Abstract
Background
The term “reversal hypothesis” has been attached to the observation that the socio-economic status (SES) gradient of non-communicable diseases (NCDs) tends to reverse over time as country's economic and social development progresses. We examined this hypothesis in China between 1991 and 2020, to test whether a divergence in the experiences of men and women can be identified and to explore the possible contributors to any disparity.
Methods
Using nine waves of China Health and Nutrition Survey (CHNS, 1991–2015) and six waves of the China Family Panel Studies (CFPS, 2010–2020), we investigated the SES-NCD gradient among participants aged 18 years and older. We examined the relationship between self-reported doctor diagnosed NCDs and SES proxied by formal educational years by gender using multivariable logistic regression and Blinder-Oaxaca decomposition models.
Findings
Prevalence of a list of NCDs in the overall population increased from 3.52 % (1991) to 13.72 % (2020). The association between SES and NCDs has shifted from positive to negative over time, shifting earlier among women compared to men. For both men and women, the difference in prevalence between high and low SES groups was mostly explained by the socio-demographic and risk behaviour variables included in the models: age, being overweight/obesity, employment status, residency, and region. However, the contribution size of these factors and consistency was different by gender and changed over time.
Conclusions
China has experienced a reversal in the SES-NCD gradient which occurred first among women. The explanations of SES differences are complex and differ by gender.
Keywords: Reversal hypothesis, Non-communicable diseases, Socioeconomic status, Gender disparities, China, Panel data
Highlights
-
•
Evidence before this study: Prior studies focus on SES and body weight, and limited research on dynamic trends and gender difference in China.
-
•
Added value of this study: This is the first study using 30-year longitudinal data in China to examine SES-NCD reversal and gender differences.
-
•
Implications of all the available evidence: Our findings highlight the need for equity-driven policies to reduce rising NCD burdens among low SES groups in China.
1. Introduction
Existing literature has established the association between socioeconomic status (SES) and the development of negative health outcomes such as non-communicable diseases (NCDs) (Wu et al., 2019). While NCDs are concentrated in low-SES populations in high income countries (HICs) (Pampel et al., 2012; Porhcisaliyan et al., 2021), patterns are much more mixed in low and middle-income countries (LMICs) (Corsi & Subramanian, 2019; Subramanian et al., 2011). For example, the prevalence of diabetes among individuals with low and high SES is 19.2 % and 25.0 % (no schooling vs. higher secondary) in India (Corsi & Subramanian, 2019), 5.47 % and 6.97 % (no formal education vs. college or above) in China (Wu et al., 2019) and 35.9 % and 29.2 % (primary school and below vs secondary and above) in Singapore (Porhcisaliyan et al., 2021), separately. The “reversal hypothesis” proposes that the burden of negative health outcomes shifts from the rich to the poor as a country's economic and social development progresses and with the process of epidemiological transition (Jones-Smith et al., 2012; McPake, 1982; Pampel et al., 2012; Zhou, 2021). However, the evidence for this hypothesis is not fully explored and the understanding of its underlying processes at earlier stages of development is limited.
NCDs pose a substantial health burden all over the world. The World Health Organisation (WHO) estimates that NCDs account for about 74 % of all deaths globally each year and 77 % of all NCD deaths are in LMICs. (World Health Organization) If NCDs become increasingly a problem for the poor, the reversal hypothesis has substantial policy implications for the reduction of poverty and health impacts. Existing studies of the reversal hypothesis mostly focus on the association between body weight and SES (Pérez-Ferrer et al., 2018) and draw from cross-sectional data (Goryakin & Suhrcke, 2014; Kinge et al., 2015). The evidence from transitioning economies is limited and findings of associations between SES and different types of NCDs are inconsistent with evidence related to metabolic syndrome related conditions more clearcut than for example for cancers, mental health conditions or chronic respiratory conditions (Corsi and Subramanian, 2012, 2019; Emamian et al., 2017; Jung et al., 2019; Monteiro et al., 2001; Nikparvar et al., 2021; Siddiquee et al., 2015; Subramanian et al., 2013; Veisani et al., 2019). A recent cross-sectional data analysis on the reversal hypothesis in China found that NCDs and risk factors were concentrated among the richer population but it did not explore trends in the relationship between SES and NCDs over time (Zhang et al., 2023).
Better understanding the processes underpinning an increasing concentration of NCDs in poorer populations as economic development and epidemiological transition progress might offer new insights into how to attenuate the process and protect poorest populations from future health burdens. Conversely, neglect of this process suggests acceptance of its inevitability. Gender differences are important because health promotion and behavioural change programs that might build on new understandings of the ‘reversal’ process have to be tailored to their target populations. Previous studies have found gender differences in the SES gradient in negative health outcomes, (Han et al., 2014) (Sánchez-Vaznaugh et al., 2009) (Yang et al., 2020), (Jiwani et al., 2019) (Zhao et al., 2018) but they have been dominated by analyses of body weight and provided limited evidence on the contributing factors or underlying explanations.
This study examines the “reversal hypothesis” in China between 1991 and 2020 and aims to clarify gender differences and to investigate the factors shaping those differences. China provides a promising setting to explore these questions. China has been undergoing rapid epidemiological and socio-demographic transition over the past 30 years and a transformation in gender roles reflected in employment and education participation. (National Bureau of Statistics of China) The experience of dramatic changes in a relatively short period allows insight into large changes within periods covered by longer-running panel data sets. Such data sets covering multiple decades as exist in China may be sufficiently long to provide new insight into the dynamic inter-relationships taking shape.
2. Methods
2.1. Data and population
This study used two longitudinal datasets: nine waves of the China Health and Nutrition Survey (CHNS, 1991–2015) and six waves of China Family Panel Studies (CFPS, 2010–2020). The two datasets allow us to examine trends over the 30-year period 1991–2020. CHNS is a panel study that collects families' and individuals’ information including nutrition intake and health status. The baseline survey was launched in nine Chinese provinces in 1989, and nine follow-up waves to date (1991, 1993, 1997, 2000, 2004, 2006, 2009, 2011 and 2015). (The University of North Carolinaa) We did not used the 1989 survey as the NCD information was not collected in this wave.
CFPS is an ongoing nationally representative longitudinal survey that collects data at individual, household and community levels in 25 provinces of China on a range of topics including socio-demographic characteristics and health status. (Institute of Social Science Survey and Peking Universitya) The baseline survey was launched in 2010, with five follow-up surveys to date (2012, 2014, 2016, 2018 and 2020). Detailed information describing the CHNS (The University of North Carolinaa; The University of North Carolinab) and CFPS (Institute of Social Science Survey and Peking Universitya; Institute of Social Science Survey and Peking Universityb) has been reported elsewhere.
To investigate how SES-NCD gradients have evolved over time, a single, long-running longitudinal dataset covering multiple decades would be ideal. In the absence of such data, we used CFPS and CHNS.
We excluded participants under 18, and those with missing information for NCDs and SES (years of formal education) and key socio-demographic variables (i.e., age and gender, after filling missing values based on information from other waves).
2.2. Measurements
Our health outcome was self-reported NCDs. In the CHNS, participants were asked about their doctor-diagnosed history of high blood pressure, diabetes, myocardial infarction, stroke or ischemic attack, cancer, and asthma. A code of 1 is given if the individuals reported any of these conditions. In the CFPS, participants were asked: “During the past six months, have you had any doctor-diagnosed chronic disease”. Each respondent who answered “Yes” to this question was asked to select the type of disease from the Disease Codebook, and the two most serious diseases were recorded if there was more than one. We categorized individuals as having NCD if they answered “Yes” to any of the NCDs on the list. The CFPS therefore identifies a more comprehensive set of NCDs and the measures from the two surveys are therefore not directly comparable. Although the NCD definitions differ slightly between those two surveys, both datasets include the most common NCDs and allow us to explore longer term trends in SES-NCD gradients over time in China.
We used participants completed years of formal education to represent SES at the individual level. Participants were categorized into two low or high SES groups based on the median years of education. There has been substantial growth and a significantly gendered pattern of educated year over the past 30 years, particularly due to the implementation of nine-year compulsory education since 1986. Consequently, we generated the quantiles of educated year by age groups and gender separately for each wave.
The following variables were included as covariates in our regression models: age (18–30, 31–40, 41–50, 51–60, 61+), gender (men, women), marital status (never married, currently married, divorced/widowed/separated), residency (urban, rural), per capita household income (Q1 to Q5, Q1 being the poorest), geographical region (western, eastern, central), employment status (no, yes), smoking (never smoked, ever smoke) and drinking (drink less than three times a week, more), overweight/obesity (if BMI ≥24 kg/m2 or not) (Fang & Liang, 2017), health insurance (no, yes).
2.3. Statistical analyses
We reported descriptive sample characteristics for baseline waves of the two datasets (i.e., 1991 in CHNS and 2010 in CFPS). We reported the prevalence of NCDs between 1991 and 2020 for the whole sample and by gender respectively.
Multivariate logistic regression models were fitted to estimate adjusted odds ratios (AOR) and 95 % confidence intervals (CI) for the associations between SES and NCDs among whole sample and by gender for each wave respectively. The regressions were adjusted by covariates. We included cluster robust standard errors at the community level to account for the observed and unobserved homogeneity across communities.
Blinder-Oaxaca (BO) decomposition models were fitted to explain the difference in the prevalence of NCDs between low and high SES groups in women and men, respectively. Based on the BO model, the NCDs prevalence gap between two groups can be decomposed into the ‘explained’ component (endowment effect), which refers to the gaps explained by differences in independent variables in the model, and the ‘unexplained’ component (coefficients effect), which captures all the differential effects of independent and unobserved variables (Okamoto et al., 2021; Rahimi & Hashemi Nazari, 2021). We divided the sample into low and high SES groups for men and women separately, then calculated the relative percentage of contribution of each variable to the difference between two SES groups. Low SES groups were treated as the reference group and covariates were controlled as independent variables. Analysis was run separately in CFPS and CHNS.
We conducted sensitivity analysis by repeating the multivariate logistic regression models in individuals under and above 50 years of age respectively. All statistical analyses were performed using STATA/MP 14.1. P-values less than 0.05 were considered to be statistically significant.
2.4. Sensitivity analyses
Given that age is closely linked with education years in China and the strong association between age and NCD prevalence, we performed a sensitivity analysis by repeating our main analysis in individuals under and above 50 years of age respectively, to investigate if the results at the population level may mask differences by age.
3. Results
We reported the results from the two datasets separately to account for any inconsistencies, given the differences in the measurement of NCD and age distribution in population between the two surveys.
3.1. Descriptive characteristics
We finally included participants from the year 1991 (n = 9936), 1993 (n = 9262), 1997 (n = 9639), 2000 (n = 10,185), 2004 (n = 10,565), 2006 (n = 10,527), 2009 (n = 10,716), 2010 (n = 32,522), 2011 (n = 13,840), 2012 (n = 30,626), 2014 (n = 30,771), 2015 (n = 14,157), 2016 (n = 28,760), 2018 (n = 27,615), and 2020 (n = 20,126). In CHNS 1991, the mean age of the study population was 40.5 years old (standard deviation [SD], 16.03 years), 49.9 % were women, the median year of formal education is 5.9 years (SD, 4.2) and 64.6 % had received 5.9 or fewer years, 75.4 % were married, 68.5 % resided in rural areas. In CFPS 2010, the mean age of the study population was 46.4 years old (SD, 15.8 years), 51.6 % were women, the median year of formal education is 6.9 years (SD, 4.8) and 64.3 % had received 6.9 or fewer years, 82.3 % were married, 53.4 % resided in rural areas. The descriptive sample characteristics of CHNS 1991 and CFPS 2010 are presented in Appendix 1 (p1-3).
