The association between material living standard and psychological distress: Results from a New Zealand population survey

Abstract

Background:

People with a low material living standard experience more psychological distress than those with a high living standard, but previous studies suggest the size of this difference is modest.

Aim:

To measure the association between living standard and psychological distress using a multidimensional measure of living standard, the Economic Living Standard Index (ELSI).

Methods:

Adults aged 25–64 years (n = 8,465) were selected from a New Zealand community survey. Logistic regression models were used to compare household income and ELSI scores as risk factors for high psychological distress, defined as a K10 score of 12 or over.

Results:

In the population, the prevalence of high psychological distress was 5.8%. The prevalence of high distress increased steeply with decreasing living standard. In the most deprived decile according to ELSI score, 24.3% had high distress, compared to 0.8% in the least deprived decile. For household income, high distress was present in 15.9% of people in the lowest decile and 2.2% of the highest decile. In fully adjusted models, ELSI score remained significantly associated with high distress but household income was not.

Conclusion:

The mental health disparity between those at opposite ends of the social spectrum is very large. Comprehensive measures such as the ELSI give a more accurate estimate of this disparity than household income.

Keywords

Psychological stress socioeconomic factors mental disorders income K10 Economic Living Standard Index New Zealand

Introduction

More than 30 years ago, Kessler remarked that the excess of psychological distress in lower social strata was one of the most consistent findings in psychiatric epidemiology (Kessler & Cleary, 1980). Psychiatric epidemiologists’ interest in the area waned during the 1980s (Dohrenwend, 1990), perhaps assuming the matter to be settled. However, it has recently been argued that the evidence on the relationship between socioeconomic status and common mental disorders remains inconsistent and beset by methodological weaknesses (Fryers, Melzer, & Jenkins, 2003; Lahelma, Laaksonen, Martikainen, Rahkonen, & Sarlio-Lähteenkorv, 2006).

One problem with the literature on socioeconomic status and mental health is a lack of consensus about which socioeconomic status measures to use. Education, employment, social class, income and wealth have all been examined, and the observed association with mental health depends on the specific measure employed (Lahelma et al., 2006).

Occupational social class has a long history of use in the United Kingdom in particular, but it is a relatively weak predictor of mental health (Fryers et al., 2003). Employment and education also have limitations, as they may be influenced by sex, age and cohort effects (Carter, Blakely, Collings, Gunasekara, & Richardson, 2009). Material living standard measures appear more promising, but a recent systematic review reported only a modest association with mental disorder, with the odds of disorder lying in the range 1.1–2.5 for the most disadvantaged group compared to the least (Fryers et al., 2003).

Income is the most commonly used measure of material living standard, but survey participants are often reluctant to disclose their income (Turrell, 2000), and there are difficulties taking into account self-employment and informal economic activities (O’Donnell, Van Doorslaer, Wagstaff, & Lindelow, 2008). Furthermore, low income has been shown to be a poor indicator of current deprivation (Perry, 2002). Longitudinal evidence suggests that income exerts only a weak effect on overall health (Imlach Gunasekara, Carter, Liu, Richardson, & Blakely, 2012), and individual deprivation is a stronger predictor of self-reported change in health status (Imlach Gunasekara, Carter, Crampton, & Blakely, 2013). Alternative measures including net wealth (Carter et al., 2009), asset ownership (Lahelma et al., 2006; Lewis et al., 1998; Weich & Lewis, 1998), housing quality (Weich & Lewis, 1998) or general living standard (Imlach Gunasekara, Carter, Crampton, & Blakely, 2013; Pfoertner, Andress, & Janssen, 2011) may be better predictors of health outcomes. The multidimensional nature of poverty was highlighted in the United Kingdom Poverty and Social Exclusion Study, with low income, deprivation, lack of necessities and social exclusion all shown to be associated with poorer mental health (Payne, 2006).

The New Zealand Ministry of Social Development developed the Economic Living Standard Index (ELSI) as a comprehensive measure of living standard (Jensen, Spittal, & Krishnan, 2005). It comprises 25 questions including items on asset ownership, economising, self-perception of economic status and social exclusion. Social exclusion in particular is a correlate of poverty that is not fully captured by traditional socioeconomic status measures (Burchardt, Le Grand, & Piachaud, 1999; Gordon et al., 2000), and it may be particularly relevant for people experiencing mental illness (Boardman, 2011). A summary of all items in the ELSI is provided in Figure 1.

