen Generation”) influences the genome through epigenetic changes, which result in increased susceptibility to future social and environmental stressors and may predispose the individual to diseases in later life. Prenatal and early postnatal stressors, such as poor nutrition, exposure to toxic substances and drugs of abuse, stress associated with low socio- economic status, childhood trauma (neglect and abuse), along with pervasive discrimination create a lifetime exposure to stress and trauma. This will result in a chronically increased activation of the biological stress response and immune system, characterized by elevated cortisol and chronic low-grade inflammation. Ultimately, this leads to increased allostatic load and long-term exposure to these bioactive substances impacts on brain structure (synaptic remodeling, neurogenesis) and function, leading to altered stress perception, emotion regulation, information processing and cognitive function. In a susceptible individual, such changes may contribute to the development of mental health problems, low educational attainment and poor lifestyle choices driven by impaired decision-making processes, which, in turn, lead to alcohol and drug abuse, unhealthy diet and resultant obesity, metabolic syndrome and cardiovascular disorders. Metabolic disturbances in turn contribute to neuroendocrine dysregulations and chronic low-grade inflammation, thus sustaining allostatic load. These, together with self-harm and suicide emerging from the above mental health problems, contribute to early death and the widening of the Indigenous life expectancy gap. Figure 1 and its legend were originally published in: Sarnyai, Berger, & Jawan (2016) Allostatic load mediates the impact of stress and trauma on physical and mental health in Indigenous Australians. Australas Psychiatry 24: 72–75.

Quantifying allostatic load

McEwen, Seeman and colleagues initially measured AL by focusing on ten AL biomarkers in the MacArthur Study of Successful Aging (Seeman, McEwen, Rowe, & Singer, 2001). These biomarkers included: epinephrine, norepinephrine, urinary cortisol, serum DHEAS, waist-to-hip ratio, total cholesterol, serum high density lipoprotein (HDL), glycosylated hemoglobin (HbA1C), systolic and diastolic blood pressure (Table 2 from Seeman et al., 2001). By incorporating AL composite measures that count the number of dysregulated biomarkers, a growing body of literature has demonstrated improved prediction of numerous deleterious outcomes in comparison to traditional biomedical methods that focus almost exclusively on normalizing clinically significant biomarker levels. Because the AL index incorporates multiple biomarkers before they fall within clinical ranges, nuanced prediction of pathological states is possible. It is pertinent to note that medical professionals already rely upon many of these biomarkers to diagnose disease; however, at present, attention is placed on clinically significant deviations from reference ranges as opposed to sub-clinical ranges which can be utilized to identify high-risk individuals.

In recent years, a number of different approaches toward calculating a composite score reflective of AL index have been employed. Such measures have varied in terms of the number of biomarkers used, the systems from which the biomarkers are derived (e.g. HPA axis biomarkers versus metabolic parameters), use of continuous or categorical measurements, definition of categorical thresholds, adjustment for medication use and weighted methods of combining biomarker data. Studies which seek to calculate AL index often categorically define biomarker levels as either “high” or” low” depending on the distribution of data within the quartiles of the study sample. This of course means that indices vary across studies, and are dependent on the biomarker distributions within populations (Gallo, Fortmann, & Mattei, 2014). It has been recently suggested that the number of biomarkers used in calculating the AL index may not lead to a difference in the overall calculated measure of AL (Berger et al., 2018).

Studying stress biology in Indigenous Australians

There have only been a handful of studies thus far which have examined the link between psychosocial stressors and individual or composite AL biomarkers in the Australian Indigenous population. The pertinent findings of these studies are summarized in Table 1. Twenty-five years ago Schmidt et al demonstrated that Australian Aboriginals in the Kimberly region had higher urinary cortisol and epinephrine excretion rates than that of an Oxford comparison population which had higher rates than world standards (Schmitt, Harrison, Spargo, Pollard, & Ungpakorn, 1995). In a follow-up study they demonstrated cyclical variation in urinary cortisol and adrenaline rates in Kimberly Aboriginal Australians, with peaks in excretion being correlated with stressful gambling activities which occurred on a weekly basis (Schmitt, Harrison, & Spargo, 1998). Both of these earlier studies were the first to point toward increased activation of the HPA-axis in a subset of Indigenous Australians, which were postulated to be a result of high levels of stress in the Aboriginal people of the Kimberly (Schmitt et al., 1995; Schmitt et al., 1998).

Table 1. Studies measuring HPA-axis activity in Indigenous Australians.