3.2. NCD prevalence by gender
The overall crude prevalence trends of NCDs among Chinese adults between 1991 and 2020 are displayed in Fig. 1a and b (detailed prevalence rates are reported in Appendix 2 p4). Overall, there were large increases in the prevalence of NCDs in the overall study population between 1991 (3.52 %) and 2015 (19.51 %) and between 2010 (11.28 %) and 2020 (13.72 %). Women consistently have had higher NCD prevalence than men in both datasets but the gap between women and men has varied over the period studied.
Fig. 1.
Prevalence (%) of NCD in subpopulations (1991–2020)
Data source: China Health and Nutrition Survey (CHNS) and China Family Panel Study (CFPS). Figures were made based on the results in Appendix 2.
Note: LowSES = low SES participants. HighSES = high SES participants. Mlow = low SES male participants. Mhigh = high SES male participants. Flow = low SES female participants. Fhigh = high SES female participants.
3.3. NCD prevalence by SES group
Fig. 1c–f shows the prevalence of NCDs by SES group for the full sample and by gender over time respectively. High SES men and women consistently have had higher NCD prevalence than their low SES counterparts in both datasets. However, the gap changed overtime. Among women, the gap widened from approximately 2.06 % in 1993 to 7.74 % in 2011, then narrowed down to almost 0 % since 2011. The gap in NCD prevalence by SES group had similar pattern in men, widening from 2.97 % in 1993 to 7.87 % in 2011, then narrowing and stabilising after 2018.
3.4. The association between SES and NCD by gender
Fig. 2 shows the adjusted associations between NCD prevalence and SES obtained from multivariate logistic regression. Generally, the plots show a positive association between NCDs and SES in the whole population, but the association shifts towards to negative over time, although the association was not statistically significant in most survey years with an exception in 1997 (AOR = 1.40 [95 % CI 1.05–1.87], (Fig. 2, Overall). By gender, the pattern was more pronounced in women. There was a significant positive association in 1997 (AOR = 1.70 [95 % CI 1.13–2.56]) and becoming negative in 2016 (AOR = 0.82 [95 % CI 0.72–0.94]). For men, we did not find statistically significant associations during 1991 and 2020, except a positive relationship in 2010 (AOR = 1.16 [95 % CI 1.00–1.35]).
Fig. 2.
Association between SES and NCDs (1991–2020)
Data source: China Health and Nutrition Survey (CHNS) and China Family Panel Study (CFPS).
Note: Odds Ratio represents the associations between SES and NCDs. Each Odds Ratio was generated from multivariate logistic regression models for the associations between SES and NCDs separately. The regressions were adjusted by covariates. Cluster robust standard errors at the community level were included in the regression models. The regression was run separately in CFPS and CHNS. Overall: odds ratio was obtained from full sample. Men: odds ratio was obtained from male participants. Women: odds ratio was obtained from female participants.
3.5. Factors explaining the NCD difference between low and high SES groups by gender
Appendix 3 (p5-19) shows detailed results from the Blinder-Oaxaca decomposition model of differences in NCD prevalence between low and high SES men and women respectively. Among men, the differences in NCD prevalence by SES group were found to be statistically significant between 1991 and 2016 except for 2010, 2014 and 2015, whereas those among women were statistically significant only in 1993 and 2015.The differences in independent variables explained a large portion of the NCD difference between low and high SES individuals.
Fig. 3 plots the factors explaining the NCD differences and the size of contribution. Among men, the differential distribution of age, overweight/obesity, employment status, residency, region, and per capita household income made a significant contribution to the SES-NCD gradient. Specifically, aged 51 and above, being overweight/obesity, unemployed, living in urban area and eastern China, and had higher per capita household income accounted for high SES men's higher NCD (Fig. 3a and b). Counter-intuitively, we found lower rates of drinking and smoking accounted for the higher NCD prevalence among high SES men than low SES men in 2012 and 2016, using data from CFPS (Fig. 3b).
Fig. 3.
Factors explaining the NCD difference between low and high SES groups (1991–2020)
Fig. 3a were made based on the results in Appendix 3 Table a&e. Data source: China Health and Nutrition Survey (CHNS).
Fig. 3b were made based on the results in Appendix 3 Table b&f. Data source: China Family Panel Study (CFPS).
Fig. 3c were made based on the results in Appendix 3 Table c&g. Data source: China Health and Nutrition Survey (CHNS).
Note: Figures only included the survey years where the NCD differences between two groups (low vs. high men/women) are significant at the 5 % level (p-value <0.05). The Y-axis represents the contributing factors, and the X-axis represents the % of the contribution of these factors (numerical value of the % are labelled at the end of the bar). The coloured bar represents the statistically significant factors, and the grey colour represents the factors that are not statistically significant. The scale range of the X-axis in most years is between −40 % and 50 %, while it is between −80 % and 50 % in 1993 and -40 % and 100 % in 2009, 2012, 2015, and 2016 due to some factors are much large/small in these years than other years. Blinder-Oaxaca (BO) decomposition models and were run separately in CFPS and CHNS.
Like men, aged 41 and above, being/obesity, unemployed, living in urban area and eastern China accounted for high SES women's higher NCD prevalence (Fig. 3c).
3.6. Sensitivity analysis
Our sensitivity analysis showed that a more significant reversal in the SES-NCD gradient happened in 2016 among those who were under 50 years compared to those who were aged 50 years and above. This implies that transition occurs following the differential experience of the next population cohort (rather than because of behaviour change in the older cohort). The association was in general positive. Similar to our main results, transition among younger women occurred earlier (results are provided in Appendix 4 p20-24).
4. Discussion
This study investigated the reversal hypothesis across the SES-NCD gradient among Chinese men and women by using two nationally representative datasets covering 1991–2020. We identified a significant increase in NCD prevalence in both genders during the period and women consistently have had higher NCD prevalence than men in both datasets. Over the past 30 years, the prevalence of NCDs was always higher in high SES women and men than in their low SES counterparts in both datasets. However, we found that the association between SES and NCDs was shifting from positive towards the negative over time, occurring first in women. As might be expected, the results from younger age groups support the reversal hypothesis more clearly.
We found a significant increase in the prevalence of NCDs in the past 30 years in China. Rapid economic growth and urbanization have accompanied growing risk factors such as unhealthy diet and lack of physical activity (Li et al., 2012) and have contributed to the increase of NCDs and their predisposing factors including obesity (Pang et al., 2022) and hypertension (Miao & Wu, 2016). We found that women have a higher prevalence of NCDs than men, and this difference persists over time, in line with previous studies on obesity (Han et al., 2014). This may be partially explained by the fact that women have larger hormonal changes throughout their lifetimes and generally longer lifespan than men (Okamoto et al., 2021), and experience higher rates of mental health pressures like anxiety (Fang & Liang, 2017), increasing their risk of developing NCDs.
We found that high SES participants had higher NCD prevalence than low SES participants in both datasets and in both genders. However, the gap between low and high SES groups in men and women narrowed. The association between SES and NCDs shifted from positive to negative over time, with an earlier transition in women, in line with the evidence on education and obesity from other countries (Pérez-Ferrer et al., 2018). In our study, we also found a reverse of the SES-overweight gradient among women but not in men (Appendix 5a p25), which is likely to contribute to the earlier transition of burden of NCDs in women in China. Nevertheless, this association between NCD and SES was only significant in a few survey years, which may indicate that China is in relatively early stages of reversal.
The observed NCD differences between low and high SES groups in men and women can be mostly explained by certain independent variables in our Blinder-Oaxaca decomposition models. However, we identified different contributing factors and varying magnitudes of their contributions to the SES-NCD gradients in men and women. Specifically, age contributed most, followed by overweight/obesity, employment, residency, and region to the SES-NCD difference in men, whereas age contributed most, followed by residency, region, overweight/obesity and employment status to the SES-NCD difference in women. We found that there are certain factors only contributing to the NCD difference in one gender. For example, living in the urban area consistently contributed to the SES-NCD gradient in women between 2004 and 2015 whereas only this was only the case in 1993 and 2011 for men. Per capita household income has consistent impact for men while not for women.
Our further analysis partially explains these differences. For example, compared to men, the difference between low and high SES the proportion of the population with overweight/obesity is smaller in men and almost disappeared after 2005 (Appendix 5a p25). It is noteworthy that higher employment rates among low SES participants accounts for a lower prevalence of NCDs among low SES participants compared with high SES participants in both women and men. We found that low SES participants had a higher employment rate than high SES participants (Appendix 5b p26), especially among participants aged 51 and above in men and aged 41 and above in women (Appendix 6a p34). This might imply higher rates of physical activity in older age groups among participants of low SES compared to their higher SES counterparts. We found a strong negative association between employment and NCDs (Appendix 6b p35).
Counter-intuitively, we found a contribution of current or previous smoking and frequent drinking contributing to the NCD difference in men in 2012 and 2016 (CFPS survey). This may be due to the economic and epidemiological shifts over time during which the factors affecting NCD prevalence have become more complex and our results may be related to omitted variable bias (Pan & Palmer, 2018). Another possible explanation is that participants may change their behaviour after NCDs have been diagnosed (Pan & Palmer, 2018), but our repeated cross-sectional design cannot infer causality. The above results may indicate that economic development and urbanization have more consistent and/or more significant impact on women's health than men's, whereas individual/household income has more impact on men's health.
Our sensitivity analysis showed that the reversal pattern was more pronounced among younger participants, indicating that general population effects may obscure important age and gender differences and that transition is a cohort more than whole population effect. Younger participants have a number of differences from older ones besides age. They have received more years of education (Appendix Table 6c p36), they may have a higher health literacy than older participants (Pérez-Ferrer et al., 2018), and they've also benefited from technical improvements and better understandings of the risk factors for NCDs at younger ages. Another possible explanation of earlier reversal of NCD prevalence in younger groups is that different generations of the Chinese population have experienced stages of the life cycle under markedly different conditions. Older participants may have experienced more disadvantaged situations such as low parental SES, food shortages, or bad health during their infancy, childhood, or adulthood than younger generations, all factors that affect NCD prevalence (Ding et al., 2021; Ogunsina et al., 2018) and may then influence the NCD-SES relationship.
5. Limitations
We present the first study that assesses the reversal hypothesis in SES-NCDs gradient using population-based longitudinal datasets in a country that has progressed from low to upper middle-income status over 30 years. However, there are several caveats when interpreting the results. However, there are several caveats when interpreting the results. Firstly, the use of self-reported information including the presence of NCDs, smoking and drinking might be subject to recall or social desirability bias. Secondly, causation between independent variables and NCDs cannot be conclusively attributed from repeated cross-sectional data. Limitations also arise from the use of two datasets that are different in some respects (e.g., different questions on self-reported NCDs and list of included NCDs, age distribution of population, questions to measure drinking behaviour). Those differences may limit the consistency of our results between the two datasets.
6. Conclusion
In conclusion, we found that the burden of NCDs is shifting towards low SES groups, and this process is occurring earlier in women. Therefore, special attention needs to be paid to disadvantaged groups in the development of prevention policies for NCDs. Our study also highlighted that the most important common factors affecting the difference of NCD prevalence between high and low SES groups were different in women and men. This implies the importance of developing gender-specific dynamic interventions to address the emerging socioeconomic disparities in NCDs.
CRediT authorship contribution statement
Xuemei Zhang: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Resources, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Tianxin Pan: Writing – review & editing, Supervision. Barbara McPake: Writing – review & editing, Supervision, Methodology.
Ethical approval
This research analysed the publicly available secondary datasets, and we consider that no ethical approval is required.
Data sharing
All data are publicly available at CHNS and CFPS website. Certain authors had full access to the data in the study.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix 1. Descriptive characteristics of sample
Table 1a.
