It has been demonstrated that socioeconomic factors affect health through material and psychosocial pathways [1]. Materially, people with higher income may have more or better material goods and health services and be much healthier as a result [1]. However, average income levels (e.g., gross national income per capita) are not always associated with health in rich countries, yet there are strong associations between income inequality and health, which suggests that psychosocial mechanisms are relevant [2–4]. For instance, in a society with greater income inequality, people in lower social positions are more likely to be exposed to behavioral risks resulting from psychosocial stress, including stress-related smoking, drinking, and eating “for comfort.” Previous ecological studies across and within countries have indicated that health indicators at the population level, such as mortality and life expectancy, are associated with income inequality [5–8], though there is still debate about the degree of this association [9, 10]. With the establishment of diverse health survey systems and the increasing availability of individual health data, more in-depth studies at the level of the individual have been conducted in Europe [11–13], the US [14–18], Japan [19, 20], as well as in the mainland China [21] and Hong Kong [22].

Chronic non-communicable diseases have become one of the most important disease burden worldwide [23, 24]. The measurements of the outcomes of health or disease are no longer limited to objective indicators (e.g, mortality, morbidity, life expectancy, etc.) but are expanding to health-related quality of life (HRQL), which reflects the subjective feeling of health [25, 26]. Many researchers have employed the self-rating of health in their individual-level studies [11–13, 17, 20, 27], which reflects personal and subjective judgments of health and conforms to the biopsychosocial medical model. EQ-5D is a useful tool for measuring HRQL in terms of disease, disability, and psychology and is widely used in many countries for its simplicity of administration [28–33]. Since 2008, EQ-5D has been included in the Chinese National Health Service Survey (NHSS), which is the largest national representative health survey.

Chiang et al. examined the changing relationship between income inequality and mortality across different stages of economic development in Taiwan [10]. He found that in the transitional process of moving from an agricultural to an industrial economy, the association between relative income and health was increasing, while the connection between absolute income and health was decreasing. However, most published studies were from societies in advanced economies and few were from developing economies due to limited access and the poor quality of individual health data. Shaanxi is an interior province located in northwestern China, and its economy is rapidly developing. With the strategies of reform and opening up and Development of the Western Regions effectively implemented in China, Shaanxi’s economy has experienced rapid development over the last 30 years [34, 35]. The gross domestic product per capita in Shaanxi reached 291 yuan, 4968 yuan, and 14,607 yuan in 1978, 2000, and 2007, respectively. However, income disparity has rapidly broadened during this period. For example, Yuchun Zhu et al. reported that the Gini coefficient in this region was 0.06 in 1978, peaked at 0.37 in 2000, and decreased a little to 0.28 in 2004 [35].

Using the data from the NHSS (2008) and the Shaanxi province census data, this study examined the status of the association between income inequality and HRQL and the differences in the association between high- and low-income groups as well as between rural and urban regions.

Income inequality has been found to be associated with many negative health outcomes, such as higher mortality, decreased life expectancy, and worse self-rated health, in advanced economies [11–18]; however, few studies have focused on developing economies. China’s economic achievement has been termed an “economic miracle.” However, the income disparities between rural and urban regions as well as between coastal and interior regions have increased dramatically over the last 30 years [39–41]. As an interior province located in the northwest of China, Shaanxi has experienced a similar social and economic development process. In this study, the associations of HRQL, measured by the EQ-5D, with household income and income inequality were examined on the basis of a provincial representative sample in Shannxi, a rapidly developing economy in China. Our findings indicated that HRQL worsened as income decreased, and the negative effect of income inequality on HRQL was greater for people with low income than those with high income and also greater for people living in rural regions than those living in urban regions.

The effect of low income on HRQL was negative in all the models, which was consistent with previous studies [14, 16] and strongly supported the hypothesis that HRQL worsens as income decreases. In terms of income inequality, our results different somewhat between low- and high-income groups. High income inequality was a significant risk predictor of HRQL in the low-income group, while the effect was not significant in the high-income group. This finding was not fully consistent with studies in advanced economies, in which income inequality was predictive of worse health for people with both low and high income [14, 16, 19]. However, our finding was insufficient to deny or even challenge the two pathways through which income inequality affects health. On the contrary, we still believe that the two pathways [2] also function as important mechanisms in developing economies such as Shaanxi. There may be three factors contributing to the association of HRQL with income inequality in Shaanxi. First, Shaanxi’s income level was still relatively low compared to other areas of China, as shown in Table 1—it had not reached the income level of a developed economy. This may have affected the results for the high-income group. Second, there might be a time lag between the prevalence of income inequality and its effects on HRQL. As Yang et al. reported in his research, in the last two decades, the incomes and the income gap in the north of Shaanxi increased dramatically while environmental pollution also increased due to the exhaustive exploitation of energy resources. The negative effects of environmental pollution on health take time to be observed [42]. Third, the population used to calculate income inequality might not be large enough and may be relatively homogeneous. Therefore, further research on the association of provincial-level income inequality with health in China should be conducted with larger samples.

As seen in Table 3, the coefficients of the models in rural regions were larger than those of urban regions, especially in the low-income group. This implied that the health disparity of people living in rural regions, which had greater income inequality, was much larger than that in urban regions. People in rural regions were much more vulnerable to economic factors, including both absolute and relative income. This suggests that it may be helpful to create different economic and health policies for rural and urban regions, and people in rural regions would achieve more health benefits if policy support could be provided to increase household income as well as reduce personal health expenditure.

This study provides a new perspective on the association between HRQL and income inequality. However, it has a few limitations. One limitation is that the household income was self-reported. Some respondents may have intentionally would underreported their income, which could lead to bias in income and the county or district Gini coefficient. However, the bias would be in the same direction for all counties and districts because most of the bias would result from underreporting, which would not change the relative position of each county and district in the Gini classification. Therefore, our conclusion would still be applicable. In addition, a cross-sectional design limited any inference regarding causation. Therefore, further longitudinal studies examining trends in income inequality, individual income, and health are needed.

Income inequality has a damaging effect on HRQL in Shaanxi, China, especially for people with low income. In addition, people living in rural regions are more vulnerable to economic factors. Therefore, Shaanxi should work to increase income levels, especially among rural residents, and make efforts to reduce income inequality.