As the population of many countries age, an increasing proportion of the global population will be in their more advanced years [34]. As such, increasing attention is being paid to the wellbeing and living standards of older people [19, 22, 35, 53, 54]. Australia is no exception, with 25% of the population expected to be aged over 65 by 2044–45 [40].

Considerable attention has been paid to the macroeconomic impacts of an older population within Australia – increased health care costs, and welfare payments, and reduced productivity from a lower aged-dependency ratio (the proportion of working aged people to the proportion of people of retirement age) [50]. At the individual level, multiple studies have investigated the living standards of older people, which have included the assessment of health, income and other aspects of living standards [4, 5, 38]. One study by McRae et al., also explored the impact of healthcare costs on the living standards of older people within Australia, concluding that 12% faced catastrophic healthcare costs [33]. Older people with depression also faced the third highest annual costs for health care. This highlights growing interest in the health - living standards nexus, including mental health issues, amongst older Australians, and how this affects the well-being of a growing proportion of the population.

There is a vast body of literature that demonstrates the association between mental illness and lower socio-economic status in people of all ages [30, 31, 39, 41]. This literature, however, assesses the impact of income poverty on later mental health status, rather than the inverse relationship - the impact of mental illness on poverty. The direction of the impact is important for assessing points for potential intervention to ensure that older people are not beset by both poor mental health and poverty.

The limited literature assessing this inverse relationship, found that 57% of people aged 45 to 64 with depression were not in the labour force, and that these people had a weekly income 73% lower than people who were employed [42–45]. Butterworth et al. looked specifically at older workers and found that those who had retired early were more likely to have mental disorders than those in the labour force, particularly amongst men [7] and similar results were also reported by Gill et al. [24]. In a more recent study Butterworth et al. used longitudinal data of all ages to show that mental health status predicted future unemployment [8] and Kiely and Butterworth [29] used longitudinal data to show that mental health predicted receiving welfare payments [29]. However, these studies focus on the impact that mental ill health has upon unemployment or income and given that older people beyond the age of 65 have low rates of labour force participation these studies of lost earnings or the impacts on employment are likely to be insensitive to the full impact of mental illness on the living standards of people beyond the traditional retirement age. This paper aims to determine whether having a high level of psychological distress increases an older individual’s risk of falling into income poverty or multidimensional poverty and how many additional people fall into poverty as a result of high psychological distress,

There were 69 records of individuals aged 65 and over in 2009 on the HILDA dataset who had high psychological distress in 2009 (102,400 people in the Australian population), and 801 records of individuals who had low psychological distress (representing 1,111,500 people in the population).

When disaggregated by sex, 51% of males with high psychological distress and 28% of males with low psychological distress fell into income poverty, and 48% of females with high psychological distress and 33% of females with low psychological distress fell into income poverty. 51% of males with high psychological distress and 14% of males with low psychological distress fell into multidimensional poverty, and 46% of females with high psychological distress and 21% of females with low psychological distress fell into multidimensional poverty.

After adjusting for age, employment in 2009, remoteness of residence in 2009, marital status in 2009 and housing tenure in 2009, males with high psychological distress had 1.68 times the risk of falling into income poverty between 2010 and 2012 (95% CI: 1.02–2.75) compared to males with low psychological distress (Table 3). There was no significant difference in the risk of falling into income poverty between females with high and low psychological distress (p = 0.1008) (Table 3).

If all cases of high psychological distress in people aged 65 years and over in 2009 had been prevented than an estimated 8.0% of income poverty cases would have been avoided between 2010 and 2012 (95% CI: 7.8% to 8.4%), and an estimated 19.5% of multidimensional poverty cases would have been avoided between 2010 and 2012 (95% CI: 19.2% to 19.9%).

The results of this paper have shown that having high psychological distress increases the risk of older males falling into income poverty compared to those with only low psychological distress; however, there was no significant difference in the risk for older females between those with high and low levels of psychological distress. This is in line with the results of previous studies, which have found that while older males are more likely to retire early after developing mental disorders, the relationship was less pronounced for females [7, 24]. To date, no studies have specifically sought to use longitudinal data to document whether mental illness is a risk factor for income poverty.

