Article Text
PDF
Abstract
Background Evidence of socioeconomic status (SES)-related health inequality is scarce in patients with cardiovascular diseases (CVDs) who need both lifestyle change and medical care, particularly in developing countries.
Methods The study employed a nationwide population-based cohort design, covering all 31 provinces of Chinese mainland from September 2014 to March 2021. Participants aged 35–75 years with self-reported CVD diagnoses were included. Information on SES and lifestyle details were collected via a questionnaire, and the unequal mortality across SES groups and the mediating effects of lifestyles were explored.
Results Among the 104 718 participants included, 27 943 (26.7%) were allocated to high SES, 35 802 (34.2%) were allocated to medium SES and 40 973 (39.1%) were allocated to low SES. During a mean follow-up of 48.9±18.9 months, 5010 deaths were recorded. Participants with low SES had a 65% (HR=1.65, 95% CI: 1.50 to 1.80) greater risk of all-cause death and a 95% (HR=1.95, 95% CI: 1.72 to 2.20) greater risk of CVD death in Chinese communities. A low SES with the worst lifestyle was associated with a significant increase in the risk of all-cause mortality by 172% (HR=2.72, 95% CI: 2.37 to 3.12) and CVD mortality by 218% (HR=3.18, 95% CI: 2.64 to 3.83) compared with a high SES with healthy lifestyle. The joint mediating effects of lifestyles on CVD mortality accounted for 19.6% (95%CI: 14.8% to 24.2%) of the excess mortality risk for individuals with low SES, and these effects varied by genders (p for interaction=0.013) and urbanity (p for interaction=0.004). Leisure-time physical activity was the strongest mediator, followed by dietary factors. For all-cause mortality, outcomes were similar to this.
Conclusions Both SES-related health inequalities and lifestyle disparities should be comprehensively considered when caring for this population, and upstream structural interventions that integrate SES and lifestyle factors and are tailored to the target population are urgently needed.
Trial registration number NCT02536456.
- Cardiovascular Diseases
- Cohort Studies
- Risk Factors
- Health Services
Data availability statement
Data are available on reasonable request. The data are not publicly available. The China Health Evaluation And risk Reduction through nationwide Teamwork only provides conditional data access for qualified researchers with legitimate requests; a formal application and research proposal is required. Please contact cvd-project@nccd.org.cn to seek approval for data access.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Prior studies reported that substantial socioeconomic status (SES)-related health inequality exists in lower SES populations, with lifestyle factors as mediators. The evidence is scarce for patients with cardiovascular disease (CVD) who need both lifestyle change and medical care, particularly in developing countries, where these situations could be more complicated.
WHAT THIS STUDY ADDS
Based on a large cohort of patients with CVD in Chinese communities, a low SES with the worst lifestyle was associated with a significant increase in the risk of all-cause mortality by 172% and CVD mortality by 218%, compared with a high SES with healthy lifestyle. The mediating effects of lifestyles accounted for nearly one-fifth of the excessive mortality risk for low SES in terms of all-cause mortality and CVD mortality, varying from genders and urban–rural status. Leisure-time physical activity was the strongest mediator, followed by dietary factors.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These findings have important implications for health policy-making. The outcome of healthcare across SES groups and the subsequent lifestyles in patients with established CVD has led to substantial gaps, and thus should be put on the agenda. A more supportive environment is needed to enable the low SES population to adopt healthier lifestyles.
Introduction
Socioeconomic status (SES), usually defined based on education, income and occupation, is an important determinant of morbidity and mortality.1 2 For example, for cardiovascular diseases (CVD), which affect more than 500 million people worldwide,3 substantial SES-related inequality in CVD burden exists—people with lower SES are at greater risk of suffering from and dying of CVD.4–6 In the causal paths between SES and health outcomes, lifestyle factors are recognised as mediators and more important intervention targets since they are largely modifiable.1 7 These studies were mainly conducted in the general population, and the evidence is scarce for patients with CVD, who may be more vulnerable to SES and need both lifestyle changes and medical care.