Figure 1.

Items included in the Economic Living Standards Index (ELSI).

Psychological distress is a non-specific construct that has proven useful for measuring current mental health status in epidemiological surveys. The K10 scale measures psychological distress in the past 4 weeks, or the worst month in the last 12 months (Kessler et al., 2003). High psychological distress (K10 ≥ 12) reliably predicts the presence of serious mental illness in population samples (Andrews & Slade, 2001; Oakley, Wells, Scott, & Mcgee, 2010).

The primary aim of the present study was to compare the ELSI to other measures of socioeconomic status, in particular household income, as predictors of current psychological distress. We hypothesised that the ELSI would be a superior measure of socioeconomic status than income, employment or education and therefore would produce more a more reliable estimate of the association between socioeconomic status and mental illness in the population.

Methods

The 2006/2007 New Zealand Health Survey involved face-to-face interviews with 12,488 adults aged 15 years and over living in permanent private dwellings. The sample was obtained using clustered sampling methods. Disproportionate sampling was used to recruit sufficient Maori (n = 3,160), Pacific (n = 1,033) and Asian people (n = 1,513) to minimise the variance of parameter estimates in these subpopulations. Calibrated weighting was applied with reference to external population benchmarks. The sampling procedures have previously been described in detail (Ministry of Health, 2008). In this study, adults aged 25–64 years (n = 8,465) were selected in order to reduce the potential influence of including younger and older people less likely to be in the paid workforce.

Measures

The K10 is a 10-item scale measuring psychological distress in the past 4 weeks (Andrews & Slade, 2001; Kessler et al., 2002). Individual items were scored 0–4 in this survey; therefore, the range of total scores was 0–40 and scores of 12+ were classed as high (Andrews & Slade, 2001; Oakley et al., 2010).

Material living standard was measured using The ELSI–short form and household income. Total scores on the ELSI can range from 0 to 31 but have also been categorised. Scores of 16 or less indicate the presence of hardship, 17–24 a comfortable living standard, 25–28 a good living standard and 29–31 a very good living standard (Jensen et al., 2005). Scores on the ELSI were grouped into approximate deciles, using the following cut-points: 16.5, 20.5, 22.5, 24.5, 25.5, 26.5, 27.5, 28.5 and 29.5 corresponding to these percentiles: 10, 20, 28, 40, 49, 59, 71, 81 and 90.

Household income was obtained by self-report, and data were missing for 10% of people aged 25–64 years. The New Zealand Ministry of Health imputed values for missing data using the SAS procedure PROC IML, based on age, whether receiving a welfare benefit, household type, health insurance or individual deprivation. Household income was equivalised to take into account household composition, using a formula that assigns a different weighting to adults and children (Jensen, 1988). To allow direct comparison between each ELSI and household income decile, the same percentile groupings were used for household income, producing cut-points at the following 2006 New Zealand dollar values: 17,972; 27,460; 34,998; 44,128; 53,882; 64,489; 74,810; 89,688; 113,633.

Covariates with a known or suspected association with socioeconomic status or psychological distress were included in the full logistic regression models. These were age (25–34, 35–44, 45–54 and 55–64), sex, prioritised ethnicity (Maori, Pacific, Asian, European or other), neighbourhood deprivation (NZDep index quintile) (Salmond, Crampton, & Atkinson, 2007), Alcohol Use Disorders Identification Test score (Saunders, Aasland, Babor, De La Fuente, & Grant, 1993), smoking status, physical health (36-item Short-Form Health Survey (SF-36) role physical, physical functioning and bodily pain subscales) (Scott, Tobias, Sarfati, & Haslett, 1999), employment status and cohabitation status.

Statistical analysis

Logistic regression models were used to measure the association between the material living standard measures and high psychological distress (K10 ≥ 12). Data came from a complex survey, and therefore, analysis was conducted using SUDAAN 10.0.1 (Research Triangle Institute). Calibrated replicate weights were used for variance estimation. All figures presented are weighted estimates. Results from logistic regressions are presented as predicted marginals. This approach, which is outlined in detail by Graubard and Korn (1999), provides for easier interpretation of the magnitude of group differences by presenting percentages instead of odds ratios. Each predicted marginal is an average outcome associated with one level of a risk factor while controlling for imbalances in covariates (if there are no covariates in the model, the predicted marginal is simply the observed percentage with the outcome). Estimates are presented with 95% confidence intervals. Supplementary details regarding statistical models may be obtained by contacting the authors.