Descriptive characteristics of sample in CHNS (1991)
| Variables | All |
Men (%) |
Women (%) |
|---|---|---|---|
| 9936 | N = 4978 (50.10) | N = 4958 (49.90) | |
| NCDs | |||
| No | 96.48 | 97.03 | 95.93 |
| Yes | 3.52 | 2.97 | 4.07 |
| Education | |||
| Low | 64.57 | 60.41 | 68.76 |
| High | 35.43 | 39.59 | 31.24 |
| Age | |||
| 18–30 | 32.52 | 32.90 | 32.13 |
| 31–40 | 23.92 | 23.84 | 24.00 |
| 41–50 | 16.46 | 16.81 | 16.10 |
| 51–60 | 13.04 | 12.82 | 13.27 |
| 61+ | 14.06 | 13.62 | 14.50 |
| Employment status | |||
| No | 19.35 | 15.02 | 23.71 |
| Yes | 80.65 | 84.98 | 76.29 |
| Overweigh | |||
| No | 81.52 | 85.29 | 78.03 |
| Yes | 18.48 | 14.71 | 21.97 |
| Health insurance | |||
| No | 68.61 | 65.45 | 71.78 |
| Yes | 31.39 | 34.55 | 28.22 |
| Smoking | |||
| Never | 65.37 | 31.97 | 95.75 |
| Ever/current | 34.63 | 68.03 | 4.25 |
| Drinking | |||
| No | 84.82 | 71.25 | 97.36 |
| Yes | 15.18 | 28.75 | 2.64 |
| Residency | |||
| Urban | 31.53 | 30.78 | 32.29 |
| Rural | 68.47 | 69.22 | 67.71 |
| Region | |||
| Western | 27.56 | 27.80 | 27.31 |
| Central | 37.92 | 38.55 | 37.29 |
| Eastern | 34.52 | 33.65 | 35.40 |
| Marital status | |||
| Never Married | 17.26 | 19.34 | 15.17 |
| Married | 75.42 | 75.42 | 75.42 |
| DWS | 7.32 | 5.24 | 9.41 |
| Household Income | |||
| Q1 (the lowest) | 20.01 | 20.02 | 20.00 |
| Q2 | 20.00 | 20.00 | 20.00 |
| Q3 | 20.01 | 20.02 | 20.00 |
| Q4 | 20.10 | 19.98 | 20.23 |
| Q5 (the highest) | 19.88 | 20.00 | 19.76 |
Table 1b.
Descriptive characteristics of sample in CFPS (2010)
| Variables | Overal32,522 | Men (%) N = 15,737 (48.39) | Women (%) N = 16,785 (51.61) |
|---|---|---|---|
| NCDs | |||
| No | 87.06 | 87.94 | 86.23 |
| Yes | 12.94 | 12.06 | 13.77 |
| Education | |||
| Low | 64.34 | 69.56 | 59.46 |
| High | 35.66 | 30.44 | 40.54 |
| Age | |||
| 18–30 | 18.52 | 17.99 | 19.01 |
| 31–40 | 18.98 | 18.49 | 19.43 |
| 41–50 | 22.80 | 22.39 | 23.18 |
| 51–60 | 19.87 | 20.19 | 19.57 |
| 61+ | 19.84 | 20.94 | 18.81 |
| Employment status | |||
| No | 20.66 | 15.82 | 25.19 |
| Yes | 79.34 | 84.18 | 74.81 |
| Overweight | |||
| No | 71.45 | 69.98 | 72.88 |
| Yes | 28.55 | 30.02 | 27.12 |
| Health insurance | |||
| No | 18.28 | 17.19 | 19.31 |
| Yes | 81.72 | 82.81 | 80.69 |
| Smoking | |||
| Never | 62.91 | 28.68 | 95.01 |
| Ever/Current | 37.09 | 71.32 | 4.99 |
| Drinking | |||
| No | 83.88 | 70.00 | 96.89 |
| Yes | 16.12 | 30.00 | 3.11 |
| Residency | |||
| Urban | 46.60 | 45.90 | 47.24 |
| Rural | 53.40 | 54.10 | 52.76 |
| Region | |||
| West | 27.33 | 27.83 | 26.87 |
| Central | 29.48 | 29.17 | 29.78 |
| East | 43.19 | 43.01 | 43.35 |
| Marital status | |||
| Never Married | 10.17 | 12.68 | 7.81 |
| Married | 82.26 | 81.88 | 82.62 |
| Divorce/Widow | 7.57 | 5.44 | 9.57 |
| Household Income | |||
| Q1 (the lowest) | 20.01 | 20.00 | 20.02 |
| Q2 | 20.00 | 20.01 | 19.99 |
| Q3 | 20.03 | 20.06 | 20.00 |
| Q4 | 20.05 | 20.07 | 20.02 |
| Q5 (the highest) | 19.92 | 19.86 | 19.97 |
Appendix 2. NCDs prevalence in subpopulations (1991–2020)
Table 2a.
NCD prevalence in subpopulations in CHNS (1991–2015)
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 1991 | 3.52 | 2.97 | 4.07 | 3.02 | 4.43 | 2.33 | 3.96 | 3.64 | 5.04 |
| 1993 | 3.88 | 3.56 | 4.19 | 3.02 | 5.53 | 2.52 | 5.49 | 3.52 | 5.58 |
| 1997 | 5.42 | 4.77 | 6.07 | 4.29 | 7.6 | 3.75 | 6.55 | 4.8 | 8.83 |
| 2000 | 6.76 | 5.87 | 7.69 | 5.58 | 8.9 | 4.49 | 8.41 | 6.74 | 9.41 |
| 2004 | 10.05 | 9.27 | 10.83 | 8.3 | 13.07 | 7.74 | 11.72 | 8.83 | 14.56 |
| 2006 | 11.29 | 11.34 | 11.25 | 9.64 | 14.23 | 10.19 | 13.45 | 9.08 | 14.95 |
| 2009 | 15.46 | 15.38 | 15.55 | 13.78 | 18.6 | 13.34 | 19.39 | 14.2 | 17.91 |
| 2011 | 19.75 | 19.59 | 19.89 | 16.61 | 24.42 | 16.54 | 24.41 | 16.68 | 24.42 |
| 2015 | 19.51 | 20 | 19.05 | 18.6 | 20.97 | 19.89 | 20.19 | 17.31 | 21.63 |
Table 2b.
NCD prevalence in subpopulations in CFPS (2010–2020)
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 2010 | 12.94 | 12.06 | 13.77 | 12.79 | 13.22 | 11.89 | 12.44 | 13.77 | 13.77 |
| 2012 | 11.28 | 10.00 | 12.50 | 10.87 | 12.16 | 9.58 | 11.09 | 12.24 | 12.95 |
| 2014 | 15.37 | 13.57 | 17.11 | 14.94 | 16.09 | 13.23 | 14.26 | 16.86 | 17.44 |
| 2016 | 15.36 | 13.38 | 17.27 | 14.74 | 16.39 | 12.63 | 14.94 | 17.11 | 17.48 |
| 2018 | 15.88 | 13.78 | 17.93 | 15.29 | 16.73 | 13.34 | 14.58 | 17.55 | 18.38 |
| 2020 | 13.72 | 12.65 | 14.80 | 13.62 | 13.87 | 12.59 | 12.73 | 14.69 | 14.95 |
Note: All = all participants. Male = male participants. Female = female participants. NCD = noncommunicable diseases. SES = socioeconomic status. CHNS=China Health and Nutrition Survey. CFPS=China Family Panel Study.
Appendix 3. Blinder-Oaxaca decomposition between low and high SES groups (1991–2020)
Table 3a.
Blinder-Oaxaca decomposition between low and high SES male in CHNS (1991–2015)
Table 3b.
Blinder-Oaxaca decomposition between low and high SES male in CFPS (2010–2020)
Table 3c.
Blinder-Oaxaca decomposition between low and high SES female in CHNS (1991–2015)
Table 3d.
Blinder-Oaxaca decomposition between low and high SES female in CFPS (2010–2020)
Table 3e.
The proportion of each factor contribute (%) to the NCD difference between low and high SES males in CHNS (1991–2015)
| Factors | 1991 | 1993 | 1997 | 2000 | 2004 | 2006 | 2009 | 2011 |
|---|---|---|---|---|---|---|---|---|
| Low-M | ||||||||
| High-M | ||||||||
| explained | 87.24 | 84.76 | 73.68 | 108.96 | 95.20 | 97.28 | 80.00 | 80.75 |
| unexplained | 12.65 | 15.15 | 26.46 | −9.00 | 4.90 | 2.74 | 20.00 | 19.37 |
| Explained | ||||||||
| 31–40 | −1.19 | −0.75 | 0.13 | −0.12 | −0.55 | −1.33 | −1.69 | 1.08 |
| 41–50 | −0.06 | 4.12 | 2.63 | −5.52 | −0.06 | 3.60 | −1.27 | 0.27 |
| 51–60 | 12.45 | 12.13 | −0.74 | 7.63 | 22.80 | −30.82 | −15.12 | −19.98 |
| 61+ | 6.28 | 19.18 | 20.73 | 28.49 | −0.06 | 22.63 | 26.46 | 42.48 |
| Q2 | −1.80 | −1.26 | −1.06 | 1.47 | −1.12 | −2.84 | −0.94 | −5.96 |
| Q3 | −0.74 | 0.24 | −0.02 | −0.20 | −2.34 | 0.66 | −0.57 | −0.34 |
| Q4 | 2.37 | 0.65 | −3.71 | −0.90 | 0.27 | 4.83 | 0.71 | 5.19 |
| Q5 | 2.98 | 0.84 | −1.94 | 15.34 | 17.25 | 40.18 | 30.08 | 2.49 |
| Employment | 26.28 | 7.29 | 12.69 | 15.70 | 10.63 | 9.37 | 15.43 | 7.51 |
| Married | 8.06 | −0.52 | 1.44 | −0.03 | 0.10 | −4.59 | −1.76 | 1.16 |
| Divorced/Widowed | 2.58 | 0.07 | −0.60 | 0.39 | −0.45 | −0.96 | −0.75 | −0.57 |
| Residency | 13.06 | 10.79 | 5.67 | 12.39 | 13.38 | 24.62 | 9.15 | 20.22 |
| Smoking | 1.52 | 1.36 | 0.93 | 2.45 | −1.11 | −0.82 | −0.91 | 0.25 |
| Drinking | 2.05 | 0.03 | −0.86 | −0.04 | 0.43 | 2.33 | 0.99 | −1.10 |
| Insurance | −8.57 | 12.38 | 8.80 | 9.80 | 13.31 | 1.26 | −0.45 | 0.10 |
| Overweight | 23.42 | 16.55 | 27.69 | 20.32 | 15.56 | 29.94 | 13.32 | 10.70 |
| Central | −0.08 | −0.07 | 0.01 | 0.48 | 0.24 | −0.64 | −0.05 | −3.62 |
| East | −1.26 | 1.81 | 1.77 | 1.25 | 6.84 | −0.35 | 7.24 | 20.82 |
| Unexplained | ||||||||
| 31–40 | 14.64 | −7.68 | 10.12 | 10.06 | 6.89 | −6.86 | 0.71 | 2.35 |
| 41–50 | −4.66 | 0.68 | 17.81 | 8.43 | 5.28 | 29.40 | 21.57 | 16.61 |
| 51–60 | 0.59 | 4.39 | −6.32 | 55.98 | 36.11 | 22.93 | 19.69 | 18.65 |
| 61+ | 80.61 | 57.01 | 34.50 | 30.88 | 17.40 | 32.02 | 54.02 | 25.99 |
| Q2 | −12.86 | 5.03 | −0.13 | 4.10 | −6.31 | −15.20 | 1.73 | −4.40 |
| Q3 | 13.57 | 7.77 | −23.68 | −5.98 | 0.48 | 3.78 | 0.13 | 2.26 |
| Q4 | −28.16 | 5.67 | −22.22 | −7.53 | 10.20 | −45.62 | 13.92 | −7.98 |
| Q5 | −14.03 | −0.68 | 14.88 | −22.51 | −6.52 | −18.58 | −5.26 | 21.90 |
| Employs | 132.14 | 36.59 | −4.91 | 3.57 | 1.49 | −39.58 | 24.88 | −56.68 |
| Married | 7.91 | 23.93 | 14.94 | −26.69 | 16.74 | 47.43 | −8.57 | 18.17 |
| Divorced/Widowed | −33.21 | −3.78 | −2.52 | 0.26 | 2.28 | −6.34 | 2.54 | −3.77 |
| Residency | 38.83 | 89.33 | −56.14 | −59.76 | 4.90 | −49.85 | −1.83 | 107.70 |
| Smoking | −47.24 | −0.34 | −25.12 | 1.91 | −17.15 | −38.67 | 2.20 | −1.40 |
| Drinking | 10.36 | 6.74 | 22.95 | −0.55 | 0.16 | −2.40 | −8.54 | −16.00 |
| Insurance | 76.02 | 7.53 | −1.62 | −21.51 | 18.36 | 36.25 | −33.70 | 37.91 |
| Overweight | −8.11 | −0.43 | −0.71 | 4.68 | 50.76 | −24.89 | −8.27 | 24.91 |
| Central | 40.87 | 7.62 | 7.69 | 15.48 | 30.05 | 19.70 | 6.80 | −16.49 |
| East | 69.90 | −17.65 | −11.37 | 38.45 | 14.92 | 39.88 | 14.91 | −1.35 |
Note: The table only included in years where the “NCD differences” between low and high SES groups were significant at the 5 % level in CHNS for males.
