Xiaoyong Hu, Conceptualization , Methodology , Writing – review & editing , 1, * Tiantian Wang, Methodology , Writing – original draft , 1 Duan Huang, Conceptualization , Resources , 2 Yanli Wang, Data curation , Methodology , 1 and Qiong Li, Methodology , Writing – review & editing 2
1 Faculty of Psychology, Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing, People’s Republic of China
Find articles by Xiaoyong Hu1 Faculty of Psychology, Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing, People’s Republic of China
Find articles by Tiantian Wang2 School of Health, Wuhan Sports University, Wuhan, People’s Republic of China
Find articles by Duan Huang1 Faculty of Psychology, Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing, People’s Republic of China
Find articles by Yanli Wang2 School of Health, Wuhan Sports University, Wuhan, People’s Republic of China
Find articles by Qiong Li Filipe Prazeres, Editor1 Faculty of Psychology, Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing, People’s Republic of China
2 School of Health, Wuhan Sports University, Wuhan, People’s Republic of China Faculty of Health Sciences - Universidade da Beira Interior, PORTUGAL Competing Interests: The authors have declared that no competing interests exist. Received 2020 Oct 25; Accepted 2021 Jul 2. Copyright © 2021 Hu et alThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
S1 Fig: Empirical mediation model of social class and physical and mental health through health self-management. Age are controlled for but are not illustrated for simplicity. **p < .01, ***p < .001.
GUID: A7D7483E-BC4A-4195-A96F-E98F0DF0D367 S1 Table: Demographic and socioeconomic data. (XLSX) GUID: 1315FA6B-BB26-42EB-A317-FE9AEB2736E3 S2 Table: The Means and Correlations among central study variables (N = 663). (XLSX) GUID: 578FE2EA-1AF0-45F1-BE12-71FD3038C624S3 Table: Test of mediation effects of health self-management on the relationship of social class to physical and mental health: Bootstrap results. (XLSX)
GUID: F3755977-84C6-450E-8EB3-AA9112F34655 S1 Data: (RAR) GUID: 190BEC3C-D5C9-462A-85CA-C7E52AFD670F Attachment: Submitted filename: Response to Reviewers Comments.docx GUID: CF7FC1E6-A5A5-47A8-8F7D-D62CB8B5E7C6All relevant data are within the manuscript and its Supporting Information files.
Studies have explored the relationship between social class and health for decades. However, the underlying mechanism between the two remains not fully understood. This study aimed to explore whether health self-management had a mediating role between social class and health under the framework of Socio-cultural Self Model.
663 adults, randomly sampled from six communities in Southwest China, completed the survey for this study. Social class was assessed using individuals’ income, education, occupation. Health self-management was assessed through evaluation of the health self-management behavior, health self-management cognition, health self-management environment. Physical health and mental health were measured by the Chinese version of Short-Form (36-item) Health Survey, which contains Physical Functioning, Role-Physical, Role-Emotional, Vitality, Mental Health, Social Function, Bodily Pain and General Health. Pearson’s correlation was used to examine the associations between major variables. Mediation analyses were performed to explore the mediating role of health self-management.
Social class positively predicted self-rated health. The lower the social class, the lower the self-reported physical and mental health. Health self-management partially mediated the relationship between social class and self-rated health. That is, the health self-management ability of the lower class, such as access to healthy and nutritious food and evaluate their own health status, is worse than that of the higher class, which leads to physical and mental health inequality between the high and the low classes.
Health self-management mediated the relationship between social class and health. Promoting health self-management abilities are conducive to improving both physical and mental health.
Resident health not only reflects the individual’s physical and psychological adaptability, but also reflects the comprehensive strength of a country or region. In recent decades, people’s health status has seen a worldwide improvement in correlation with the economic and social development of the country. However, from the perspective of different social groups, the economic development only highlights the class gap in health [1–3]. That is, economic and social development improves the overall health of the whole society, but widens the health gap between different demographics. Therefore, an increasing number of researchers realize that only when social determinants that lead to health differences are fully taken into account, can health intervention that reduce disease and save lives be effective [2, 3].