Results

The full survey had a weighted response rate of 67.9%. All subsequent reported results relate to adults aged 25–64 years. Men made up 47.9% of this population. The distribution of prioritised ethnicity was 11.3% Maori, 4.9% Pacific, 8.8% Asian and 75.1% European or other ethnicity. Men were under-represented in the sample relative to women but calibration to the known sex distribution in the population was applied.

K10 scores ranged from 0 to 40. A K10 score of 0–5 was present in 80.6% (79.4, 81.7), 6–11 in 13.6% (12.8, 14.5), 12–19 in 4.2% (3.7, 4.8) and 20–40 in 1.6% (1.3, 2.0). Combining the four categories into two, high psychological distress (K10 ≥ 12) was present in 5.8% (5.2, 6.4).

Table 1 shows the prevalence of high distress according to age and gender. The prevalence of high distress tended to decline with age and was less for males than females.

Table 1.

High psychological distress (K10 ≥ 12), by gender and age, for ages 25–64 years.

Age (years)	Percentage with high psychological distress
Age (years)	Males: % (CI)	Females: % (CI)
25–34	6.2 (4.3, 8.8)	7.4 (6.0, 9.2)
35–44	5.3 (3.8, 7.3)	7.4 (6.0, 9.0)
45–54	5.2 (3.6, 7.4)	5.0 (3.7, 6.8)
55–64	4.0 (2.9, 5.5)	5.5 (4.0, 7.4)
Total	5.2 (4.4, 6.3)	6.4 (5.6, 7.2)

CI: confidence interval.

Table 2 shows the results of a series of logistic regression models examining ELSI and household income as predictors of high psychological distress (K10 ≥ 12). Unadjusted univariate models are presented for ELSI and household income, followed by results from a model (model 3) including ELSI and household income with no other covariates, then a model with ELSI, household income, age and gender (model 4) and finally a fully adjusted model (model 5) with ELSI, household income and all covariates. As shown in Table 2, there was a strong univariate association with psychological distress for both ELSI and income, but in the fully adjusted model, only ELSI remained a predictor of high distress.

Table 2.

Material living standard and high psychological distress.

Decile^a	Percentage^b with high psychological distress (K10 ≥ 12)
	Models 1 and 2: unadjusted, % (95% CI)		Model 3:^c % (95% CI)		Model 4:^d % (95% CI)		Model 5:^e % (95% CI)
	ELSI	Household income	ELSI	Household income	ELSI	Household income	ELSI	Household income
1	24.3 (21.3, 27.7)	15.9 (13.0, 19.2)	21.0 (17.6, 24.3)	8.3 (6.3, 10.2)	21.1 (17.8, 24.4)	8.3 (6.3, 10.3)	12.0 (9.8, 14.2)	6.4 (4.7, 8.1)
2	9.7 (7.6, 12.2)	11.0 (8.6, 13.9)	8.9 (6.7, 11.0)	7.0 (5.2, 8.8)	8.9 (6.7, 11.1)	7.0 (5.1, 8.8)	6.7 (5.2, 8.1)	5.9 (4.4, 7.4)
3	5.4 (3.7, 7.8)	7.2 (5.4, 9.5)	5.2 (3.3, 7.1)	5.3 (3.8, 6.8)	5.2 (3.3, 7.2)	5.3 (3.8, 6.8)	4.4 (2.8, 6.0)	4.9 (3.5, 6.3)
4	4.8 (3.3, 7.0)	5.3 (3.9, 7.2)	4.9 (3.1, 6.7)	4.6 (3.2, 6.0)	4.9 (3.1, 6.8)	4.6 (3.2, 6.1)	4.9 (3.0, 6.8)	5.4 (3.8, 7.0)
5	5.0 (3.0, 7.9)	4.5 (3.3, 6.1)	5.0 (2.7, 7.4)	4.4 (3.1, 5.7)	5.0 (2.7, 7.4)	4.4 (3.1, 5.8)	5.6 (3.1, 8.2)	5.0 (3.6, 6.5)
6	3.2 (2.1, 4.8)	4.5 (3.0, 6.7)	3.3 (2.0, 4.6)	5.6 (3.6, 7.7)	3.3 (2.0, 4.6)	5.6 (3.6, 7.6)	3.8 (2.4, 5.3)	5.8 (3.7, 7.9)
7	2.2 (1.3, 3.7)	3.7 (2.6, 5.4)	2.3 (1.1, 3.5)	5.0 (3.2, 6.9)	2.3 (1.1, 3.5)	5.0 (3.1, 6.9)	2.9 (1.3, 4.5)	6.3 (4.2, 8.5)
8	1.8 (1.0, 3.1)	2.0 (1.2, 3.4)	1.9 (0.1, 3.0)	3.4 (1.7, 5.0)	1.9 (0.8, 3.0)	3.4 (1.7, 5.0)	2.9 (1.4, 4.5)	4.0 (2.1, 6.0)
9	1.9 (1.1, 3.5)	3.5 (2.3, 5.3)	2.1 (0.1, 3.3)	6.1 (3.7, 8.5)	2.1 (0.8, 3.3)	6.1 (3.7, 8.5)	3.4 (1.5, 5.4)	7.2 (4.4, 10.0)
10	0.8 (0.3, 1.9)	2.2 (1.3, 3.6)	0.9 (0, 1.6)	5.1 (2.5, 7.6)	0.8 (0.1, 1.6)	5.1 (2.5, 7.6)	1.5 (0.2, 2.7)	5.9 (3.0, 8.8)
Wald F, p	45.2,<0.001	18.1,<0.001	28.1,<0.001	2.6,0.009	29.5,<0.001	2.5,0.01	10.7,<0.001	1.08, 0.39