In addition, such evidence is scarce in developing countries such as China, where these situations could be more complicated. In recent decades, developing countries have experienced rapid economic growth and urbanisation, resulting in substantial changes in the SES structure of their population. Moreover, lifestyles across SES groups have also changed dramatically. For example, in the 1993 China Health and Nutrition Survey, unhealthy lifestyles were more prominent in the high SES group, but recent studies revealed clustering of lifestyle risk factors in the low SES population.8 A comprehensive understanding of these associations and mediations is essential for developing targeted health promotion strategies that may also alleviate health disparities due to SES.6 7 9
Therefore, we conducted this analysis based on data drawn from a large-scale nationwide population cohort in China. This study aimed to quantitatively assess SES-related health disparities and measure the mediating effects of lifestyles on all-cause mortality and CVD mortality among patients with established CVD in Chinese communities.
Methods
Study design and population
The China Health Evaluation And risk Reduction through nationwide Teamwork is a government-funded public health programme that focuses on CVD risk screening and intervention in the community-based population of China. In this project, from September 2014 to March 2021, local residents aged 35–75 years were recruited. The design of this project (formerly named China Patient-centred Evaluative Assessment of Cardiac Events Million Persons Project) has been detailed previously.7 The details are also described in the online supplemental methods S1. In this study, participants with a self-reported history of coronary heart disease (ie, myocardial infarction, percutaneous coronary intervention or coronary artery bypass grafting) or stroke were included.
Supplemental material
Those with missing gender, age, income, education, occupation or lifestyle information were excluded.
Data collection and definitions
Standardised in-person interviews were conducted at baseline by trained personnel using electronic questionnaires with a logical-check function. Information on sociodemographic status, type of transportation, lifestyles, medical history and medication use (see online supplemental methods S2) was collected. Body weight, blood pressure, blood lipids and glucose were measured using a standardised protocol and unique devices.
The dimensions of lifestyles included smoking, alcohol intake, leisure-time physical activity (LTPA), housework, diet and body weight. The definition of lifestyles is described in the online supplemental methods S3.
Outcome ascertainment
We obtained mortality data from the National Mortality Surveillance System and Vital Registration of the Chinese Center for Disease Control and Prevention. The death records in this system are reported by healthcare institutions almost in real time and then checked against local residential records and health insurance records annually, and the analyses used mortality data available up to 31 December 2021. All events were coded using the International Classification of Diseases, Tenth Revision (ICD-10). The outcomes of interest were all-cause mortality and CVD mortality (I00–I99).10
Statistical analysis
We calculated frequencies and percentages for categorical variables and means and SD or medians for continuous variables.
We aimed to categorise each participant into one of the three SES groups (ie, high, medium or low SES) through latent class analysis (LCA) based on the previous literature.7 11 12 Self-reported education level, annual family income level and occupation were selected to measure SES.7 The details are shown in online supplemental methods S4, tables S1 and S2.
A multilevel Cox proportional hazard regression model with province as a random effect was used to estimate the effect of inequality in SES on mortality, with high SES as a reference. The proportional hazard assumption was tested by the Schoenfeld residuals test (online supplemental table S3). In model 1, sociodemographic variable and patient’s CVD characteristic variables were adjusted, including gender, age, urbanity, marriage, health insurance, type of CVD disease, duration of CVD and type of transportation. In model 2, we additionally adjusted for lifestyle factors, history of hypertension, dyslipidaemia, diabetes and medications. In addition to the analysis of the whole CVD population, subgroup analysis was also conducted on gender, age, region and urban and rural areas, and the heterogeneity among subgroups was tested using p values for interaction.
To estimate the effect of lifestyles on mortality among the overall CVD population and different SES (low SES, medium SES, high SES) subgroups, a multilevel Cox proportional hazard regression model was used to calculate HRs. First, the number of risk lifestyle factors classified into three groups, 0–1, 2, and ≥3 (mainly according to the number of participants at each score), was included to estimate their cumulative effect on mortality. Second, the binary categories of the six lifestyle factors were included in another model to estimate the effect of each lifestyle factor on mortality. Furthermore, we explored the inclusion of interaction terms in our population-wide model to test whether there is a trend linearity in the effects of these lifestyles across different SES levels, ranging from high to low SES.