Study details	Study group(s) and sample sizes	Allostatic load biomarkers	Pertinent Findings
Schmitt et al., 1995	Aboriginals from the Kimberley region (3x communities–Derby, Kalumburu, Kupungarri): n = 145  Compared with data from UK (Oxford) residents–Pollard’s 1993 study of South Oxford residents.	Urine cortisol  Urine adrenaline	All Aboriginal communities demonstrated a very high cortisol excretion rate compared to a sample of UK (Oxford) residents   (p < .01 Derby community,   p < .01 Kalumburu community,   p < .01 Kupungarri community).   Aboriginal communities excrete adrenaline at a slightly higher rate than that found in Oxford, however only the Derby males had statistically significant results (p < .05).
Schmitt et al., 1998	Total: n = 62 Aboriginal Australians from the Kimberley region.    Monday-Tuesday group: n = 34.  Thursday-Friday group: n = 28.	Urine epinephrine  Urine cortisol  Stature  Weight  SBP DBP	Higher mean urine epinephrine in the Thursday-Friday group (p = .004).  Higher mean urine cortisol in the Thursday-Friday group (p = .027).  Diastolic blood pressure 6 mmHg higher in the Thursday-Friday group.
Berger et al., 2017	Total: n = 22  Indigenous: n = 11  Non-indigenous: n = 11	Endogenous cortisol levels (saliva)  HRV	Indigenous participants: blunted cortisol awakening response (p = .002).   Association between self-reported discrimination and HRV (p = .040).
Davison et al., 2019a	Total: n = 344  Indigenous: n = 268  Non-Indigenous: n = 76	Hair cortisol  Hair cortisone  Total hair glucocorticoid	Total: no difference in hair cortisol, cortisol or total glucocorticoid between Indigenous and non-Indigenous participants (p = .87).   Women: higher levels of hair cortisone in Indigenous women compared to non-Indigenous (p = .003).  Men: lower levels of cortisone in Indigenous men compared to non-Indigenous men (p = .037).   No association between hair cortisone/cortisol and perceived stress.
Berger et al., 2019	Total: n = 329 (all Indigenous)  Torres Strait Cohort: n = 207  North Queensland Cohort: n = 122	Hair cortisol  HR SBP DBP  IL-6  TNF-α CRP Triglycerides  HDL LDL  Total cholesterol  Glucose  Glycosylated hemoglobin  BMI	Hair cortisol correlated with age (increasing age increasing hair cortisol, p < .001), but not different between male and females in either screening programs (p = .67, p = .75).  Overall depression scores (aPHQ-9) were unrelated to hair cortisol (p = .25 and p = .94) and AL (p = .30 and p = .88) Anhedonia (p < .007) and insomnia (p < .006) sub-scores were each significantly associated with AL in one study site
Davison et al., 2019b	Total: n = 245 Indigenous: n = 179 Non-Indigenous: n = 66	Fingernail cortisol	Similar median fingernail cortisol levels between Indigenous (4.36pg/mg, interquartile range 2.2 − 10.0pg/mg) and non-Indigenous (3.87 pg/mg, interquartile range 2.0–9.7pg/mg) participants (p = .68) Indigenous participants had higher fingernail cortisol levels when adjustments were made for levels of emotional distress and exposure to stressful events (Geometric mean: 1.82, 95%CI: 1.07–3.09, p = .027) A negative association was found between fingernail cortisol in the Indigenous group and increasing number of stressful events (Geometric mean: 0.89, 95%CI: 0.84–0.95, p = .001)

SBP: systolic blood pressure; DBP: diastolic blood pressure; HRV: heart rate variability; HR: heart rate; SBP: systolic blood pressure; DBP: diastolic blood pressure; IL-6: interleukin 6; TNF-α: tumor necrosis factor alpha; CRP: c-reactive protein; HDL: high density lipoprotein; LDL: low density lipoprotein.

Table 2. The 10-parameter allostatic load index (Seeman et al., 2001).

System	Allostatic load marker
Sympathetic nervous system	12-h urinary epinephrine (μg/g creatine) 12-h urinary norepinephrine (μg/g creatine)
HPA axis	12-h urinary cortisol (μg/g) Serum DHEAS (μg/dL)
Metabolism	Waist-to-hip ratio Total cholesterol Serum HDL cholesterol (mg/dL) Glycosylated hemoglobin (%) (HbA1c)
Cardiovascular	Diastolic blood pressure (mm Hg) Systolic blood pressure (mm Hg)

DHEA: dihydroepiandrosterone sulfate; HDL: high-density lipoprotein.