ELSI: Economic Living Standard Index; CI: confidence interval.

Approximate deciles with cut-points at the following percentiles: 10, 20, 28, 40, 49, 59, 71, 81 and 90.

Predicted marginal means from logistic regression models.

Model with ELSI and household income, no other covariates.

Model with ELSI, household income, age and sex.

Fully adjusted model including ELSI, household income and all covariates.

There was only a modest bivariate correlation between household income and ELSI score (r = 0.385). As shown in Table 3, hardship is concentrated in people with household incomes below the median, but it is not confined to those in the lowest income decile.

Table 3.

Unadjusted prevalence of financial hardship by household income deciles, for ages 25–64 years.

Decile	Household income range^a	Percentage with current financial hardship^b (95% CI)
1	<17,972	34.3 (30.7, 38.1)
2	17,972–27,460	20.9 (18.2, 23.8)
3	27,460–34,998	15.0 (12.0, 18.6)
4	34,998–44,128	9.9 (8.1, 12.1)
5	44,128–53,882	9.2 (7.2, 11.8)
6	53,882–64,489	4.3 (2.9, 6.5)
7	64,489–74,810	3.8 (2.1, 6.8)
8	74,810–89,688	1.7 (1.0, 3.1)
9	89,688–113,633	1.8 (1.0, 3.2)
10	113,633+	0.5 (0.2, 1.4)
Overall		9.8 (9.0, 10.5)

CI: confidence interval.

Equivalised household income, in 2006 New Zealand dollars.

Economic Living Standard Index score of 16 or less.

Discussion

This study demonstrates the association between living standard and psychological distress in a large New Zealand population sample. In particular, we compared household income to a multidimensional measure known as the ELSI. Some of this dataset has been previously analysed in a report to the New Zealand Ministry of Health, the purpose of which was to compare the ELSI to another living standard measure (the NZDep) in relation to health more broadly (Tobias & Mason, 2010).

Of those in the most deprived decile, according to their ELSI score, 24% had high psychological distress in the past 4 weeks. This was compared to 5.8% of the population overall. The percentage of high distress declined steeply between decile 1 and decile 2 and then maintained a steady downward trend between deciles 2 and 10. The percentage of high distress was only 0.8% among the least deprived decile. The pattern of association remained after accounting for a range of covariates, but it was reduced in magnitude.

Household income has been extensively used to measure living standard in population surveys. There was a strong univariate association between household income and psychological distress, but the association did not persist after adjustment for covariates including ELSI score. Analyses using household income were complicated by 10% of data being missing, highlighting a further limitation of this measure.

Composite measures of living standard such as the ELSI have the advantage of being able to assess a range of dimensions of poverty that are not adequately captured by measuring income. These include social exclusion and lack of access to necessities, which have previously been shown to predict adverse health outcomes (Burchardt et al., 1999; Gordon et al., 2000; Payne, 2006).