To examine how lifestyles mediate the relationship between SES and mortality among the CVD population, we performed a mediation analysis. All variables except lifestyles in the above model 2 were adjusted as confounders in the mediation analysis; lifestyles were selected and regarded as mediating variables. In brief, a mediator is defined as a variable that is on the causal pathway between the predictor and outcome of interest (online supplemental figure S1). The mediation effect size of lifestyle on the relationship between SES and mortality was reported as the mediation proportion (per cent, the total indirect effect divided by the total effect) for both low SES and medium SES populations, with high SES as a reference. Subgroup mediation analysis was also performed for male and female populations, as well as for urban and rural areas. The method was implemented using the ‘mma’ package in the statistics software R (V.4.2.1) and explained in detail elsewhere.13 Additionally, we calculated the joint mediation effect of these lifestyles using the ‘mma’ package, while the total effect was obtained by simply summing the absolute values of the mediation effects of statistically significant factors, without considering the direction of the mediation effects or those mediation effects without statistical significance. All the other analyses were performed using SAS V.9.4. We considered two-sided p values<0.05 to be significant.
Results
Of the 117 079 participants with established CVD, 104 718 participants with complete data were ultimately included in the study (online supplemental figure S2). The participants had a mean age of 61.8±8.0 years, and 49.1% were women. Based on the result of LCA, among them, 27 943 (26.7%) were allocated to high SES, 35 802 (34.2%) were allocated to medium SES and 40 973 (39.1%) were allocated to low SES. The characteristics of low SES were more likely to be female, older, living in western China, living in rural areas, working as a farmer, less educated and having lower income (table 1).
Baseline characteristics of the included participants
Overall, there were 28.5%, 35.6% and 35.9% of the participants having 0–1, 2 and ≥3 lifestyle risk factors, respectively. Unhealthy diet (65.0%) and low LTPA (52.6%) were the two most prevalent lifestyle risk factors. Lifestyle risk factors were more prevalent among participants with lower SES. For participants with high, medium and low SES, the proportions of participants with ≥3 numbers of unhealthy lifestyles were 26.4%, 34.8% and 43.3%, respectively (p for trend<0.001), and the proportions of participants with only 0 to 1 unhealthy lifestyle decreased from 40.8%, 30.1% to 18.8%, respectively (p for trend<0.001) (table 1).
During a mean follow-up of 48.9±18.9 months, 5010 deaths were recorded in this study cohort. There was a 65% greater risk of all-cause mortality (adjusted HR=1.65, 95% CI: 1.50, 1.80) and a 95% greater risk of CVD mortality (adjusted HR=1.95, 95% CI: 1.72, 2.20) for participants with low SES than for their counterparts with a high SES. Subgroup analyses across gender, age, region and urbanity subgroups demonstrated that the impact of SES on all-cause and CVD mortality showed no statistical significance (all interaction p values>0.05, table 2). The crude effects of low SES in model 1 are displayed in online supplemental table S4, while comparisons of medium SES with high SES are presented in online supplemental tables S5 and S6.
The effect of low SES on all-cause mortality and CVD mortality compared with high SES group (model 2)
Among the six lifestyle risk factors, for the overall patients, low LTPA was the strongest, with an adjusted HR of 1.38 (95% CI: 1.30, 1.47), followed by insufficient housework (HR=1.35, 95% CI: 1.27, 1.43) and unhealthy diet (HR=1.11, 95% CI: 1.04, 1.18), the association of housework time quartiles and intensity of LTPA with all-cause mortality and CVD mortality is shown in online supplemental table S7. Unhealthy lifestyles were indiscriminately associated with significantly higher mortality across SES groups, and the results are shown in figure 1. The HR of low LTPA was even larger in the high SES group (p for interaction=0.029). Similar situations have also been observed for CVD mortality (p for interaction=0.054, figure 1). The rates of all-cause mortality and CVD mortality in the overall population and across SES and lifestyles are shown in online supplemental table S8.
The effect of lifestyle factors on all-cause mortality (a) and CVD mortality (b) among participants in different SES groups with established CVD. BMI, body mass index; CVD, cardiovascular disease; LTPA, leisure-time physical activity; SES, socioeconomic status.
SES and unhealthy lifestyles both independently increased the risk of all-cause mortality and CVD mortality. The combination of a low SES with the worst lifestyle (≥3 unhealthy lifestyles) was associated with a significant increase in the risk of all-cause mortality by 172% (HR=2.72, 95% CI: 2.37, 3.12) and CVD mortality by 218% (HR=3.18, 95% CI: 2.64, 3.83) compared with a high SES with healthy lifestyle (0–1 number of unhealthy lifestyles) (figure 2). The association between SES and the risk of all-cause mortality and CVD mortality becomes progressively stronger as the number of unhealthy lifestyles increases, and when lifestyles are maintained at a satisfactory level, low SES continues to be a significant risk factor (figure 2). The effect of risk lifestyle numbers (with the best lifestyle as reference) on all-cause mortality and CVD mortality among participants in different SES groups is shown in online supplemental table S9.
