PHWR Public Health Weekly Report

Statistics Korea predicts that the percentages of the Korean population aged 65 years or older, 75 years or older, and 85 years or older will rise from 17.4%, 7.3%, and 1.8% in 2022 to 47.7%, 31.9%, and 14.3% in 2072 [1]. Excluding Sejong-si, the Chungcheong region is expected to transition to a super-aged society by 2025 [2]. The rate of aging in the Chungcheong region is notably swift compared to other regions, with a higher proportion of older adult residents than in other areas [3].

In a society with an aging population, the importance of health management becomes even more critical. Implementing prevention-focused strategies is crucial to decrease the substantial costs associated with managing chronic diseases that would otherwise be necessary. With the exception of Sejong-si, which is in the early stages of an aging population, Daejeon, Chungcheongnam-do, and Chungcheongbuk-do have low rates of treatment, indicating challenges in chronic disease management and highlighting disparities between regions [4,-6].

Deaths from chronic diseases accounted for 79.6% of all deaths in the Republic of Korea (ROK) in 2021. The 2019 Korea Health Panel found that 60% of the older adult population (aged ≥65 years) had multiple chronic conditions. In contrast, the 2018 U.S. National Health Interview Survey (NHIS) reported that 68 million individuals, equivalent to 27.2% of adults, were affected by at least 2 chronic diseases [7].

Chronic diseases necessitate continuous management and regulation to prevent deterioration. Worsening of chronic illnesses can lead to a reduced quality of life due to the burden of management, changes in employment status, financial strain from medical expenses, and social challenges [8]. Specifically, managing blood pressure and blood glucose levels for hypertension and diabetes requires long-term healthcare services [9].

In 2018, the Ministry of Health and Welfare initiated the establishment and administration of a team and committee to promote chronic disease management in primary healthcare. This initiative involved integrating and connecting a pilot project for community-based primary healthcare and a project on fees for chronic disease management. This represents the largest government-led chronic disease management effort to date. However, there are areas that need improvement, such as a high patient dropout rate and insufficient monitoring and management of patients at the local government level [10]. The root causes of these issues include a lack of self-management skills for chronic diseases, inadequate education and counseling, and low medical fees [11].

Without a chronic disease management strategy in place for the entire Chungcheong region (including Daejeon, Sejong, Chungcheongbuk-do, Chungcheongnam-do), the region’s chronic disease-related statistics are worse compared to the national average and the previous year. Additionally, there is a growing health disparity between different parts of the region [11]. To address this issue, we analyzed the sociodemographic characteristics of the Chungcheong region and explored if major health-related indicators show a spatial correlation with neighboring regions. Our goal is to propose an effective chronic disease management plan.

In order to manage chronic health strategically in the Chungcheong region, we compared health behaviors and outcomes among 250 cities, counties, and districts in the ROK. We utilized open data from the Korean Statistical Information Service (KOSIS) and the 2022 Community Health Survey to enable analysis at the city, county, and district levels. Major health outcomes indices included hypertension prevalence, diabetes prevalence, and depression experience rate. Health behaviors examined were current smoking rate, monthly alcohol consumption, high-risk drinking rate, exercise performance rate, walking performance rate, healthy lifestyle performance rate, obesity rate, perceived stress, and subjective health status. Additionally, regional characteristics such as aging index, ratio of single-person households, gross regional domestic product (GRDP), number of hospitals per 100,000 population, number of essential medical clinics per 100,000 population, and unmet healthcare needs were considered. Mean values per city, county, or district were calculated for the Community Health Survey, with data representing a sample of 231,785 adult participants aged 19 years and older from 2022, accounting for individual weights. The KOSIS data were categorized for the 250 cities, counties, and districts.

We conducted univariate analysis of the health indices mentioned above to examine their spatial distribution characteristics. Our goal was to determine whether each index’s distribution was linked to neighboring areas, leading to the formation of geographical clusters. In addition to the health indices data, we also utilized administrative region boundaries data for cities, counties, and districts from censuses (SGIS).

Spatial analysis weighting was conducted using the contiguity-based queen method. By specifying the contiguity order as 1, we identified neighboring regions as those directly sharing a border. Isolated areas such as islands were not included in the calculations.