In addition to showing the higher risk of income poverty, this study goes further by using a multidimensional measure of poverty, which captures a broader spectrum of factors that influence living standards. The results have shown that there is a significantly higher risk of both older males and females falling into multidimensional poverty amongst those with high psychological distress. The living standards of those in multidimensional poverty are seen to be poorer than those who are in income poverty but have no further forms of disadvantage, as those in multidimensional poverty not only have the burden of low income, but also have poor health or a relatively poor level of education attainment acting as barriers to improving their income, or indeed acting as a drain on their income (in the case of poor health [32]). As such, these findings identify older adults with high psychological distress as being a key target population for policies to improve living standards of vulnerable populations. Similarly, interventions to prevent high levels of psychological distress developing in older adults may be seen to have the additional indirect benefits of preventing cases of poverty. This study has indicated a need for policy to consider the multi-faceted needs of older people within the population. Health care and income support are generally delivered in silos, with little recognition of how health influences economic status and how economic status influences health. A more holistic approach to people’s wellbeing may be required, with better communication between sectors.

Studies that have looked at the costs of mental illness in terms of lost income have generally focused on lower labour force participation rates as a driver of low income, both within Australian and internationally ([6, 7, 24, 25, 36]; D Schofield et al., 2011; [46, 51]). However, given the older age group lower labour force participation is likely to only be part of the reason for lower income reported in this study, due to the majority of older people with and without high psychological distress being out of the labour force. None-the-less those with low psychological distress did have a higher proportion of people in employment and so the analysis adjusted for labour force status.

The high risk of falling into multidimensional poverty, even after controlling for employment status, may be explained by older people with high psychological distress and low income and poor overall health accessing their savings or accumulated wealth stocks to pay for their health condition (or conditions). The results did show that a high proportion of people with high psychological distress who were in multidimensional poverty had low income and poor health, rather than just low income and a low level of education attainment. Individuals aged 45 to 64 in Australia who had depression or another mental illness were more likely to have significantly less wealth or none at all [42, 44, 45]. A recent study of the out-of-pocket medical costs faced by older Australians listed depression as the third most costly condition, only behind heart disease and cancer [32], and another study has shown that over 40% of Australians with depression skip health care due to the cost [12]. In addition to the potential for medical costs to reduce the amount of wealth held by older people with psychological distress, it is known that those with depression and other mental health problems are less likely to hold income producing assets such as investment properties and shares [44, 45]. Thus, even if developing high psychological distress did not result in a drawdown of total wealth, it may result in a change in wealth portfolio structure to safer and less management intensive assets, which also produce lower returns, hence negatively affecting income.

Even after removing the SF-36 MCS, which measures overall mental health, from the measure of multidimensional poverty, both males and females with high psychological distress were still more likely to be multidimensionally poor, having low income plus either poor overall health status or an insufficient level of education attainment. This is likely to be explained by the findings of other studies, which have shown that amongst older adults, depression does increase the risk of a decline in physical health [23, 37].

The key limitation of this study is that it is based on self-reported data. Firstly, the study relies on responses to the Kessler-10 survey instrument and does not measure clinically diagnosed psychological distress. The measure also only asks respondents about their experiences in the previous 4 weeks. Despite this, the K10 is a validated survey instrument shown to have good validity and reliability (as noted in the methodology section). Furthermore, cognitive ability at baseline is a further potential confounder that was not available on the dataset and thus not included in the analysis. It should also been noted that although the HILDA dataset is accompanied by population weights, and weighted results are reported, the sample was truncated, with those already in income poverty excluded. This may have affected the accuracy of the population weights.

Overall the results of this study have shown that older adults with high psychological distress have a higher risk of falling into income poverty and multidimensional poverty. Nearly half of older adults with high psychological distress in 2009 would fall into income poverty and multidimensional poverty by 2012. Even after adjusting for potential confounders, older males with high psychological distress had 1.7 times the risk of falling into income poverty and 3.4 times the risk of falling into multidimensional poverty, and older females had 2.2 times the risk of falling into multidimensional poverty, than their counterparts with low physiological distress. To date, these additional costs of psychological distress have not been quantified.