Social class refers to groups in different positions within the social hierarchy, which was formed by economic and political reasons, among others; there are objective differences in social resources (income, education, and occupation) and subjective differences in perceived social status among these groups [4]. Many studies demonstrate that social classes can positively predict individual health [5–11]. Compared with the lower class, the upper classes have longer life spans, a better health status, and less possibility of suffering from a physical disability [12, 13]. For example, data from the United Kingdom, France, Switzerland, Portugal, Italy, the United States and Australia, published in Lancet, shows that lower class is associated with shorter life expectancy [14]. Compared with wealthier people, those in the lower class are approximately 1.5 times more likely to die before the age of 85: among those with a lower socioeconomic status, 55,600 (15.2% males and 9.4% females) died before the age of 85. This study also estimates that lower socioeconomic status can shorten life expectancy by 2.1 years (41% males and 27% females), and indicates that social class, like traditional risk factors such as hypertension, obesity, high alcohol consumption, and a sedentary lifestyle, should be a key factor of health risk [14]. In addition, compared with the upper classes, the lower class experience less happiness [1, 15, 16]; more negative emotions and stress [17, 18]; and more psychological symptoms [19, 20]. For example, research found that social class is closely related to depressive symptoms [21]; that social class is an important factor affecting anxiety; and that lower class children have more anxiety and that their anxiety is often related to psychopathology [22, 23].
Why is it that the lower the social class, the worse the health status? The Socio-cultural Self Model points out that the self is the core mechanism that leads to differences in the health between higher and lower classes [24]. The Socio-cultural Self Model focuses on the role of the self, and regards the self as the product of interaction between individual characters and environmental factors within a specific social and cultural background. The model notes that individual characteristics and environmental factors indirectly influence behavior through the self, which is shaped by social, cultural and individual factors [24]. Under the theoretical framework of the Socio-cultural Self Model, researchers suggested that self-management abilities, such as health self-management, is an important factor explaining class inequalities in health [25–27]. Health self-management is the ability of an individual to take control of their own health [28]. Specifically, it is the ability of an individual to analyze and evaluate her or his own health status and influencing factors in life, and make lifestyle changes, such as be able to actively seek health consultation and guidance, and take intervention measures such as medical treatment, appropriate exercise, or reasonable nutrition for health risk factors, and maintain his or her own health [28–31].
Several studies have confirmed that there are significant differences in the health self-management capabilities among different classes [32, 33]. For example, one research team conducted multi-factor logistic regression analysis on the health self-management of 1028 participants in Hangzhou and found that social class significantly predicted health self-management. Specifically, those with a higher education background have higher health awareness and higher level of health self-management [33]. In addition, a further study shows that health self-management is positively related with individual health status [34, 35].
Actually, many researchers have explored the mediators between social class and health from different aspects, including social (the socioeconomic position of an individual in adulthood [36], adult socioeconomic status [37]), psychological (perceived discrimination [38] , social relationships [39], social support [37]), and physical (metabolic alterations [40], cortisol slope [41], subjective social status [13], negative emotions [17]) factors. However, to the best of our knowledge, no studies have yet to directly examine the mediating role of self between social class and health. According to the Socio-cultural Self Model, self is the core mechanism of social class affecting health, and social and individual factors can only work through self. Therefore, in this study, we aimed to test the mediating role of self between social class and health. Under the framework of Socio-Cultural-Self Model, we hypothesized that social class was positively associated with physical and mental health, and that this relationship was mediated by health self-management.
After obtaining the ethical approval from the Research Ethics Committee of Southwest University (IRB NO. H19070), we collected data from Chongqing, a provincial municipality directly under the central government in Southwest China. Consistent with the purpose to investigate social class and health disparities, we sought a sample that is socially and economically diverse. Sample selection and recruitment began with the type of housing (villa, apartment, low-rent housing). In Yuzhong District and Rongchang District of Chongqing City, one affluent community(villa), one middle class communities(apartment), and one lower class communities (low-rent housing) for each district were randomly selected, and then 125 subjects were randomly selected in each community for questionnaire survey. During the investigation, a total of six well-trained investigators (psychology graduate students) whom divided into three groups, with the support of the Community Neighborhood Committee, investigated six communities one by one through the household survey. Before the survey was distributed, one of the authors explained the purpose of the study to the subjects. It was guaranteed that all of the responses would be kept confidential. All participants provided written informed consent. The survey involved 98 questions and took 20–30 minutes to complete. Totally, surveys were distributed to 750 adults of whom 663 responded to the full survey, resulting in a response rate of 88.4%. The remaining 87 subjects either refused to accept the survey, or more than 50% of the survey questions were not answered. Of these subjects, 41 live in low-rent houses, 28 live in buildings, and 17 live in villas. Among the participants, 55.2% of them were female, 44.8% were male. The average age of the sample was 33.73 years (SD = 12.61), and ranged from 18 to 65 years old.
