Socioeconomic Status (SES) is defined as the combined economic and social status of an individual, and health disparities disproportionately affecting people of lower SES have been a popular topic in the field of cardiovascular research [1]. Simultaneously, the validity of indicators measuring SES remains a major concern [1]. Although some published traditional indices such as income, education, and occupation have been responsible for successfully impacting a substantial proportion of cardiovascular events, it has been argued that no single parameter fully captures SES [2]. Notwithstanding, most studies in cardiovascular health employ a single SES predictor as a proxy for the effects of socioeconomic disadvantage [3, 4].

Using a single SES predictor oversimplifies the multifaceted and entwined dimensions of socioeconomic disadvantage and may erroneously overestimate the impact of a variable, ultimately underestimating the true impact of social hierarchy on health outcomes [5]. Even when using other social variables as covariates, adjustments are likely to be incomplete and result in limited reliability [5]. In real-life scenarios, different SES indicators tend to cluster and intersect together [6], which underscores the theory of intersectionality. The theory of intersectionality suggests that various socially constructed categories of discrimination (i.e., SES, race or ethnicity, etc.) interact to contribute to health disparities [7]. Thus, there is a need for novel markers of SES developed through rigorous design and methods.

Cluster analysis can allow the meaningful interpretation of the complex concept of SES and facilitate the identification of a vulnerable group of low SES individuals [8]. Understanding the potential clustering of SES risk markers could facilitate an etiologic understanding of cardiovascular health and identify possible interventions that could address mutually reinforcing contributors to Cardiovascular Disease (CVD) [9]. Many western studies support the feasibility of applying clustering to assess various SES conditions and associated cardiovascular health [10, 11]. For instance, Aungkulanon et al. [10] constructed a composite socioeconomic index using cluster analysis and explored the correlation between the composite score and cardiovascular mortality in Thailand. Additionally, Mirowsky et al. [11] examined residential-level SES clusters in central North Carolina with their clustering technique, which enables the examination of its relationship to various cardiovascular and metabolic biomarkers.

In South Korea (hereafter “Korea”), older adults are the most vulnerable population with greater socioeconomic disadvantages. A high percentage of the aging population is suffering from CVD or is at risk. Importantly, such cardiovascular health conditions are closely related to low SES [12, 13]. However, no study has explored clusters of older adults based on various intersecting SES indicators or how clustering contributes to cardiovascular health in Korea. Although previous Korean studies employed a wide range of SES factors as predictors [14, 15], each of which was investigated separately for its effect on cardiovascular health, this approach could obscure important social gradients in cardiovascular health.

Using cluster analysis, older adults were categorized into low-, middle-, and high-SES clusters. Older adults of lower SES clusters exhibited significantly higher CVD risk factors. Importantly, our approach provided a detailed overview of important SES factors that mutually contribute to CVD risk. For instance, we demonstrated that lower SES clusters comprised older adults who lived at the intersection of less favorable material conditions (e.g. low income, no home ownership, no private insurance, or unemployed) had adverse outcomes, which highlights “the material hypothesis” [23, 24]. This approach differs from early Korean studies of older adults which tended to report health outcomes of those in income poverty vs. those above the poverty line [12, 13]; these studies overlooked the significance of material resources for explicating the health of this population group, which protected household budgets from expensive care and ensured access to healthcare.

Among the material factors that contributed to adverse cardiovascular health, and when combined with other SES risk factors, the findings on employment was particularly noticeable: no individuals in the low-SES cluster were employed at the time of the survey, whereas those in the high-SES cluster were all employed. Although Korea's average age of labor market exit is the highest among OECD countries at 72.90 years for men and 73.10 years for women [19], unemployment among older adults remains a burgeoning problem despite the recognized value of this demographic in Korean society [25]. Thus, our results indicate the need to urgently address the impact of unemployment on CVD risk, especially among older adults in the low-SES cluster identified in our analysis.

The study included marital status and living arrangements (i.e. indices of social networks and support) as SES factors and demonstrated that those who were nonmarried/ unpartnered or lived alone mostly belonged to the low-SES cluster and had higher CVD risk factors. Studies emphasizing this “social-relational” aspect posits that health disparities cannot be entirely explained by well-known financial barriers or material factors [26, 27]. Yet, such social-relational aspects as disparity factors relating to cardiovascular health among older individuals have not received much attention in research in Korean [13]. This discussion is important as an increasing number of older adults are living alone in Korea [28]. The actual remarriage or cohabitation rate among oldest-old adults remains very low due to traditional social norms and lower economic status [29]. Our findings demonstrate that the proportion of people aged 75 and over were more concentrated in the lower SES cluster.