The presence of high psychological distress in the past 4 weeks, defined as a K10 score of 12 or more, predicts serious mental illness (Kessler et al., 2003; Oakley et al., 2010). Our findings suggest that people with the lowest living standard are at far greater risk of having a current serious mental illness than those with a comfortable living standard or better. The strength of association between material living standard and mental illness demonstrated in this study substantially exceeds what has been described in most previous studies (Fryers et al., 2003; Laaksonen et al., 2007; Lahelma et al., 2006) and is somewhat greater than that reported in a recent study using another, more basic, combined hardship measure (Butterworth, Olesen, & Leach, 2012). This may be due to the ELSI being a more elaborate measure than those used in most previous studies. A second possible explanation for the current results is that there is a greater association between hardship and psychological distress in New Zealand in comparison to other populations studied, although this seems unlikely. A further consideration is that the psychological distress found among people living in poverty might represent a fluctuating normal response to environmental stressors, rather than a clinically significant mental disorder.

Much of the recent focus in the scientific literature has been on understanding the effect of inequality on health at a population level and attempting to address it via social policy (Wilkinson, 2009). A simpler observation is that at an individual level, social deprivation is likely to be relevant for mentally ill people presenting for treatment even if it is not causally related to the mental illness. Conversely, people presenting to social welfare services are likely to have poor mental health. The common co-occurrence of mental illness and poverty has important implications for service provision.

Concerningly, our findings show very high levels of mental health inequality in New Zealand in spite of a well-developed social welfare system, state-funded healthcare and levels of income inequality comparable to other Organisation for Economic Co-operation and Development (OECD) countries (Pickett, 2010).

The main limitation of this study is that it is based on cross-sectional data and therefore cannot test whether psychological distress causes or results from deprivation, although previous studies suggest that there is a causal relationship in both directions (Boardman, 2011; Burchardt et al., 1999). This was a household survey and therefore excludes individuals in institutions such as prisons and residential facilities, in whom both financial hardship and psychological distress are likely to be over-represented. The response rate was 67%, which is consistent with other similar studies, but it is possible that differential non-response could have biased these findings.

Conclusion

People with a low material living standard experience very high levels of psychological distress. Composite measures that capture a range of aspects of poverty appear more useful than income for determining the extent of this disparity.

Footnotes

Acknowledgements

The New Zealand Crown is the owner of the copyright of the data.

Conflict of interest

The results presented in this article are the work of the authors, and the authors take full responsibility for the research outputs. The authors declare no competing interests and received no external funding to conduct this research.

Funding

The New Zealand Health Survey 2006/2007 was funded by the New Zealand Ministry of Health.

References

Andrews

Slade

(2001). Interpreting scores on the Kessler Psychological Distress Scale (K10). Australian and New Zealand Journal of Public Health, 25, 494–497.

Boardman

(2011). Social exclusion and mental health – How people with mental health problems are disadvantaged: An overview. Mental Health and Social Inclusion, 15, 112–121.

Burchardt

Le Grand

Piachaud

(1999). Social exclusion in Britain 1991–1995. Social Policy & Administration, 33, 227–244.

Butterworth

Olesen

S. C.

Leach

(2012). The role of hardship in the association between socio-economic position and depression. Australian and New Zealand Journal of Psychiatry, 46, 364–73.

Carter

K. N.

Blakely

Collings

Gunasekara

F. I.

Richardson

(2009). What is the association between wealth and mental health? Journal of Epidemiology & Community Health, 63, 221–226.

Dohrenwend

B. P.

(1990). Socioeconomic status (SES) and psychiatric disorders. Social Psychiatry & Psychiatric Epidemiology, 25, 41–47.

Fryers

Melzer

Jenkins

(2003). Social inequalities and the common mental disorders. Social Psychiatry & Psychiatric Epidemiology, 38, 229–237.

Gordon

Adelman

Ashworth

Bradshaw

Levitas

Middleton

… Townsend

(2000). Poverty and social exclusion in Britain. York, UK: Joseph Rowntree Foundation.

Graubard

B. I.

Korn

E. L.

(1999). Predictive margins with survey data. Biometrics, 55, 652–659.

10.

Imlach Gunasekara

Carter

Crampton

Blakely

(2013). Income and individual deprivation as predictors of health over time. International Journal of Public Health, 58, 501–511.

11.

Imlach Gunasekara

Carter

K. N.

Liu

Richardson

Blakely

(2012). The relationship between income and health using longitudinal data from New Zealand. Journal of Epidemiology & Community Health, 66, e12.