The combined effect of risk lifestyles and SES on all-cause mortality (a) and CVD mortality (b) among participants with established CVD. Note: Adjusted variables: gender, age, urbanity, marriage, health insurance, type of CVD disease (coronary heart disease, stroke or both), duration of CVD, type of transportation, history of hypertension, dyslipidaemia and diabetes, and medications. CVD, cardiovascular disease; SES, socioeconomic status.
Lifestyle factors indirectly influence the relationship between SES and all-cause death among patients with CVD (table 3 and online supplemental table S10). In the overall cohort, the total mediating effects (indirect effects) accounted for 25.7% in low SES and 19.8% in medium SES to the extra risk of CVD mortality, respectively, and the joint mediated effect proportion was 19.6% (95% CI: 14.8%, 24.2%) and 19.1% (95% CI: 13.9%, 24.3%). The size of the joint mediating effect proportion in the male population was much greater than that in the female population (25.6% vs 18.4% for CVD mortality in low SES, p for interaction=0.013). Also, the size of the mediating effect in the urban population was much greater than that in the rural population (28.9% vs 9.3% for CVD mortality in low SES, p for interaction=0.004) (table 3). The result for all-cause mortality was similar to those for CVD mortality (online supplemental table S10).
The mediated effect proportion of lifestyle factors to excess CVD mortality risk among those with low SES and medium SES compared with those with high SES
Among the lifestyle risk factors, LTPA, the strongest mediator, was found to mediate 17.8% (95% CI:14.0, 21.6) in low SES and 14.6% (95%CI: 11.2%, 18.0%) in medium SES to the excessive CVD mortality risk compared with those of high SES, followed by unhealthy diet (4.8% in low SES and 5.2% in medium SES). Insufficient housework (−2.1%) and smoking (−1%) showed a negative mediation proportion, which means the two risk factors were more prevalent in the population with high SES, and the distribution of this factor across the population and its effect paradoxically results in more CVD mortality among those with high SES. No significant mediating effects were observed for risk drinking (p=0.402) and low or high BMI (p=0.885) (table 3). The mediated effect proportion (%) of lifestyle factors to excess all-cause mortality risk among those with low SES and medium SES compared with those with high SES is shown in online supplemental table S10.
Discussion
Based on a large cohort of patients with CVD in China, this study revealed that individuals with low SES were at 65% excess risk of all-cause death and 95% excess risk of CVD death in Chinese communities, with the associations varying by gender, age, region and urbanity. The combination of SES and lifestyle could effectively stratify the risk of death in patients with CVD in communities, and a low SES with the worst lifestyle was associated with a significant increase in the risk of all-cause mortality by 172% and CVD mortality by 218% compared with a high SES with healthy lifestyle. Unhealthy lifestyles mediated approximately one-fifth of SES-related inequity in mortality, but substantial heterogeneity in the proportions existed across subgroups. Low LTPA was consistently the most important lifestyle mediator.
SES-related health inequity is a hotspot issue for research on both health promotion and system performance.2 6 9 Analysis of patients with CVD who needed both lifestyle change and medical care provided a unique scenario to generate new knowledge on this topic in the following aspects. First, even in such a high-risk population, unhealthy lifestyles have mediated a notable proportion of SES-related inequity in mortality. Previous studies that were mostly conducted in the general population found that 20%–30% of the associations between SES and health outcomes (ie, CVD incidence and mortality) could be explained by lifestyle.7 14 15 The present study confirmed the substantial mediating effects of lifestyle in a large-scale cohort of participants with coronary heart disease or stroke.