Table 3f.
The proportion of each factor contribute (%) to the NCD difference between low and high SES males in CFPS (2010–2020)
| Year | 2012 | 2016 |
|---|---|---|
| Low-M | ||
| High-M | ||
| explained | 13.09 | 91.75 |
| unexplained | 87.04 | 8.25 |
| Explained | ||
| 31–40 | 0.72 | 3.43 |
| 41–50 | −33.33 | −30.41 |
| 51–60 | −23.27 | −20.86 |
| 61+ | −2.77 | 86.60 |
| Q2 | −3.09 | 2.40 |
| Q3 | −1.70 | 4.88 |
| Q4 | −0.21 | −2.48 |
| Q5 | 39.44 | −8.90 |
| Employment | −2.70 | 29.97 |
| Married | 1.39 | −3.57 |
| Divorced/Widowed | 1.54 | −0.07 |
| Residency | −6.54 | 6.98 |
| Smoking | 13.58 | 8.93 |
| Drinking | 14.88 | 9.83 |
| Insurance | −3.42 | 0.14 |
| Overweight | 12.72 | 5.36 |
| Central | 2.70 | 3.05 |
| East | 3.14 | −3.47 |
| Unexplained | ||
| 31–40 | −11.05 | 9.14 |
| 41–50 | −5.91 | 19.62 |
| 51–60 | −10.68 | 13.16 |
| 61+ | 20.86 | 38.83 |
| Q2 | −5.86 | −19.69 |
| Q3 | 14.44 | 1.52 |
| Q4 | 45.12 | −9.59 |
| Q5 | 68.52 | −33.02 |
| Employs | −54.81 | −3.92 |
| Married | 133.95 | −32.58 |
| Divorced/Widowed | 2.79 | 7.15 |
| Residency | 57.16 | 20.79 |
| Smoking | −23.95 | −57.73 |
| Drinking | 44.81 | 17.59 |
| Insurance | 85.19 | 15.46 |
| Overweight | 41.48 | 47.42 |
| Central | 16.30 | −36.08 |
| East | 24.20 | −59.45 |
Note: The table only included in years where the “NCD differences” between low and high SES groups were significant at the 5 % level in CFPS for males.
Table 3g.
The proportion of each factor contribute (%) to the NCD difference between low and high SES females in CHNS (1991–2015)
| Factors | 1993 | 1997 | 2000 | 2004 | 2006 | 2009 | 2011 | 2015 |
|---|---|---|---|---|---|---|---|---|
| Low-F | ||||||||
| High-F | ||||||||
| Explained | 40.30 | 21.77 | 47.38 | 85.25 | 81.62 | 123.12 | 88.90 | 106.42 |
| unexplained | 59.88 | 78.11 | 52.62 | 14.91 | 18.22 | −23.12 | 11.11 | −6.54 |
| Explained | ||||||||
| 31–40 | 1.84 | −0.25 | 1.93 | −0.97 | −0.72 | −2.55 | −0.30 | 0.98 |
| 41–50 | 8.98 | 3.73 | 4.21 | −4.34 | −3.64 | −11.50 | −8.69 | −10.74 |
| 51–60 | 21.20 | 10.92 | 31.68 | 16.33 | 10.63 | 11.21 | 14.80 | −22.76 |
| 61+ | −72.46 | −12.30 | −8.43 | 40.82 | 25.42 | 71.15 | 33.55 | 93.97 |
| Q2 | −9.16 | 4.52 | −2.46 | −3.10 | 0.35 | −1.12 | 4.17 | 2.12 |
| Q3 | 1.85 | 0.03 | −1.23 | −0.43 | −0.84 | −1.48 | 1.10 | 0.06 |
| Q4 | 13.83 | 0.12 | 4.21 | 0.80 | 2.25 | 1.38 | 0.02 | 3.60 |
| Q5 | 22.63 | −2.81 | 8.22 | 13.74 | 10.90 | 15.18 | −5.69 | −0.99 |
| Employs | 3.01 | 7.10 | 5.65 | 3.48 | 9.83 | 7.61 | 5.37 | −1.37 |
| Married | −0.09 | 0.28 | 1.26 | −0.17 | 2.04 | 3.89 | 0.10 | 1.04 |
| Divorced/Widowed | 7.90 | −2.53 | −0.60 | 0.19 | −0.36 | 0.00 | −1.43 | 0.23 |
| Residency | 28.20 | −0.81 | 8.40 | 20.24 | 18.68 | 21.94 | 26.66 | 26.46 |
| Smoking | 0.77 | 1.25 | −0.24 | −0.09 | 0.28 | 0.95 | −0.26 | 0.05 |
| Drinking | −0.19 | 0.03 | 0.27 | 0.10 | 0.21 | −1.17 | 0.18 | 0.00 |
| Insurance | −5.27 | 2.42 | −23.77 | −4.32 | 1.50 | 0.83 | 0.24 | 0.10 |
| Overweight | 20.18 | 7.00 | 10.34 | −0.10 | −2.39 | −4.25 | 3.02 | −6.46 |
| Central | −6.71 | 2.90 | −0.83 | 1.34 | 3.11 | 0.50 | −4.89 | −7.10 |
| East | 3.52 | 0.16 | 8.90 | 1.71 | 4.38 | 10.57 | 20.92 | 27.24 |
| Unexplained | ||||||||
| 31–40 | −7.84 | −3.76 | 13.25 | 7.39 | 2.33 | −0.49 | 2.93 | 1.20 |
| 41–50 | 57.90 | 8.99 | 26.70 | 5.39 | 2.31 | 8.83 | 10.52 | −11.38 |
| 51–60 | 30.42 | 33.87 | 8.30 | 22.47 | 12.13 | 11.84 | 2.13 | −2.06 |
| 61+ | 3.37 | 47.47 | 30.63 | 22.47 | 34.15 | 14.94 | 24.87 | 8.00 |
| Q2 | 5.97 | 10.09 | −24.14 | 8.83 | −7.21 | −1.58 | 1.48 | 4.40 |
| Q3 | −11.38 | 8.71 | −30.63 | 0.10 | −20.52 | −7.27 | 2.58 | 35.41 |
| Q4 | −27.78 | −2.58 | −65.45 | 3.05 | −20.52 | −34.78 | −1.04 | 48.05 |
| Q5 | −10.84 | 22.74 | −62.57 | −7.51 | −24.81 | −42.29 | 22.58 | 52.33 |
| Employs | −4.94 | 11.31 | −63.61 | −6.26 | 2.13 | −1.63 | −12.33 | −22.96 |
| Married | 34.07 | 40.09 | −84.29 | −52.83 | 44.41 | 73.52 | −10.20 | 33.46 |
| Divorced/Widowed | 12.87 | −13.82 | 8.80 | −15.06 | −10.69 | −10.20 | 6.06 | 1.64 |
| Residency | 37.72 | 42.86 | −3.12 | 21.44 | −25.57 | −100.59 | 53.83 | 120.43 |
| Smoking | −1.93 | 0.41 | −9.27 | 0.14 | −5.33 | −8.00 | −1.52 | 4.40 |
| Drinking | −9.82 | −3.80 | 9.06 | 2.81 | 0.84 | 4.41 | −1.40 | 1.39 |
| Insurance | 46.71 | 8.85 | 54.71 | 40.65 | 6.94 | 41.11 | −36.10 | 23.15 |
| Overweight | 57.31 | 6.04 | 26.44 | 36.54 | 28.79 | −13.06 | 5.66 | −23.35 |
| Central | 9.76 | 1.53 | −18.01 | 16.38 | −9.69 | −27.08 | −17.22 | −19.09 |
| East | 34.73 | 17.21 | −18.53 | 4.96 | −5.02 | −12.83 | −23.60 | 1.03 |
Note: The table only included in years where the “NCD differences” between low and high SES groups were significant at the 5 % level in CHNS for females.
Appendix 4. Sensitivity analysis
Fig. 4a.
NCD prevalence among population≤50 (1991–2020).
Fig. 4b.
NCD prevalence among population >50 (1991–2020)
Data source: China Health and Nutrition Survey (CHNS) and China Family Panel Study (CFPS).
Note: LowSES = low SES participants. HighSES = high SES participants. Mlow = low SES male participants. Mhigh = high SES male participants. Flow = low SES female participants. Fhigh = high SES female participants.
Fig. 4c.
Association between SES and NCD among population in different age groups in CHNS (1991–2015)
Data source: China Health and Nutrition Survey (CHNS) and China Family Panel Study (CFPS).
Note: Odds Ratio represents the associations between SES and NCDs. Each Odds Ratio was generated from multivariate logistic regression models for the associations between SES and NCDs in participants aged at 50 and under or above separately. The regressions were adjusted by covariates. Cluster robust standard errors at the community level were included in the regression. regression models. Overall (≤50): odds ratio was obtained from full sample that participants aged at 50 and under. Men (≤50): odds ratio was obtained from male participants aged at 50 and under. Women (>50): odds ratio was obtained from female participants aged at 50 and under. Overall (>50): odds ratio was obtained from full sample that participants aged at 50 above. Men (>50): odds ratio was obtained from male participants aged at 5o above. Women (≤50): odds ratio was obtained from female participants aged at 50 above.
Fig. 4d.
Association between SES and NCD among population in different age groups in CFPS (2010–2020)
Data source: China Health and Nutrition Survey (CHNS) and China Family Panel Study (CFPS).