Two decades later, Berger et al demonstrated significant differences in the circadian cortisol excretion profiles between Indigenous Australian and non-Indigenous university students in North Queensland (Berger et al., 2017). Indigenous participants had significantly lower levels of salivary cortisol sustained throughout the day, reflective of a lower total cortisol output (Berger et al., 2017). Moreover, a significant flattening of the cortisol awakening response (CAR) was observed in Indigenous Australians, which was associated with the experience of chronic stress and specific stressors such as childhood adversity (Berger et al., 2017). A suppressed CAR is also associated with chronic diseases and risk for mental disorders (Boehringer et al., 2015; Joseph & Golden, 2017; Nederhof et al., 2015). Therefore, the observed flattening of CAR in Indigenous participants may indicate increased risk of developing chronic non-communicable disorders, including neuropsychiatric disorders (Berger et al., 2017). They also found that heart rate variability (HRV), a marker of autonomic nervous system instability (ANS), was associated with increased levels of self-reported discrimination during a laboratory-based acute psychosocial stress (Berger et al., 2017). This raises the importance of measuring biomarkers other than those directly related to the HPA axis to predict disease risk in Indigenous Australians.

Davison et al published a study in 2019 exploring the relationship between cortisol, cortisone and total glucocorticoids amongst Indigenous Australians and non-Indigenous Australians (Davison et al., 2019a). Cortisol and cortisone levels were measured from hair samples. Along with their hair samples, participants were also asked to quantify their emotional well-being and emotional stress, using quantified scales, as a means of assessing a possible link between perceived stress and glucocorticoid levels. No significant relationship was found between these two variables. However, more detailed analysis revealed that Indigenous women had higher levels of hair cortisone compared to non-Indigenous women, whilst Indigenous men had lower hair cortisone levels compared to non-Indigenous men (Davison et al., 2019a). Indigenous women–as opposed to their non-Indigenous counterparts–are subjected to a range of unique psychosocial stressors including higher rates of teenage pregnancy and domestic violence, which may account for the elevated levels of cortisone (Davison et al., 2019a). Moreover, the decreased level of cortisone seen Indigenous males may be due to suppression of HPA axis activity in response to chronic stress, as diurnal glucocorticoid profiles are known to be flattened in this instance (Faresjo, Theodorsson, & Chatziarzenis, 2013; Miller, Chen, & Zhou, 2007).

Recently, Berger et al made direct comparisons between hair cortisol levels, AL (calculated by amalgamating the values for various endocrine, immune, metabolic and cardiovascular biomarkers listed in Table 1) and depressive symptoms in two separate, anthropologically distinct Indigenous Australian cohorts within the context of community health checks (Berger et al., 2019). The first cohort consisted of individuals aged 15 years and older was situated across various sites in the Torres Strait. The second cohort was comprised in individuals aged 15–24 years of age, and was situated in Yarrabah; a rural North Queensland Aboriginal Community (Berger et al., 2019). Both hair cortisol levels and AL were found to be significantly higher in the Torres Strait population than the North Queensland population. Furthermore, within the North Queensland population, insomnia and anhedonia–two hallmark symptoms of depression–were positively correlated with increased AL (Berger et al., 2019). These results implicating a positive relationship between insomnia and anhedonia and AL suggest that multisystem dysregulation may influence the presence of these individual hallmark symptoms of depression (Berger et al., 2019). More work is needed in the area to determine the relationship between composite measures of AL and the manifestation of other mental and physical disorders.

In a very recent study, Davison et al. demonstrated the utility, practicality and cultural appropriateness of collecting fingernail cortisol samples in Indigenous cohorts. Since fingernail cortisol can be obtained and measured non-invasively, and has been previously shown to be reflective of cortisol levels seen in saliva and hair, it was deemed to be a culturally appropriate and robust stress biomarker (Davison et al., 2019b). This study compared demographic data, substance misuse, emotional wellbeing, perceived stress, exposure to stressful events and fingernail cortisol levels between two separate but complimentary cohorts from Australia’s Northern Territory. The first cohort consisted of 179 Indigenous Australians and the second was comprised 66 non-Indigenous Australians, with participants in both cohorts between aged 21–28 years (Davison et al., 2019b). While there was no significant difference in fingernail cortisol levels between Indigenous and non-Indigenous Australians, Indigenous participants showed higher cortisol levels when adjusted for levels of emotional distress and exposure to stressful life events. Indigenous participants were also shown to have experienced a significantly higher rate of stressful life events than their non-Indigenous counterparts (Davison et al., 2019b). Moreover, a significant negative association was found in the Indigenous cohort between fingernail cortisol and increasing number of stressful life events experienced, which may represent hypofunction of the HPA axis as a result of repeated stressors (Davison et al., 2019b).