Second, the mediating effects of unhealthy lifestyles on mortality varied substantially across genders and urbanity. Prior studies on health disparities used to pay particular attention to vulnerable groups, such as women, or underdeveloped regions, such as rural areas.16 17 Nevertheless, regarding the mediating effects of lifestyle, our analysis identified their counterparts (ie, male and urban residents) as noteworthy groups. There are two major explanations, first, in male and urban residents, the SES-related inequity in mortality was relatively small because it already represents a privileged section in the SES spectrum of the entire population. Second, in populations with richer resources and better access, differences in SES may lead to larger differences in lifestyles. For instance, rural residents who have poor environments and limited facilities for physical exercise or women who usually have less leisure time than men due to having been occupied by household duties, high purchasing power or health literacy have little to do with the habit of LTPA.18–21
Third, low LTPA, which had strong associations with both SES and mortality, acted as the most important mediator among lifestyle risk factors, followed by diet. Previous studies in the USA and the Netherlands have suggested that physical activity can significantly improve the body’s functions and overall condition, particularly by enhancing cardiorespiratory fitness, thereby reducing the incidence and mortality of CVD significantly.21 22 Meanwhile, physical activity mediates the association of SES with metabolic risk factors and cardiorespiratory fitness for the incidence of CVD, and we should emphasise and focus on reducing SES-related inequities by increasing and promoting regular physical activity.21 The scarcity of household chores contributed more to health losses among high SES individuals, although to a modest extent, yet this is a noteworthy direction for future research. For diet, the previously assessed relationships remained largely consistent—low-income groups were more likely to have difficulty affording healthy food,1 while those with less education tended to have limited awareness of healthy diet and nutrition knowledge.23 24
This new knowledge has noteworthy implications for health policy-making. First, lifestyle disparities across SES groups have led to substantial gaps in population health and thus should be put on the agenda. China has made great achievements in care access and equity in recent decades through establishing universal social medical insurance and strengthening the clinical capacity of hospitals and primary care institutions.25 However, healthy lifestyle facilitating policies, particularly for the low SES population, are needed in the new era plan. Second, mitigating SES-related health inequity requires comprehensive actions. In addition to the long-term strategy through improving education and income in the low SES population, as China did to promote compulsory education and eradicate extreme poverty, there is a short-term but probably more feasible strategy by modifying lifestyle risk factors, which needs a focus on male and urban populations with potentially larger disparities. Third, LTPA and diet should be highlighted as intervention targets for mitigating SES-related inequity in mortality. Policy-makers could consider fostering a more friendly environment for the low SES population to follow healthy lifestyles, in which low-income individuals can get access to physical exercise facilities at low cost in their near neighbourhood, and low-education individuals can easily understand how different types of food can affect their health. A collaborative alliance involving the government, schools, hospitals and social capital is required to form a concerted effort, grounded in the community environment, emphasising education as a comprehensive means and making sustained efforts with a focus on promoting physical exercise and healthy eating habits.6 9 21
This study should be interpreted in the context of several potential limitations. First, information on socioeconomics, CVD history and lifestyle was collected based on self-reports. Second, potential changes in lifestyles during the follow-up period were not captured, which may influence the associations. Third, the SES information of these patients may be altered by the disease since it was collected during the course of CVD rather than before or at the onset of the disease; thus, the analysis might still overestimate the associations.
Conclusions
Both SES-related health inequalities and lifestyle disparities should be comprehensively considered when caring for this population, and upstream structural interventions tailored to the target population are needed to reduce CVD burdens.
Data availability statement
Data are available on reasonable request. The data are not publicly available. The China Health Evaluation And risk Reduction through nationwide Teamwork only provides conditional data access for qualified researchers with legitimate requests; a formal application and research proposal is required. Please contact cvd-project@nccd.org.cn to seek approval for data access.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants. The central ethics committee at China National Center for Cardiovascular Diseases approved the project (2014-574). All enrolled participants provided written informed consent. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank the contributions that have been made by study teams at the Chinese National Center for Cardiovascular Diseases, and the local sites in the collaborative network in the realms of study design and operations.
References
Footnotes
YW, KP, WX and XH are joint first authors.
Contributors YWa, KP and XH were responsible for writing the original draft. YWa was responsible for the data analysis, and BC was responsible for checking the data analysis. XLiu, YL, JL, YS, GH, XZ, AT, WR, CW, YT, YWu, CL and WP were responsible for reviewing, revising and commenting on the article. YY, WX, JC and LS were responsible for project administration and data collection. XLi was responsible for conceptualisation and critical revisions. XLi and SH were responsible for the supervision of the project and are responsible for the overall content as guarantors. All authors interpreted the data and approved the final version of the article. YWa, BC, XLi and SH had access to the raw data.
Funding This work was supported by the National High Level Hospital Clinical Research Funding (2022-GSP-GG-4), the Guangdong Basic and Applied Basic Research Foundation (2021A1515110307) and the Shenzhen Clinical Research Center for Cardiovascular Diseases Fund (20220819165348002). The funders of the study had no role in study design, data collection, data analysis, data interpretation or writing of the report.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.