Social class was measured by objective socio-economic status (or objective SES) indicators, which involves an intersection of different factors, including income, education, and occupation [42, 43]. To accurately model the interaction of these factors, psychologists have developed many methods to estimate social class, such as factor analysis method, regression equation method, and weighted mean method. Among them, the factor analysis method is widely used [44]. Therefore, we performed factor analysis via structural equation modeling, with participants’ reported income, education, and occupation. Specifically, the factor analysis method includes the following steps:
Firstly, participations rate their education, occupation, and per capita annual household income. According to the standards used by OCED-2012, participations rated their education on the following scale:1 = Never been to school, 2 = elementary school, 3 = junior high school, 4 = secondary technical school, 5 = general high school, 6 = vocational training after high school, 7 = college, 8 = undergraduate, 9 = postgraduate. According to the Chinese Professional Reputation Index [45], participants rated their vocation on the 24 different occupations scale. For example, 1 = senior leading cadres of the government, 21 = ordinary farmers and fishermen, 24 = engaged in such jobs as nanny, hourly worker, manual tricycle driver, etc. Participants rated their per capita annual household income from the following scale: 1 = Less than 3000 RMB, 2 = RMB 3000–5000, 3 = RMB 5000–12000, 4 = RMB 12000–20000, 5 = RMB 20000–30000, 6 = RMB 30000–55000, 7 = Over 55000 RMB.
The second step is to deal with the missing values in each variable by replacing with sequence mean. The third step is to transform the three variables of education, occupation, and per capita annual household income into standard scores. Then, principal component analysis is carried out and the social class variables are calculated according to the following formula: social × Z education + β2 × Z occupation + β3 × Z income) / εƒ, in which β1, β2 and β3 are factor loads and εƒ is the characteristic root of the first factor.
The health self-management ability scale was developed by Zhao and Huang [46], and includes health self-management behavior (14 items), health self-management cognition (14 items), health self-management environment (10 items) scale, for a total of 38 items. Using the Likert-5 scoring method, each item is scored from 1 to 5. The sum of the three dimensions is the total score of health self-management ability. The higher the total score, the better the health self-management ability. In this study, the internal consistency coefficient (Cronbach’s alpha) is 0.94.
Physical and mental health status was measured by the Short-Form (36-item) Health Survey (SF-36). SF-36 has excellent psychological measurement characteristics, and is widely used in the measurement of the general population health status [47]. The Chinese version of SF-36 has sufficient reliability and validity, and its psychological characteristics are similar to the test results of American population samples [48].
The SF-36 consists of 36 items and 8 sub-dimensions, and the applicable population is adults over 14 years old. Excluding one item in the questionnaire that measures health transition, the remaining 35 items belong to eight dimensions related to health: Physical Functioning (PF, 10 items), Role-Physical (RP, 4 items), Role-Emotional (RE, 3 items), Vitality (VT, 4 items), Mental Health (MH, 5 items), Social Function (SF, 2 items), Bodily Pain (BP, 2 items), General Health (GH, 5 items). Each dimension score is the weighted sums of the questions in the section and is directly transformed into a 0–100 scale on the basic assumption that each question carries equal weight. These eight dimensions can be aggregated into two independent comprehensive dimensions: Physical Component Summary (PCS, α = 0.87) and Mental Component Summary (MCS, α = 0.69). Physical Component Summary consists of PF, RP, BP, and GH and Mental Component Summary consists of RE, VT, MH, SF. The higher the score, the higher the degree of people’s mental and physical health status.
Several sociodemographic factors were taken into account, including gender (0 = male, 1 = female) and age.
SPSS 20.0 was used for data analyses. Firstly, we investigated the general tendency among social class, health self-management, and mental and physical health. We calculated the mean and standard deviation for each variable and the Pearson correlation coefficients between the variables.