This apparent underactivity of the HPA axis with increasing exposure to stressful events is consistent with the finding of decreased levels of hair cortisone observed in the Indigenous male cohort of previous study by Davison et al, and is also in line with the finding of decreased CAR in Indigenous individuals subjected to chronic stressors by Berger et al. (2017), (Davison et al., 2019a). Importantly, all three studies, which are the only investigations carried out over the last decade that directly compare differences in stress biomarkers between Indigenous and non-Indigenous cohorts, point toward suppression of the HPA axis in response to chronic stress. Therefore–despite the relative paucity of literature concerning AL and stress biomarkers in Indigenous Australians–the impact of chronic, population-specific stressors on increasing AL and suppressing the HPA axis in Indigenous Australians is beginning to be characterized.

More broadly, there have also been very few studies to investigate the links between stress, trauma and AL in Indigenous populations across the globe. Among Canadian Inuit of Nunavik it has been demonstrated that household crowding is a source of chronic stress and that higher household crowding was significantly associated with elevated AL levels (Riva et al., 2014). Furthermore, there were greater odds of being at risk on at least seven physiological indicators included in the AL index, especially among women and independently of individuals’ characteristics (Riva et al., 2014). More recently, Thayer et al studied associations of early life trauma, post-traumatic stress disorder (PTSD) and AL in an American Indian population. Native Americans, like the Indigenous Australian population and other First Nations people, are affected by colonization, breakdown of community and cultural ties, higher incidence of chronic diseases and shorter life expectancy (Paradies, 2016). They found that early life trauma was related to PTSD, which in turn was related to elevated allostatic load in adulthood (Thayer, Barbosa-Leiker, & McDonell, 2017). Moreover, quite recently, it was demonstrated in a western Canadian cohort of Indigenous adults that AL index–taken as the composite of seven biomarkers–was seen to increase with increased levels of racial discrimination (Currie, Copeland, Metz, Chief Moon-Riley, & Davies, 2020). Another significant finding of this study was that high levels of cultural continuity–or increased engagement in practices and values of one’s Indigenous heritage–act to decrease the impact of racial discrimination on elevating AL in Indigenous Canadian adults (Currie et al., 2020). Overall, the results of such studies in distant geographical areas suggest that the AL concept might be useful in assessing the impact of stress and trauma related to the experience of Indigenous individuals in many countries, not just Australia, when it comes to mental health.

Using the allostatic load concept to address Indigenous Australian health disparities

Given the unique set of chronic stressors affecting the Indigenous Australian population, the AL model might serve as an important framework when addressing the significant mental health disparities that plagues this population. Increased AL has long been known to increase the mortality attributable to a range of disease states, including cardio-metabolic, respiratory, infectious diseases and a host of other communicable and non-communicable diseases (Robertson, Beveridge, & Bromley, 2017). Moreover, in light of the findings of the limited number of studies discussed earlier–which demonstrate differences in HPA axis activity in Indigenous Australians and suggest that AL may be implicated in the symptomatology of depression–there is emerging evidence to suggest that measurement of AL may be of benefit to preventative healthcare in Indigenous Health.

One particular aspect of the AL model which lends itself well to the Indigenous Australian population is that AL biomarkers give objective measurements of cumulative disease risk, prior to the manifestation of pathological states (Gallo et al., 2014). In the context of Indigenous health, stressors are prevalent throughout pre-natal stages of life, childhood, adolescence and adulthood. By quantifying the effects of epigenetic mechanisms and persisting transgenerational trauma, these measurements may be useful in a number of areas of preventative health across the lifespan of Indigenous Australians (Barker, 1995; Babenko et al., 2015). In light of recent results that demonstrated increased levels of glucocorticoids in young Indigenous women, one such area of preventative healthcare could be to fortify efforts to prevent domestic violence and teenage pregnancy and their health effects, which are two stressors which are disproportionately high in this particular population (Davison et al., 2019a). Moreover, following on from the study which demonstrates increased prevalence of depressive symptoms in Indigenous Australians, preventative mental health interventions could be directed toward Indigenous individuals with a high AL index prior to the manifestation of psychiatric disorders (Berger et al., 2015).