Note: Odds Ratio represents the associations between SES and NCDs. Each Odds Ratio was generated from multivariate logistic regression models for the associations between SES and NCDs in participants aged at 50 and under or above separately. The regressions were adjusted by covariates. Cluster robust standard errors at the community level were included in the regression. regression models. Overall (≤50): odds ratio was obtained from full sample that participants aged at 50 and under. Men (≤50): odds ratio was obtained from male participants aged at 50 and under. Women (>50): odds ratio was obtained from female participants aged at 50 and under. Overall (>50): odds ratio was obtained from full sample that participants aged at 50 above. Men (>50): odds ratio was obtained from male participants aged at 5o above. Women (≤50): odds ratio was obtained from female participants aged at 50 above.
Appendix 5. Descriptive characteristics of each factor in Blinder-Oaxaca decomposition (1991–2020)
Table 5a.
Overweight prevalence (%) by SES and gender (1991–2020)
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 1991 | 18.48 | 14.71 | 21.97 | 11.62 | 19.37 | 19.99 | 26.18 |
| 1993 | 21.04 | 17.32 | 24.47 | 14.26 | 22.79 | 21.94 | 29.52 |
| 1997 | 26.12 | 23.01 | 29.09 | 18.51 | 30.71 | 26.66 | 34.33 |
| 2000 | 32.25 | 29.78 | 34.63 | 25.15 | 37.99 | 32.36 | 38.88 |
| 2004 | 35.40 | 33.86 | 36.88 | 29.21 | 41.04 | 36.91 | 36.84 |
| 2006 | 37.31 | 36.55 | 38.02 | 33.09 | 42.76 | 38.86 | 36.59 |
| 2009 | 39.84 | 40.21 | 39.49 | 37.16 | 46.28 | 40.31 | 38.02 |
| 2011 | 44.62 | 45.91 | 43.44 | 42.16 | 51.76 | 42.66 | 44.54 |
| 2015 | 49.20 | 50.89 | 47.69 | 47.79 | 56.43 | 49.07 | 45.64 |
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 2010 | 28.55 | 30.02 | 27.12 | 27.52 | 35.68 | 25.55 | 29.32 |
| 2012 | 30.53 | 32.42 | 28.66 | 30.16 | 38.19 | 27.70 | 30.26 |
| 2014 | 33.08 | 35.42 | 30.77 | 33.58 | 39.11 | 30.71 | 30.86 |
| 2016 | 35.65 | 38.36 | 32.93 | 36.39 | 42.12 | 33.09 | 32.75 |
| 2018 | 37.96 | 40.44 | 35.49 | 38.20 | 44.44 | 36.58 | 34.22 |
| 2020 | 38.63 | 42.38 | 34.81 | 40.53 | 45.09 | 37.02 | 31.94 |
Note: All = all participants. Male = male participants. Female = female participants.
Table 5b.
Employment rate (%) by SES and gender (1991–2020)
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 1991 | 80.65 | 84.98 | 76.29 | 81.94 | 78.29 | 88.02 | 80.36 | 76.58 | 75.66 |
| 1993 | 79.85 | 84.6 | 75.03 | 81.42 | 76.79 | 87.17 | 79.87 | 75.80 | 73.45 |
| 1997 | 76.91 | 82.26 | 71.48 | 79.62 | 71.64 | 85.61 | 76.37 | 73.95 | 66.06 |
| 2000 | 72.69 | 78.23 | 66.86 | 77.09 | 64.62 | 83.69 | 68.20 | 70.14 | 60.86 |
| 2004 | 62.33 | 71.32 | 53.45 | 65.15 | 57.46 | 74.06 | 66.95 | 56.86 | 47.07 |
| 2006 | 61.78 | 70.2 | 53.63 | 65.34 | 55.48 | 71.84 | 67.16 | 58.87 | 44.71 |
| 2009 | 60.57 | 70.07 | 51.5 | 64.21 | 53.78 | 73.34 | 63.62 | 55.16 | 45.07 |
| 2011 | 59.48 | 68.83 | 50.83 | 63.16 | 53.99 | 71.92 | 63.95 | 54.68 | 45.39 |
| 2015 | 49.04 | 58.56 | 40.13 | 46.72 | 52.77 | 54.83 | 65.10 | 38.60 | 42.40 |
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 2010 | 79.34 | 84.18 | 74.81 | 78.20 | 81.41 | 84.14 | 84.27 | 71.70 | 79.40 |
| 2012 | 53.34 | 61.11 | 45.96 | 52.14 | 55.92 | 59.96 | 64.09 | 43.78 | 49.87 |
| 2014 | 71.52 | 79.33 | 64.01 | 75.42 | 65.14 | 82.75 | 72.37 | 67.17 | 59.79 |
| 2016 | 71.60 | 78.87 | 64.56 | 75.26 | 65.45 | 82.68 | 71.02 | 66.94 | 61.27 |
| 2018 | 72.65 | 79.79 | 65.67 | 76.34 | 67.30 | 83.74 | 72.70 | 67.77 | 63.16 |
| 2020 | 73.49 | 81.41 | 65.55 | 77.01 | 68.59 | 85.55 | 75.30 | 68.04 | 62.24 |
Note: All = all participants. Male = male participants. Female = female participants. Low SES = low SES participants. High SES = high SES participants.
Table 5c.
Sample distribution (%) in different age groups by SES and gender (1991–2020)
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | |
| 1991 | 40.51 | 20.25 | 16.40 | 11.24 | 11.61 | 21.31 | 29.33 | 17.45 | 15.22 | 16.69 | 37.28 | 19.54 | 12.91 | 12.53 | 17.75 | 20.79 | 33.83 | 23.11 | 14.91 | 7.36 |
| 1993 | 39.92 | 23.24 | 16.02 | 10.69 | 10.13 | 22.70 | 20.20 | 24.28 | 16.29 | 16.53 | 34.75 | 20.99 | 14.18 | 11.56 | 18.51 | 17.53 | 28.82 | 28.29 | 17.46 | 7.90 |
| 1997 | 34.30 | 20.68 | 20.42 | 13.40 | 11.20 | 23.14 | 20.82 | 24.44 | 13.09 | 18.51 | 29.38 | 22.75 | 19.63 | 10.21 | 18.04 | 19.39 | 18.46 | 28.35 | 21.25 | 12.55 |
| 2000 | 29.67 | 20.99 | 25.04 | 13.14 | 11.16 | 25.84 | 19.92 | 16.40 | 17.70 | 20.14 | 22.67 | 24.70 | 21.64 | 11.90 | 19.11 | 20.14 | 15.80 | 27.61 | 20.59 | 15.86 |
| 2004 | 16.06 | 25.38 | 22.69 | 16.93 | 18.94 | 15.08 | 15.77 | 23.53 | 26.64 | 18.98 | 13.60 | 26.24 | 26.18 | 16.69 | 17.29 | 15.00 | 13.48 | 18.62 | 25.72 | 27.18 |
| 2006 | 10.13 | 21.99 | 22.02 | 27.12 | 18.74 | 16.68 | 16.90 | 29.31 | 14.87 | 22.23 | 10.15 | 24.70 | 25.62 | 20.54 | 19.00 | 14.24 | 12.07 | 19.70 | 26.92 | 27.07 |
| 2009 | 10.14 | 19.71 | 23.14 | 26.54 | 20.46 | 17.28 | 14.72 | 21.72 | 18.08 | 28.20 | 10.26 | 20.95 | 25.92 | 22.23 | 20.63 | 14.34 | 11.42 | 16.55 | 27.06 | 30.63 |
| 2011 | 13.48 | 14.19 | 23.59 | 27.69 | 21.06 | 7.26 | 16.80 | 23.67 | 19.44 | 32.83 | 13.88 | 15.73 | 26.59 | 21.22 | 22.58 | 8.92 | 15.23 | 17.91 | 27.00 | 30.93 |
| 2015 | 13.26 | 10.62 | 21.03 | 20.62 | 34.47 | 11.95 | 18.06 | 21.96 | 27.27 | 20.76 | 12.30 | 12.12 | 24.74 | 24.72 | 26.12 | 10.90 | 16.55 | 15.84 | 18.35 | 38.35 |
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | |
| 2010 | 13.24 | 19.09 | 24.41 | 22.62 | 20.64 | 28.85 | 17.12 | 17.78 | 14.63 | 21.62 | 17.7 | 26 | 19.56 | 16.48 | 20.59 | 20.94 | 10.29 | 28.48 | 24.1 | 16.19 |
| 2012 | 16.96 | 16.33 | 25.68 | 19.43 | 21.60 | 29.71 | 16.71 | 16.66 | 16.40 | 20.52 | 19.67 | 19.80 | 22.34 | 15.62 | 22.58 | 23.47 | 10.70 | 27.25 | 23.10 | 15.48 |
| 2014 | 15.48 | 15.44 | 27.08 | 20.04 | 21.97 | 29.69 | 15.10 | 13.37 | 16.51 | 25.32 | 19.01 | 19.08 | 21.02 | 18.73 | 22.16 | 24.39 | 10.04 | 24.80 | 19.24 | 21.53 |
| 2016 | 19.88 | 14.04 | 23.71 | 20.25 | 22.12 | 22.50 | 17.58 | 12.68 | 16.58 | 30.65 | 24.82 | 16.56 | 18.76 | 17.01 | 22.84 | 19.46 | 11.05 | 22.63 | 21.41 | 25.45 |
| 2018 | 17.19 | 14.11 | 23.45 | 22.29 | 22.96 | 20.83 | 19.89 | 12.19 | 13.99 | 33.10 | 19.70 | 17.09 | 18.60 | 19.03 | 25.59 | 18.33 | 13.09 | 21.64 | 21.05 | 25.88 |
| 2020 | 18.60 | 16.76 | 21.45 | 24.70 | 18.49 | 24.14 | 24.60 | 12.98 | 11.62 | 26.65 | 20.86 | 18.51 | 24.68 | 17.53 | 18.42 | 23.18 | 21.49 | 9.25 | 22.27 | 23.81 |
Table 5d.
Sample distribution (%) in different region by SES and gender (1991–2020)
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| West | Central | East | West | Central | East | West | Central | East | West | Central | East | |
| 1991 | 29.63 | 39.47 | 30.89 | 25.01 | 37.14 | 37.85 | 28.37 | 39.04 | 32.59 | 24.98 | 33.44 | 41.58 |
| 1993 | 30.49 | 37.44 | 32.08 | 26.17 | 38.01 | 35.81 | 28.42 | 39.18 | 32.40 | 26.96 | 34.73 | 38.31 |
| 1997 | 29.09 | 48.09 | 22.82 | 25.96 | 47.57 | 26.47 | 29.56 | 45.98 | 24.46 | 24.10 | 49.67 | 26.23 |
| 2000 | 28.05 | 44.70 | 27.25 | 22.48 | 45.22 | 32.30 | 27.29 | 44.77 | 27.94 | 22.88 | 44.32 | 32.80 |
| 2004 | 27.76 | 43.11 | 29.12 | 21.01 | 43.40 | 35.59 | 25.87 | 41.89 | 32.25 | 20.14 | 44.23 | 35.63 |
| 2006 | 26.73 | 42.49 | 30.78 | 19.98 | 44.57 | 35.46 | 27.69 | 40.81 | 31.49 | 19.55 | 44.19 | 36.26 |
| 2009 | 27.26 | 42.94 | 29.80 | 18.87 | 42.64 | 38.49 | 25.92 | 43.24 | 30.84 | 20.22 | 42.51 | 37.27 |
| 2011 | 31.54 | 34.41 | 34.05 | 18.08 | 27.78 | 54.13 | 31.18 | 36.22 | 32.60 | 18.99 | 25.49 | 55.52 |
| 2015 | 31.08 | 35.80 | 33.12 | 17.30 | 29.81 | 52.90 | 31.17 | 36.58 | 32.25 | 17.91 | 28.17 | 53.93 |
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Western | Central | Eastern | Western | Central | Eastern | Western | Central | Eastern | Western | Central | Eastern | |
| 2010 | 32 | 28.4 | 40 | 19 | 30.91 | 49.89 | 33.5 | 29.24 | 37.26 | 17.15 | 30.57 | 52.29 |
| 2012 | 31.26 | 28.84 | 39.89 | 19.49 | 32.44 | 48.07 | 32.74 | 29.27 | 37.99 | 16.9 | 32.14 | 50.95 |
| 2014 | 31.35 | 28.79 | 39.86 | 19.29 | 32.58 | 48.13 | 32.84 | 29.98 | 37.18 | 18.39 | 31.41 | 50.2 |
| 2016 | 32.44 | 28.07 | 39.49 | 20.69 | 31.16 | 48.16 | 34.77 | 29.34 | 35.89 | 18.97 | 30.36 | 50.68 |
| 2018 | 34.90 | 26.64 | 38.46 | 22.45 | 29.56 | 47.99 | 36.68 | 27.97 | 35.35 | 20.06 | 30.43 | 49.51 |
| 2020 | 34.47 | 27.23 | 38.30 | 24.17 | 29.44 | 46.4 | 34.47 | 28.55 | 36.98 | 21.49 | 30.98 | 47.53 |
Table 5e.