Next, we tested the hypothesized mediational model with social class as a predictor, health self-management as a mediator, and general health as an outcome variable, using the AMOS 23.0 statistical package. Concretely, we examined the mediation mechanism of social class on mental and physical health. For the path coefficients a maximum likelihood estimation method was used, and 95% bias-corrected confidence intervals were calculated for all effects using 1,000 bootstrap samples.
The sample characteristics are shown in Table 1 . Junior high school accounts for the largest proportion of subjects’ education level (29.9%). As for occupation, a large proportion of subjects were ordinary farmers and fishermen (33.5%). In terms of income, 52.5% of the subjects had annual household income per capita less than 5000 RMB.
| Characteristics | n(%) |
|---|---|
| Female | 366(55.2) |
| Age(years) | |
| 18–25 | 214(32.3) |
| 26–35 | 186(28) |
| 36–45 | 138(20.8) |
| 46–55 | 88(13.3) |
| 56–65 | 37(5.6) |
| Education | |
| 1 = Never been to school | 53(8) |
| 2 = Elementary school | 113(17) |
| 3 = Junior high school | 198(29.9) |
| 4 = Secondary technical school | 31(4.7) |
| 5 = General high school | 128(19.3) |
| 6 = Vocational training after high school | 43(6.4) |
| 7 = College | 67(10.1) |
| 8 = Undergraduate | 27(4.1) |
| 9 = Postgraduate | 3(0.5) |
| Occupational status | |
| 1 = Senior leading cadres of party and government (cadres at ministerial level or above) | 0(0) |
| 2 = Senior professional and technical personnel, such as university professors, well-known scientists and so on | 1(0.2) |
| 3 = Party and government middle-level leading cadres | 5(0.8) |
| 4 = Leaders of government-affiliated institutions | 10(1.5) |
| 5 = Ordinary cadres of party and government organs and institutions | 12(1.8) |
| 6 = Professional and technical personnel in the fields of media, justice and education | 9(1.4) |
| 7 = Director, manager, and middle management of an enterprise | 7(1.1) |
| 8 = Law enforcement officers from taxation and other departments of the Public Security Bureau | 11(1.7) |
| 9 = Ordinary civil servants in party and government agencies and public institutions | 5(0.8) |
| 10 = Medical, engineering, economic and senior professional and technical personnel | 18(2. 7) |
| 11 = Private entrepreneur | 23(3.5) |
| 12 = Factory directors, managers of collective enterprises and middle-level managers of secondary industry enterprises | 17(2.6) |
| 13 = Middle and low-level professional technical personnel | 35(5.3) |
| 14 = Party and government organs and institutions logistics, political work, secretary, financial personnel, etc | 19(2.9) |
| 15 = All kinds of enterprise logistics, political, administrative personnel, salesman, distribution personnel, etc | 38(5.7) |
| 16 = Rural professionals, such as veterinarians, village doctors, etc | 47(7.1) |
| 17 = Small shopkeepers, owners of small workshops and other self-employed persons | 42(6.3) |
| 18 = General staff in business services | 11(1.7) |
| 19 = Industrial workers, production workers in manufacturing, including skilled workers and unskilled workers, etc. | 58(8.7) |
| 20 = professional farmer | 14(2.1) |
| 21 = Ordinary farmer and fishermen | 222(33.5) |
| 22 = Individual laborer | 27(4.1) |
| 23 = Heavy manual workers, such as porters, stevedores, miners, builders, etc | 24(3.6) |
| 24 = Engaged in nanny and part time worker, such as tricycle driver | 8(1.2) |
| Annual household income per capita | |
| 1 = Less than 3000 RMB, | 135(20.4) |
| 2 = RMB 3000–5000, | 213(32.1) |
| 3 = RMB 5000–12000, | 100(15.1) |
| 4 = RMB 12000–20000, | 116(17.5) |
| 5 = RMB 20000–30000, | 73(11) |
| 6 = RMB 30000–55000, | 26(3.9) |
| 7 = Over 55000 RMB. | 0(0) |
Due to the use of self-reported data, Harman’s single-factor test was used to rate common method bias. Exploratory factor analysis revealed that the first factor accounted for 17.77% of the total variance and did not explain most of the variance (<40%). Results showed that there was no common method bias in this study.