12.

Jensen

(1988). Income equivalences and the estimation of family expenditures on children. Wellington, New Zealand: Department of Social Welfare.

13.

Jensen

Spittal

Krishnan

(2005). ELSI short form: User manual for a direct measure of living standards. Wellington, New Zealand: Centre for Social Research and Evaluation.

14.

Kessler

R. C.

Andrews

Colpe

L. J.

Hiripi

Mroczek

D. K.

Normand

S.-L. T.

… Zaslavsky

A. M.

(2002). Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine, 32, 959–976.

15.

Kessler

R. C.

Barker

P. R.

Colpe

L. J.

Epstein

J. F.

Gfroerer

J. C.

Hiripi

… Zaslavsky

A. M.

(2003). Screening for serious mental illness in the general population. Archives of General Psychiatry, 60, 184–189.

16.

Kessler

R. C.

Cleary

P. D.

(1980). Social class and psychological distress. American Sociological Review, 45, 463–478.

17.

Laaksonen

Martikainen

Lahelma

Lallukka

Rahkonen

Head

Marmot

(2007). Socioeconomic circumstances and common mental disorders among Finnish and British public sector employees: Evidence from the Helsinki Health Study and the Whitehall II Study. International Journal of Epidemiology, 36, 776–786.

18.

Lahelma

Laaksonen

Martikainen

Rahkonen

Sarlio-Lähteenkorv

A. S.

(2006). Multiple measures of socioeconomic circumstances and common mental disorders. Social Science & Medicine, 63, 1383–1399.

19.

Lewis

Bebbington

Brugha

Farrell

Gill

Jenkins

Meltzer

(1998). Socioeconomic status, standard of living, and neurotic disorder. The Lancet, 352, 605–609.

20.

Ministry of Health (2008). Methodology report for the 2006/07 New Zealand Health Survey. Wellington, New Zealand: Ministry of Health.

21.

Oakley

B. M. A.

Wells

J. E.

Scott

K. M.

Mcgee

M. A.

(2010). The Kessler Psychological Distress Scale in Te Rau Hinengaro: The New Zealand Mental Health Survey. Australian and New Zealand Journal of Psychiatry, 44, 314–322.

22.

O’Donnell

Van Doorslaer

Wagstaff

Lindelow

(2008). Analyzing health equity using household survey data: A guide to techniques and their implementation. Washington, DC: The World Bank.

23.

Payne

(2006). Mental health, poverty and social exclusion. In Pantazis

Gordon

Levitas

(Eds.), Poverty and social exclusion in Britain. Bristol, UK: The Policy Press, pp. 285-311.

24.

Perry

(2002). The mismatch between income measures and direct outcome measures of poverty. Social Policy Journal of New Zealand, 19, 101–127.

25.

Pfoertner

T.-K.

Andress

H.-J.

Janssen

(2011). Income or living standard and health in Germany: Different ways of measurement of relative poverty with regard to self-rated health. International Journal of Public Health, 56, 373–384.

26.

Pickett

K. E.

(2010). Inequality: An underacknowledged source of mental illness and distress. British Journal of Psychiatry, 197, 426–428.

27.

Salmond

Crampton

Atkinson

(2007). NZDep 2006 index of deprivation (Department of Public Health ed.). Wellington, NZ: University of Otago, Wellington.

28.

Saunders

J. B.

Aasland

O. F.

Babor

T. F.

De La Fuente

J. R.

Grant

(1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on early detection of persons with harmful alcohol consumption – II. Addiction, 88, 791–804.

29.

Scott

K. M.

Tobias

M. I.

Sarfati

Haslett

S. J.

(1999). SF-36 health survey reliability, validity and norms for New Zealand. Australian and New Zealand Journal of Public Health, 23, 401–406.

30.

Tobias

M. J.

Mason

(2010). Living standards and health: New Zealand 2006/07. Wellington, New Zealand: Ministry of Health.

31.

Turrell

(2000). Income non-reporting: Implications for health inequalities research. Journal of Epidemiology & Community Health, 54, 207–214.

32.

Weich

Lewis

(1998). Material standard of living, social class, and the prevalence of the common mental disorders in Great Britain. Journal of Epidemiology & Community Health, 52, 8–14.

33.

Wilkinson

R. G.

(2009). The spirit level: Why more equal societies almost always do better. London: Allen Lane.