Sample distribution (%) in different residency by SES and gender (1991–2020)
| Year | LU | LR | HU | HR | MLU | MLR | MHU | MHR | FLU | FLR | FHU | FHR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1991 | 27.46 | 72.54 | 47.02 | 52.98 | 25.54 | 74.46 | 38.76 | 61.24 | 26.28 | 73.72 | 45.51 | 54.49 |
| 1993 | 24.99 | 75.01 | 47.04 | 52.96 | 23.93 | 76.07 | 39.35 | 60.65 | 23.61 | 76.39 | 44.56 | 55.44 |
| 1997 | 25.36 | 74.64 | 45.35 | 54.65 | 23.88 | 76.12 | 40.35 | 59.65 | 25.25 | 74.75 | 44.95 | 55.05 |
| 2000 | 23.74 | 76.26 | 44.35 | 55.65 | 22.88 | 77.12 | 41.59 | 58.41 | 24.32 | 75.68 | 43.98 | 56.02 |
| 2004 | 25.11 | 74.89 | 44.91 | 55.09 | 24.91 | 75.09 | 42.81 | 57.19 | 25.61 | 74.39 | 50.19 | 49.81 |
| 2006 | 23.68 | 76.32 | 44.49 | 55.51 | 25.39 | 74.61 | 45.06 | 54.94 | 24.01 | 75.99 | 48.33 | 51.67 |
| 2009 | 23.50 | 76.50 | 44.62 | 55.38 | 24.21 | 75.79 | 46.84 | 53.16 | 24.98 | 75.02 | 50.40 | 49.60 |
| 2011 | 25.02 | 74.98 | 52.16 | 47.84 | 29.36 | 70.64 | 55.76 | 44.24 | 27.33 | 72.67 | 58.91 | 41.09 |
| 2015 | 22.58 | 77.42 | 46.31 | 53.69 | 28.05 | 71.95 | 50.97 | 49.03 | 26.28 | 73.72 | 56.91 | 43.09 |
| Year | MLU | MLR | MHU | MHR | FLU | FLR | FHU | FHR |
|---|---|---|---|---|---|---|---|---|
| 2010 | 36.90 | 63.10 | 66.48 | 33.52 | 33.60 | 66.40 | 67.26 | 32.74 |
| 2012 | 36.75 | 63.25 | 63.81 | 36.19 | 33.99 | 66.01 | 67.40 | 32.60 |
| 2014 | 39.11 | 60.89 | 64.79 | 35.21 | 36.02 | 63.98 | 65.41 | 34.59 |
| 2016 | 40.94 | 59.06 | 66.33 | 33.67 | 37.90 | 62.10 | 66.64 | 33.36 |
| 2018 | 41.75 | 58.25 | 65.75 | 34.25 | 37.80 | 62.20 | 67.17 | 32.83 |
| 2020 | 40.38 | 59.62 | 66.12 | 33.88 | 40.01 | 59.99 | 69.24 | 30.76 |
Note: MLU: low SES male participant in urban area. MLR: low SES male participant in rural area. MHU: high SES male participant in urban area. MHR: high SES male participant in rural area. FLU: low SES female participant in urban area. FLR: low SES female participant in rural area. FHU: high SES female participant in urban area. FHR: high SES female participant in rural area.
Table 5f.
Health insurance rate (%) by SES and gender (1991–2020)
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 1991 | 31.39 | 34.55 | 28.22 | 20.31 | 51.52 | 23.17 | 51.89 | 17.79 | 51.07 |
| 1993 | 26.02 | 29.39 | 22.62 | 15.90 | 45.73 | 19.02 | 48.52 | 12.87 | 42.69 |
| 1997 | 25.42 | 27.87 | 22.94 | 17.79 | 40.27 | 19.58 | 42.33 | 16.11 | 37.85 |
| 2000 | 20.86 | 22.43 | 19.20 | 13.13 | 35.05 | 13.94 | 37.95 | 12.29 | 31.98 |
| 2004 | 26.58 | 28.74 | 24.45 | 18.28 | 40.92 | 19.27 | 43.85 | 17.37 | 37.72 |
| 2006 | 49.83 | 51.78 | 47.95 | 45.00 | 58.39 | 46.54 | 61.42 | 43.46 | 55.61 |
| 2009 | 91.04 | 91.60 | 90.50 | 91.78 | 89.67 | 92.51 | 89.82 | 91.05 | 89.54 |
| 2011 | 94.92 | 95.14 | 94.71 | 94.63 | 95.36 | 94.89 | 95.54 | 94.37 | 95.20 |
| 2015 | 97.13 | 97.26 | 97.01 | 97.13 | 97.12 | 97.36 | 97.08 | 96.91 | 97.15 |
| Year | All | Male | Female | Low SES | High SES | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|---|---|
| 2010 | 81.72 | 82.81 | 80.69 | 82.36 | 80.46 | 83.26 | 81.67 | 81.38 | 79.63 |
| 2012 | 87.20 | 87.77 | 86.66 | 88.40 | 84.62 | 88.57 | 85.67 | 88.22 | 83.84 |
| 2014 | 91.07 | 91.41 | 90.75 | 91.82 | 89.86 | 91.77 | 90.67 | 91.87 | 89.26 |
| 2016 | 91.4 | 91.76 | 91.06 | 91.56 | 91.14 | 91.75 | 91.78 | 91.34 | 90.66 |
| 2018 | 91.93 | 92.10 | 91.77 | 91.59 | 92.44 | 91.71 | 92.81 | 91.44 | 92.15 |
| 2020 | 90.11 | 90.53 | 89.69 | 89.46 | 91.02 | 90.08 | 91.19 | 88.81 | 90.85 |
Note: All = all participants. Male = male participants. Female = female participants.
Table 5g.
Smoking prevalence (%) by SES and gender (1991–2020)
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 1991 | 34.63 | 68.03 | 4.25 | 69.09 | 66.44 | 4.53 | 3.66 |
| 1993 | 33.50 | 65.20 | 4.60 | 66.71 | 62.51 | 4.43 | 4.93 |
| 1997 | 31.77 | 59.35 | 4.13 | 62.26 | 54.31 | 4.53 | 3.24 |
| 2000 | 31.67 | 59.80 | 4.76 | 63.48 | 53.26 | 4.93 | 4.43 |
| 2004 | 33.59 | 63.41 | 4.19 | 65.72 | 59.72 | 3.99 | 4.56 |
| 2006 | 32.78 | 62.85 | 3.74 | 65.58 | 57.80 | 3.42 | 4.30 |
| 2009 | 32.08 | 61.77 | 3.79 | 64.50 | 56.37 | 3.69 | 3.98 |
| 2011 | 31.31 | 61.50 | 3.38 | 65.72 | 54.83 | 3.50 | 3.22 |
| 2015 | 26.95 | 54.37 | 2.28 | 57.55 | 48.76 | 2.24 | 2.34 |
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 2010 | 37.09 | 71.32 | 4.99 | 75.41 | 61.98 | 5.37 | 4.44 |
| 2012 | 34.71 | 66.70 | 4.24 | 70.04 | 57.99 | 4.55 | 3.67 |
| 2014 | 35.67 | 66.95 | 4.27 | 71.14 | 58.35 | 4.47 | 4.01 |
| 2016 | 35.05 | 65.38 | 4.08 | 69.54 | 56.65 | 4.24 | 3.87 |
| 2018 | 41.64 | 78.20 | 5.89 | 81.25 | 72.72 | 6.51 | 5.15 |
| 2020 | 35.74 | 67.91 | 3.42 | 71.73 | 62.28 | 3.92 | 2.76 |
Note: All = all participants. Male = male participants. Female = female participants.
Table 5h.
Drinking prevalence (%) by SES and gender (1991–2020)
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 1991 | 15.18 | 28.75 | 2.64 | 26.35 | 32.35 | 2.85 | 2.18 |
| 1993 | 16.35 | 31.12 | 2.68 | 30.16 | 32.84 | 2.84 | 2.35 |
| 1997 | 15.43 | 28.02 | 2.68 | 27.46 | 29.02 | 2.58 | 2.89 |
| 2000 | 16.31 | 30.12 | 3.02 | 30.12 | 30.12 | 3.16 | 2.76 |
| 2004 | 15.68 | 29.77 | 1.91 | 30.81 | 28.11 | 2.01 | 1.74 |
| 2006 | 15.08 | 28.63 | 2.12 | 29.96 | 26.16 | 2.29 | 1.83 |
| 2009 | 13.82 | 26.26 | 2.03 | 27.34 | 24.13 | 1.89 | 2.28 |
| 2011 | 14.47 | 27.84 | 2.15 | 30.14 | 24.20 | 2.17 | 2.12 |
| 2015 | 11.03 | 22.07 | 1.14 | 24.49 | 17.82 | 1.13 | 1.16 |
| Year | All | Male | Female | Low SES in men | High SES in men | Low SES in women | High SES in women |
|---|---|---|---|---|---|---|---|
| 2010 | 16.12 | 30.00 | 3.11 | 31.51 | 26.54 | 3.46 | 2.59 |
| 2012 | 15.80 | 29.36 | 2.89 | 30.88 | 25.41 | 3.10 | 2.50 |
| 2014 | 15.56 | 28.79 | 2.86 | 30.96 | 24.39 | 3.06 | 2.59 |
| 2016 | 14.69 | 26.98 | 2.80 | 28.85 | 23.13 | 2.90 | 2.66 |
| 2018 | 15.45 | 27.95 | 3.23 | 30.12 | 24.05 | 3.41 | 3.01 |
| 2020 | 12.87 | 23.37 | 2.33 | 25.56 | 20.15 | 2.56 | 2.02 |
Note: All = all participants. Male = male participants. Female = female participants.
Table 5i.