In keeping with the suggestion that prominent social determinants driving health inequalities in the Australian Indigenous population are marginalization and disempowerment, routine measurement of AL biomarkers could direct social and culture-specific services to focus on at-risk individuals (Marmot, 2011). Such services, which already exist in the form of Aboriginal Community-Controlled Health Services (ACCHS), could then further direct efforts toward these high-risk individuals in order to increase their sense of belonging and connectedness to land, which is an important cultural component of wellbeing among Indigenous Australians (Freeman et al., 2014; Burgess, Johnston, Bowman, & Whitehead, 2005). The crucial protective factor of being connected to one’s culture, which has recently been demonstrated to attenuate the impact of racial discrimination on AL in Indigenous Canadians (Currie et al., 2020), is highly applicable in the Indigenous Australian context. As such, finding new ways of increasing cultural continuity may serve as a focus of public health efforts to lessen the burden of mental illness in Indigenous Australians

Overall, the consideration of AL biomarkers may be conducive to addressing the array of systemic sociocultural and political issues, which have underpinned Aboriginal and Torres Strait Islander people for decades. However, more empiric research in the area of AL and Indigenous mental health is required to better understand the relationships between social factors, chronic stress and the manifestation of disease. That said, recent advances in this area have provided a robust foundation for further culturally appropriate investigation into AL and Indigenous mental health disparities (Berger et al., 2018; Berger et al., 2019; Davison et al., 2019a; Davison et al., 2019b). The way forward in approaching these inequalities in psychiatry will involve utilizing and expanding on this foundation with the view to reorient policy and health services and further close the gap in mental health outcomes between Indigenous and non-Indigenous Australians.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

S. Ketheesan

Dr S. Ketheesan graduated with an MBBS from James Cook University, Townsville, Australia and is currently Year 1 Psychiatry Registrar at the Royal Brisbane and Women's Hospital, Brisbane, Australia. He has been interested in the role of stress in Indigenous mental health and has been active in research since his first year at university.

M. Rinaudo

Dr M. Rinaudo graduated with an MBBS from James Cook University, Townsville, Australia and is currently a Resident Medical Officer who worked at the Royal Brisbane and Women's Hospital, Brisbane, Australia over the last year. She has a keen interest in childhood development.

M. Berger

Dr M. Berger is a medical graduate of University of Vienna, Austria and obtained his PhD in Dr Sarnyai's lab in 2018. He is currently a Research Fellow at Orygen, the National Centre of Excellence in Youth Mental Health and University of Melbourne, Australia. He is the recipient of a fellowship grant from the Melbourne Academic Centre for Health.

M. Wenitong

Dr M. Wenitong is from Kabi Kabi tribal group of South Queensland. He is the Senior Medical Officer at Apunipima Cape York Health Council, where he is working on health reform across the Cape York Aboriginal communities. Dr Wenitong is a past president and founder of the Australian Indigenous Doctors Association. He is a council member of the Australian Institute of Aboriginal and Torres Strait Islander Studies and a member of the Queensland Aboriginal and Torres Strait Islander Advisory Council.

R. P. Juster

Dr R. P. Juster's research interests include the study of allostatic load, a measure of the long-term consequences of the effects of chronic stress in people. In his studies, he takes into account variables related to gender and sex to identify possible differences and explanations. He is interested in both the biological and social determinants of chronic stress. He obtained his PhD at the University of Montreal, Canada where he is now an Assistant Professor, following postdoctoral research at Columbia University, New York, USA.

B. S. McEwen

Prof B. S. McEwen was one of the leading stress researchers of the last 50 years, who discovered that hormone receptors are localized in the brain and developed the concept of allostatic load in medicine. He received his PhD from Rockefeller University where he spent his whole career. He has been recognized with many prestigious awards and prizes, including the Pasarow Neuropsychiatry Award, the Ipsen Fondation Neuroplasticity and Endocrine Regulation Prizes, the Scolnick Prize in Neuroscience from the McGovern Institute of MIT, and the William James Lifetime Achievement Award for Basic Research presented in honor of the 25th Anniversary of American Psychological Society. He was also an elected member of the National Academy of Sciences, the National Academy of Medicine, and the American Society of Arts and Sciences. He passed away while this manuscript was in the final stage of the review process. This paper is dedicated to his memory.

Z. Sarnyai

Prof Z. Sarnyai received his MD and PhD from the University of Szeged in Hungary and postgraduate training at the Department of Psychiatry at Harvard Medical School and at The Rockefeller University. He was previously University Lecturer in the Department of Pharmacology, University of Cambridge and a Fellow of Pembroke College, Cambridge. He was appointed as Lady Davis Visiting Professor at Technion-Israel Institute of Technology for 6 months in 2019. He was awarded the Curt Richter Prize by the International Society of Psychoneuroendocrinology; the DuPont-Warren Award by the Department of Psychiatry, Harvard Medical School; and the Brain Research Foundation (formerly NARSAD) Young Investigator Award.

Notes

* This article is dedicated to the memory of Professor Bruce S. McEwen.