This study included gender-related data for analysis and evaluated intersecting identities with other SES factors, which is a study design differing from most Korean studies of disparity that adopt separate gender analysis for research in this field [12, 13]. Whilst there is no optimal research method for gender analysis, there is a growing awareness of the importance for research to move beyond simple binary thinking to consider how gender differences vary within specific situations or contexts; intersectionality is a way of thinking about this more complex profile of disparity [30]. Consequently, this study found that there are more women with multiple simultaneous socioeconomic disadvantages (e.g., no private insurance, unemployed, etc.) than men in the lower SES cluster-i.e., the group with significantly higher CVD risk factors. This agrees with the argument that connects an increased rate of cardiovascular events with the perception of “women's inferiority.” Within patriarchal society, elderly Korean women have existed in a subservient position to men; this culture subordinates low-SES elderly Korean women socially and economically, prevents their pursuit of health resources, and makes them more vulnerable to facing undesirable health events [31]. Considering this situation in Korea with our findings emphasizes the need to improve the cardiovascular health of disadvantaged elderly women and ensure their health equity.

Clinically, intersecting SES risk factors identified in this study can suggest vulnerability markers, identifying individuals who need either more medical attention or closer follow-up care. This approach could also identify subgroups who would benefit from structural interventions, such as alleviating socioeconomic barriers to hospital access or specialized cardiovascular care. Further, the regression analyses revealed that the lower SES clusters have a significantly higher prevalence of diabetes, hypertension, total cholesterol, and obesity than the higher ones. Importantly, these CVD risk factors did not decrease in recent years, particularly in those with low SES in Korea [32]. This underscores the need to focus on improving screening, early detection, and treatment for patients with such risk factors. The prevalence of metabolic syndrome was significantly higher in this cluster. This additional finding on “integrated CVD risk” evaluation is also important, as existing clinical guidelines commonly recommend applying primary interventions according to the magnitude of the integrated risk [33].

In this study, we identified “upstream” SES factors that affect cardiovascular health simultaneously with individual- level factors: people on public assistance or Korean Medicaid or those who reside in rural areas constitute a high proportion of the lower SES clusters with higher CVD risk. Such results are important for Korea, where the government's socioeconomic support system for older adults remains unsatisfactory, which raises concerns regarding health-related issues [34]. The high out-of-pocket payment in public assistance and limited benefit coverage in the Korean Medicaid system work as barriers to healthcare utilization, particularly in impoverished older adults. Additionally, the introduction of the Designated Doctor System to lessen Medicaid expenditures results in inconvenience in hospital use, particularly among the rural Medicaid beneficiaries [35]. From a policy standpoint, this provides strong support for policies that reduce overall CVD risk disparities.

We acknowledge several limitations of this study. First, our study design could not conclude causality between SES and cardiovascular health. Second, as some information was self-reported (e.g., SES, pre-existing medical conditions, etc.), we could not entirely exclude the possibility of response bias. Third, the results may not be widely applicable since SES conditions for older adults vary between different countries depending on their historical, social, and cultural contexts. Fourth, SES variables included in the secondary dataset potentially lacked depth because they were operationally defined by a single survey item or a subset of test items. For instance, unemployment does not necessarily reflect real income or welfare, for it can be argued that unemployed older adults today are materially better off than when fully employed. Additionally, living arrangements may include a range of relationship variations that are not reflected by any singular indicator. Fifth, education was not considered as the relative homogeneity of the sample concerning education variables (i.e., the majority of older adults belonged in the “little to no education” category), limiting our ability to detect education-related disparities in the outcome. Sixth, biomarkers reported in the dataset were measured at one point in time, which may affect the statistical validity of our conclusion. Last, the single imputation used in this study may reduce data variability, resulting in the underestimation of variances and standard errors.

Despite these limitations, this study demonstrates several strengths. First, an integrated socioeconomic indicator was developed and tested based on important SES indicators that adequately reflect disparities in cardiovascular health among older adults, which can be considered in future research. Second, the large sample size provided statistical power to determine robust relationships within clusters. Third, we used population-based data, significantly reducing the likelihood of selection bias. Last, this study provided preliminary evidence for the disparity mechanisms in CVD risk factors, which can suggest a clear role for the prevention of CVD morbidity and mortality, particularly for those with low SES.

Our study approach improved understanding of intersecting SES risk factors as associated with cardiovascular health among the older adult population. Importantly, we found out that CVD risk factors are significantly higher in the lower SES cluster which has a higher portion of those aged over 75 years, women, with a lower income, without private insurance and homeownership, single or living alone, unemployed, on public assistance, Korean Medicaid beneficiaries, and living in rural areas.

Our study has several implications for future research, clinical studies, and policy. First, studies targeting older adults should adopt fundamentally different conceptual approaches for measuring SES associated with cardiovascular health. Second, several intersecting SES risk factors identified in this study could be vulnerability indicators, identifying individuals who require more targeted screening, early detection, or focused medical attention. Last, the primary challenge for public professionals is to design tailored health policies that consider the socioeconomic variability within the older adult population.