Per capita household income distribution (%) by SES and gender in CHNS (1991–2015)
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | Q5 | Q1 | Q2 | Q3 | Q4 | Q5 | Q1 | Q2 | Q3 | Q4 | Q5 | Q1 | Q2 | Q3 | Q4 | Q5 | |
| 1991 | 26.00 | 21.93 | 19.25 | 17.97 | 14.85 | 10.97 | 17.06 | 21.18 | 23.01 | 27.78 | 24.53 | 23.55 | 20.00 | 16.27 | 15.65 | 10.10 | 12.24 | 20.01 | 28.89 | 28.76 |
| 1993 | 23.86 | 21.82 | 19.84 | 17.53 | 16.96 | 12.95 | 16.70 | 20.26 | 24.62 | 25.48 | 24.59 | 22.63 | 19.61 | 17.26 | 15.91 | 10.63 | 14.64 | 20.79 | 25.67 | 28.28 |
| 1997 | 23.57 | 22.58 | 20.24 | 17.14 | 16.48 | 13.89 | 15.59 | 19.52 | 24.99 | 26.01 | 23.69 | 22.91 | 20.20 | 17.12 | 16.09 | 12.10 | 13.71 | 19.69 | 26.21 | 28.29 |
| 2000 | 23.82 | 23.25 | 21.08 | 17.86 | 14.00 | 13.05 | 14.04 | 18.23 | 23.73 | 30.95 | 24.78 | 22.97 | 20.77 | 17.07 | 14.42 | 11.34 | 14.64 | 18.63 | 25.41 | 29.98 |
| 2004 | 25.78 | 22.49 | 20.99 | 18.55 | 12.19 | 10.79 | 16.08 | 18.38 | 22.33 | 32.42 | 24.58 | 23.50 | 20.40 | 18.44 | 13.07 | 11.43 | 13.45 | 19.25 | 22.91 | 32.97 |
| 2006 | 25.34 | 23.67 | 21.37 | 17.28 | 12.34 | 10.26 | 13.26 | 17.53 | 24.96 | 34.00 | 25.46 | 23.26 | 21.08 | 17.91 | 12.29 | 10.92 | 14.31 | 18.15 | 23.54 | 33.08 |
| 2009 | 24.63 | 23.67 | 20.62 | 17.62 | 13.46 | 10.95 | 12.84 | 18.86 | 24.60 | 32.74 | 25.18 | 22.97 | 21.10 | 17.69 | 13.06 | 11.42 | 14.33 | 18.10 | 24.12 | 32.02 |
| 2011 | 26.16 | 24.56 | 20.70 | 16.38 | 12.19 | 10.36 | 12.79 | 18.91 | 25.70 | 32.25 | 26.37 | 25.06 | 21.69 | 15.91 | 10.98 | 11.11 | 12.91 | 17.64 | 25.72 | 32.62 |
| 2015 | 26.03 | 22.99 | 19.88 | 16.84 | 14.25 | 9.65 | 14.78 | 20.20 | 25.49 | 29.89 | 27.16 | 24.32 | 20.24 | 15.52 | 12.76 | 9.66 | 13.75 | 19.86 | 26.29 | 30.45 |
Appendix 6. Employment rate, education, and age
Table 6a.
Employment rate (%) in different age groups (1991–2020)
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | |
| 1991 | 94.31 | 98.40 | 97.38 | 79.56 | 39.76 | 82.45 | 99.68 | 98.18 | 83.46 | 19.19 | 91.05 | 93.58 | 88.48 | 57.59 | 26.38 | 82.19 | 94.59 | 89.80 | 44.74 | 9.65 |
| 1993 | 92.22 | 95.15 | 95.74 | 78.54 | 43.58 | 78.80 | 96.98 | 96.95 | 84.77 | 25.81 | 90.72 | 90.62 | 84.49 | 60.54 | 31.46 | 77.65 | 89.69 | 91.33 | 42.54 | 13.56 |
| 1997 | 90.95 | 94.17 | 93.94 | 82.10 | 42.29 | 71.75 | 96.17 | 94.47 | 80.30 | 23.29 | 89.39 | 86.99 | 80.98 | 63.95 | 29.64 | 65.37 | 88.09 | 84.21 | 51.34 | 14.97 |
| 2000 | 88.51 | 90.92 | 92.21 | 81.34 | 39.26 | 63.93 | 91.74 | 92.31 | 70.44 | 20.69 | 83.52 | 83.16 | 80.22 | 66.95 | 26.79 | 61.63 | 86.86 | 77.49 | 48.89 | 16.67 |
| 2004 | 81.98 | 87.62 | 82.42 | 77.44 | 35.30 | 64.57 | 90.60 | 85.89 | 68.00 | 24.54 | 64.18 | 72.55 | 69.16 | 53.11 | 19.43 | 68.12 | 77.82 | 64.85 | 37.55 | 11.60 |
| 2006 | 77.06 | 88.08 | 85.52 | 74.30 | 33.70 | 64.47 | 91.23 | 87.62 | 69.82 | 20.74 | 65.29 | 75.24 | 72.39 | 57.51 | 21.88 | 63.80 | 79.08 | 66.20 | 36.40 | 11.38 |
| 2009 | 80.11 | 90.79 | 86.43 | 74.39 | 36.90 | 66.45 | 89.19 | 87.96 | 73.72 | 23.99 | 57.38 | 73.94 | 75.34 | 52.64 | 17.75 | 67.49 | 82.82 | 68.99 | 33.78 | 19.80 |
| 2011 | 81.79 | 90.48 | 88.45 | 74.83 | 37.06 | 73.96 | 93.53 | 89.84 | 72.72 | 19.98 | 59.30 | 75.23 | 76.35 | 55.09 | 18.57 | 70.16 | 82.16 | 71.66 | 30.79 | 16.71 |
| 2015 | 69.54 | 76.94 | 74.21 | 60.23 | 28.24 | 65.69 | 89.76 | 90.48 | 66.56 | 18.43 | 47.78 | 62.12 | 51.88 | 34.01 | 13.09 | 61.48 | 83.44 | 65.87 | 31.02 | 9.98 |
| Year | Low SES in men |
High SES in men |
Low SES in women |
High SES in women |
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | 18–30 | 31–40 | 41–50 | 51–60 | 61+ | |
| 2010 | 81.46 | 86.34 | 84.56 | 85.80 | 81.49 | 60.59 | 95.12 | 93.90 | 93.86 | 91.40 | 73.55 | 75.51 | 71.21 | 71.67 | 65.85 | 62.75 | 91.71 | 83.64 | 82.99 | 79.49 |
| 2012 | 60.74 | 71.38 | 72.53 | 61.83 | 34.08 | 64.94 | 89.61 | 85.38 | 66.62 | 22.80 | 36.79 | 57.94 | 59.96 | 44.57 | 20.90 | 55.02 | 79.17 | 65.51 | 35.06 | 16.36 |
| 2014 | 84.40 | 94.68 | 93.90 | 85.48 | 56.98 | 69.17 | 96.94 | 95.94 | 84.53 | 41.06 | 58.64 | 80.74 | 85.37 | 72.24 | 41.26 | 54.70 | 85.76 | 84.08 | 54.02 | 30.64 |
| 2016 | 80.25 | 94.54 | 94.02 | 89.22 | 59.18 | 64.74 | 96.05 | 94.53 | 84.18 | 44.41 | 56.74 | 79.57 | 87.06 | 74.29 | 46.85 | 57.45 | 85.55 | 85.24 | 59.44 | 33.87 |
| 2018 | 81.50 | 95.23 | 94.75 | 89.62 | 61.42 | 66.76 | 96.70 | 96.30 | 82.43 | 49.20 | 55.12 | 78.41 | 85.41 | 76.52 | 51.08 | 60.96 | 83.57 | 86.95 | 60.63 | 36.57 |
| 2020 | 77.75 | 96.09 | 95.12 | 92.47 | 63.51 | 66.35 | 96.82 | 96.63 | 82.57 | 50.00 | 48.98 | 73.78 | 84.85 | 78.02 | 51.83 | 65.23 | 82.10 | 84.27 | 60.59 | 34.40 |
Table 6b.
Association between employment and NCDs (1991–2020)
| Year | All |
Male |
Female |
|||
|---|---|---|---|---|---|---|
| AOR | (95 %CI) | AOR | (95 %CI) | AOR | (95 %CI) | |
| 1991 | 0.49∗∗∗ | (0.35–0.69) | 0.44∗∗∗ | (0.24–0.82) | 0.49∗∗∗ | (0.33–0.71) |
| 1993 | 0.63∗∗∗ | (0.46–0.88) | 0.65∗ | (0.42–1.01) | 0.66∗ | (0.43–1.02) |
| 1997 | 0.51∗∗ | (0.29–0.87) | 0.49∗ | (0.23–1.01) | 0.49∗∗∗ | (0.28–0.84) |
| 2000 | 0.60∗∗∗ | (0.46–0.77) | 0.57∗∗∗ | (0.37–0.87) | 0.61∗∗∗ | (0.45–0.82) |
| 2004 | 0.68∗∗∗ | (0.55–0.83) | 0.60∗∗∗ | (0.45–0.79) | 0.74∗∗ | (0.57–0.96) |
| 2006 | 0.62∗∗∗ | (0.50–0.76) | 0.61∗∗∗ | (0.45–0.84) | 0.61∗∗∗ | (0.47–0.78) |
| 2009 | 0.63∗∗∗ | (0.53–0.74) | 0.54∗∗∗ | (0.42–0.68) | 0.72∗∗∗ | (0.58–0.91) |
| 2011 | 0.65∗∗∗ | (0.57–0.74) | 0.57∗∗∗ | (0.47–0.68) | 0.72∗∗∗ | (0.61–0.86) |
| 2015 | 0.70∗∗∗ | (0.62–0.80) | 0.63∗∗∗ | (0.53–0.75) | 0.77∗∗∗ | (0.64–0.94) |
| Year | All |
Male |
Female |
|||
|---|---|---|---|---|---|---|
| AOR | (95 %CI) | AOR | (95 %CI) | AOR | (95 %CI) | |
| 2010 | 1.14 | (0.97–1.34) | 1.19 | (0.95–1.49) | 1.12 | (0.95–1.33) |
| 2012 | 0.86∗∗∗ | (0.77–0.95) | 0.85∗∗ | (0.75–0.98) | 0.85∗∗ | (0.75–0.96) |
| 2014 | 0.70∗∗∗ | (0.63–0.79) | 0.66∗∗∗ | (0.57–0.77) | 0.73∗∗∗ | (0.63–0.83) |
| 2016 | 0.69∗∗∗ | (0.62–0.77) | 0.59∗∗∗ | (0.51–0.69) | 0.76∗∗∗ | (0.66–0.87) |
| 2018 | 0.71∗∗∗ | (0.65–0.77) | 0.64∗∗∗ | (0.56–0.74) | 0.74∗∗∗ | (0.66–0.83) |
| 2020 | 0.68∗∗∗ | (0.62–0.76) | 0.59∗∗∗ | (0.50–0.69) | 0.75∗∗∗ | (0.66–0.87) |
∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.
Table 6c.
Mean educated year in different age groups (1991–2020)
| Year | 18–30 |
31–40 |
41–50 |
51–60 |
61+ |
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| 1989 | 7.93 | 3.23 | 6.05 | 3.87 | 4.74 | 3.94 | 2.69 | 3.86 | 1.70 | 3.22 |
| 1991 | 8.19 | 3.05 | 6.81 | 3.96 | 5.23 | 3.83 | 3.28 | 4.15 | 2.40 | 3.72 |
| 1993 | 8.29 | 2.89 | 7.41 | 3.91 | 5.51 | 3.83 | 3.96 | 4.26 | 2.28 | 3.69 |
| 1997 | 8.55 | 2.86 | 8.27 | 3.44 | 5.93 | 3.91 | 4.74 | 4.08 | 2.57 | 3.76 |
| 2000 | 9.21 | 2.82 | 8.64 | 3.07 | 6.91 | 3.99 | 5.50 | 3.85 | 3.21 | 4.19 |
| 2004 | 9.66 | 2.76 | 8.62 | 3.00 | 8.24 | 3.67 | 6.35 | 3.78 | 4.46 | 4.49 |
| 2006 | 10.30 | 2.92 | 8.67 | 3.28 | 8.69 | 3.66 | 6.24 | 4.25 | 4.41 | 4.67 |
| 2009 | 10.36 | 2.89 | 8.76 | 3.15 | 8.64 | 3.38 | 6.83 | 4.19 | 4.74 | 4.43 |
| 2011 | 11.46 | 3.10 | 9.98 | 3.56 | 9.17 | 3.44 | 7.59 | 4.01 | 5.56 | 4.56 |
| 2015 | 11.82 | 3.12 | 10.76 | 3.41 | 9.25 | 3.59 | 8.84 | 3.70 | 6.46 | 4.44 |
| Year | 18–30 |
31–40 |
41–50 |
51–60 |
61–70 |
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| 2010 | 10.04 | 3.87 | 7.95 | 4.45 | 7.28 | 4.26 | 5.73 | 4.73 | 3.94 | 4.49 |
| 2012 | 9.67 | 3.97 | 7.76 | 4.71 | 6.47 | 4.57 | 5.48 | 4.99 | 3.40 | 4.40 |
| 2014 | 10.71 | 3.63 | 8.83 | 4.27 | 7.11 | 4.17 | 6.76 | 4.60 | 4.28 | 4.37 |
| 2016 | 11.44 | 3.62 | 9.46 | 4.19 | 7.15 | 4.22 | 7.11 | 4.43 | 4.55 | 4.45 |
| 2018 | 11.87 | 3.55 | 10.16 | 4.11 | 7.42 | 4.30 | 7.11 | 4.32 | 4.84 | 4.51 |
| 2020 | 12.55 | 3.22 | 10.92 | 3.92 | 8.23 | 4.28 | 7.23 | 4.25 | 5.47 | 4.66 |
Note: SD = standard deviation.
Data availability
Data will be made available on request.
References
- Corsi D.J., Subramanian S.V. Association between socioeconomic status and self-reported diabetes in India: A cross-sectional multilevel analysis. BMJ Open. 2012;2(4) doi: 10.1136/bmjopen-2012-000895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Corsi D.J., Subramanian S.V. Socioeconomic gradients and distribution of diabetes, hypertension, and obesity in India. JAMA Network Open. 2019;2(4) doi: 10.1001/jamanetworkopen.2019.0411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ding S., Chen J., Dong B., Hu J. Association between parental socioeconomic status and offspring overweight/obesity from the China Family Panel Studies: A longitudinal survey. BMJ Open. 2021;11(4) doi: 10.1136/bmjopen-2020-045433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Emamian M.H., Fateh M., Hosseinpoor A.R., Alami A., Fotouhi A. Obesity and its socioeconomic determinants in Iran. Economics and Human Biology. 2017;26:144–150. doi: 10.1016/j.ehb.2017.03.005. [DOI] [PubMed] [Google Scholar]
- Fang C., Liang Y. Social disparities in body mass index (BMI) trajectories among Chinese adults in 1991-2011. International Journal for Equity in Health. 2017;16(1):146. doi: 10.1186/s12939-017-0636-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goryakin Y., Suhrcke M. Economic development, urbanization, technological change and overweight: What do we learn from 244 Demographic and Health Surveys? Economics and Human Biology. 2014;14:109–127. doi: 10.1016/j.ehb.2013.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Han L.L., Wang Y.X., Li J., et al. Gender differences in associations of serum ferritin and diabetes, metabolic syndrome, and obesity in the China Health and Nutrition Survey. Molecular Nutrition & Food Research. 2014;58(11):2189–2195. doi: 10.1002/mnfr.201400088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Institute of Social Science Survey, Peking University China family panel studies. Available: https://www.isss.pku.edu.cn/cfps/download/index#/literature.
- Institute of Social Science Survey, Peking University China family panel studies. Available: https://www.isss.pku.edu.cn/cfps/en/documentation/user/index.htm?CSRFT=UTOZ-LYR6-I1IW-PNRM-W3BF-PGD3-HY2W-SZEO.
- Jiwani S.S., Carrillo-Larco R.M., Hernández-Vásquez A., et al. The shift of obesity burden by socioeconomic status between 1998 and 2017 in Latin America and the Caribbean: A cross-sectional series study. Lancet Global Health. 2019;7(12):e1644–e1654. doi: 10.1016/S2214-109X(19)30421-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones-Smith J.C., Gordon-Larsen P., Siddiqi A., Popkin B.M. Emerging disparities in overweight by educational attainment in Chinese adults (1989–2006) 2012;36(6):866–875. doi: 10.1038/ijo.2011.134. JIjoo. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jung L., De Neve J.W., Chen S., et al. The interaction between district-level development and individual-level socioeconomic gradients of cardiovascular disease risk factors in India: A cross-sectional study of 2.4 million adults. Social Science & Medicine. 2019;239 doi: 10.1016/j.socscimed.2019.112514. 1982. [DOI] [PubMed] [Google Scholar]
- Kinge J.M., Strand B.H., Vollset S.E., Skirbekk V. Educational inequalities in obesity and gross domestic product: Evidence from 70 countries. Journal of Epidemiology & Community Health. 2015;69(12):1141–1146. doi: 10.1136/jech-2014-205353. [DOI] [PubMed] [Google Scholar]
- Li L., Guo Y., Chen Z., Chen J., Peto R. Epidemiology and the control of disease in China, with emphasis on the Chinese Biobank Study. Public Health. 2012;126(3):210–213. doi: 10.1016/j.puhe.2011.11.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McPake B. Socio-economic gradients of health and health behaviors: From non-communicable diseases to breast feeding and back. Social Science & Medicine. 1982;2024 doi: 10.1016/j.socscimed.2023.116535. [DOI] [PubMed] [Google Scholar]
- Miao J., Wu X. Urbanization, socioeconomic status and health disparity in China. Health & Place. 2016;42:87–95. doi: 10.1016/j.healthplace.2016.09.008. [DOI] [PubMed] [Google Scholar]
- Monteiro C.A., Conde W.L., Popkin B.M. Independent effects of income and education on the risk of obesity in the Brazilian adult population. The Journal of Nutrition. 2001;131(3) doi: 10.1093/jn/131.3.881S. 881s-6s. [DOI] [PubMed] [Google Scholar]
- National Bureau of Statistics of China Annual economic data-GDP growth (percentage) https://data.stats.gov.cn/easyquery.htm?cn=C01 Available:
- Nikparvar M., Khaladeh M., Yousefi H., Vahidi Farashah M., Moayedi B., Kheirandish M. Dyslipidemia and its associated factors in southern Iranian women, Bandare-Kong Cohort study, a cross-sectional survey. Scientific Reports. 2021;11(1):9125. doi: 10.1038/s41598-021-88680-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ogunsina K., Dibaba D.T., Akinyemiju T. Association between life-course socio-economic status and prevalence of cardio-metabolic risk reactors in five middle-income countries. Journal of Global Health. 2018;8(2) doi: 10.7189/jogh.08.020405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Okamoto S., Kobayashi E., Murayama H., Liang J., Fukaya T., Shinkai S. Decomposition of gender differences in cognitive functioning: National Survey of the Japanese elderly. BMC Geriatrics. 2021;21(1):38. doi: 10.1186/s12877-020-01990-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pampel F.C., Denney J.T., Krueger P.M. Obesity, SES, and economic development: A test of the reversal hypothesis. Social Science & Medicine. 2012;74(7):1073–1081. doi: 10.1016/j.socscimed.2011.12.028. 1982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pan T., Palmer M. Risk factors and non-communicable disease diagnosis in China. China Economic Review. 2018;50:72–84. [Google Scholar]
- Pang Y., Han Y., Yu C., et al. The role of lifestyle factors on comorbidity of chronic liver disease and cardiometabolic disease in Chinese population: A prospective cohort study. 2022;28 doi: 10.1016/j.lanwpc.2022.100564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pérez-Ferrer C., McMunn A., Zaninotto P., Brunner E.J. The nutrition transition in Mexico 1988-2016: The role of wealth in the social patterning of obesity by education. Public Health Nutrition. 2018;21(13):2394–2401. doi: 10.1017/S1368980018001167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Porhcisaliyan V.D., Wang Y., Tan N.C., Jafar T.H. Socioeconomic status and ethnic variation associated with type 2 diabetes mellitus in patients with uncontrolled hypertension in Singapore. BMJ Open Diabetes Research & Care. 2021;9(1) doi: 10.1136/bmjdrc-2020-002064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rahimi E., Hashemi Nazari S.S. A detailed explanation and graphical representation of the Blinder-Oaxaca decomposition method with its application in health inequalities. Emerging Themes in Epidemiology. 2021;18(1):12. doi: 10.1186/s12982-021-00100-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sánchez-Vaznaugh E.V., Kawachi I., Subramanian S.V., Sánchez B.N., Acevedo-Garcia D. Do socioeconomic gradients in body mass index vary by race/ethnicity, gender, and birthplace? American Journal of Epidemiology. 2009;169(9):1102–1112. doi: 10.1093/aje/kwp027. [DOI] [PubMed] [Google Scholar]
- Siddiquee T., Bhowmik B., Da Vale Moreira N.C., et al. Prevalence of obesity in a rural Asian Indian (Bangladeshi) population and its determinants. BMC Public Health. 2015;15:860. doi: 10.1186/s12889-015-2193-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Subramanian S.V., Corsi D.J., Subramanyam M.A., Smith G.D. Jumping the gun: The problematic discourse on socioeconomic status and cardiovascular health in India. International Journal of Epidemiology. 2013;42(5):1410–1426. doi: 10.1093/ije/dyt017. [DOI] [PubMed] [Google Scholar]
- Subramanian S., Perkins J.M., Özaltin E., Smith Gjtajocn D. Weight of nations: A socioeconomic analysis of women in low-to middle-income countries. 2011;93(2):413–421. doi: 10.3945/ajcn.110.004820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- The university of North Carolina at Chapel Hill and the Chinese Center for disease control and prevention. China Health and Nutrition Survey. Available:https://chns.cpc.unc.edu/about/design/sample/. Accessed September 8 2025].
- The university of North Carolina at Chapel Hill and the Chinese Center for disease control and prevention. China Health and Nutrition Survey. Available:https://www.cpc.unc.edu/projects/china/data/documentation. Accessed March 11 2022].
- Veisani Y., Jenabi E., Nematollahi S., Delpisheh A., Khazaei S. The role of socio-economic inequality in the prevalence of hypertension in adults. Journal of Cardiovascular and Thoracic Research. 2019;11(2):116–120. doi: 10.15171/jcvtr.2019.20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Health Organization Noncommunicable diseases. https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases Available:
- Wu H., Bragg F., Yang L., et al. Sex differences in the association between socioeconomic status and diabetes prevalence and incidence in China: Cross-sectional and prospective studies of 0.5 million adults. Diabetologia. 2019;62(8):1420–1429. doi: 10.1007/s00125-019-4896-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang Y., Kelifa M.O., Yu B., Herbert C., Wang Y., Jiang J. Gender-specific temporal trends in overweight prevalence among Chinese adults: A hierarchical age-period-cohort analysis from 2008 to 2015. Global Health Research and Policy. 2020;5:42. doi: 10.1186/s41256-020-00169-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang X., Pan T., McPake B. Re-examining the reversal hypothesis: A nationwide population-based study of the association between socioeconomic status, and NCDs and risk factors in China. SSM - Population Health. 2023;21 doi: 10.1016/j.ssmph.2022.101335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhao P., Gu X., Qian D., Yang F. Socioeconomic disparities in abdominal obesity over the life course in China. International Journal for Equity in Health. 2018;17(1):96. doi: 10.1186/s12939-018-0809-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhou M. The shifting income-obesity relationship: Conditioning effects from economic development and globalization. SSM - Population Health. 2021;15 doi: 10.1016/j.ssmph.2021.100849. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be made available on request.