To verify the spatial autocorrelation of each explanatory variable across all 250 cities, countries, and districts nationwide, we calculated the global Moran’s I and tested the statistical significance using 9,999 permutations [12,13]. Moreover, we utilized a local indicator of spatial association (LISA) cluster map of the local Moran’s I to investigate the geographical distribution of clusters demonstrating spatial correlation [14]. Our results were visualized through plots indicating hot spots with a significant concentration of positive values (high-high, HH), cold spots with high concentrations of negative values (low-low, LL), and regions where the central region and neighboring areas showed inverse values, indicating positive and negative correlation (high-low [HL] or low-high [LH], respectively). All spatial analysis was conducted using GeoDa (GeoDa ver.1.22.; the University of Chicago).

We divided 17 cities and provinces into 250 cities, counties, and districts to investigate major health indices. We discovered that, in comparison to other regions, most of the Chungcheong area had higher rates of single-person households and a higher GRDP. The number of hospitals per 100,000 population was low, but there were many essential medical clinics (Supplementary Table 1). Unmet healthcare needs, rate of walking performance, rate of healthy lifestyle performance, and obesity rate were relatively low. However, the current smoking rate, monthly alcohol consumption, high-risk drinking rate, and rate of moderate or vigorous physical activity were high (Supplementary Table 2). Additionally, perceived stress, subjective health status, hypertension prevalence, diabetes prevalence, and depression experience rate were higher compared to other cities and provinces (Table 1, Supplementary Table 3).

Table 1 Status of major indicators related to chronic diseases by region

Variable	Chungcheong	Jeolla	Gyeong-sang	Gangwon	Metropolitan	Except Chungcheong average	Nationwide
Number of city & town	36	44	75	18	77	214	250
Socio-demographic characteristics
Ageing index (65 years+/ <15 years pop)	2.6	3.2	3.0	2.9	1.6	2.7	2.6
Single-person households rate (%)	36.9	37.5	36.7	37.2	33.5	36.2	35.9
GRDP (1,000,000 won)	60.6	34.5	46.9	37.4	61.0	45.0	50.4
Number of hospitals (per 100,000)	7.7	11.9	10.2	5.6	6.6	8.6	8.7
Number of Essential medical hospitals (per 100,000)	16.0	14.2	15.5	11.6	19.2	15.1	16.2
Unmet medical needs rate (%)	5.7	6.4	6.4	6.2	5.0	6.0	5.8
Health behavior
Current smoking (%)	18.8	17.1	18.2	20.4	17.4	18.3	18.0
Monthly drinking rate (%)	50.5	45.1	49.8	51.5	54.6	50.3	50.7
High risk alcohol consumption (%)	15.1	12.7	14.8	18.2	14.8	15.1	14.7
Moderate-to-vigorous physical activity (%)	23.2	24.4	23.3	22.2	21.4	22.8	22.8
Walking practice rate (%)	43.2	43.5	42.5	37.2	54.0	44.3	45.9
Health living practice rate (%)	18.1	20.8	18.7	15.6	21.3	19.1	19.6
Obesity rate (%)	31.6	31.8	30.5	35.4	31.2	32.2	31.4
Stress perception rate (%)	21.9	20.3	19.7	20.8	23.2	21.0	21.3
Self-rated health status (%)	44.4	41.2	39.7	42.0	47.3	42.6	43.2
Health outcome
Prevalence of hypertension (%)	28.4	29.6	26.3	31.7	22.8	27.6	26.5
Prevalence of diabetes (%)	12.2	13.1	11.9	12.8	10.1	12.0	11.7
Prevalence of depression (%)	7.7	7.1	7.1	7.0	7.4	7.2	7.2

GRDP=gross regional domestic product..

We analyzed the spatial distribution of major health indices and resources in 250 cities, counties, and districts nationwide. We calculated global and local Moran’s I to evaluate clustering characteristics. When we examined sociodemographic characteristics and healthcare resources, all six indices showed a Z-score of ≥1.96, indicating significant results with a pseudo p-value of <0.05. In the Chungcheong region, Boryeong and Buyeo were HH for the aging index, while Cheonan, Jincheon, Cheongju, part of Daejeon, and Sejong were LL, demonstrating a concentration of regions with the LL pattern. The ratio of single-person households, number of essential medical clinics, and unmet healthcare needs did not show significant spatial autocorrelation. GRDP exhibited an HH pattern in Cheonan and parts of Cheongju, and an LL pattern in Geumsan. The number of hospitals per 100,000 population displayed an LL trend in Dangjin and Hongseong (Figure 1).

Figure 1. Socio-demographic and spatial cluster characteristics of healthcare resource indicators in 250 cities, counties, and districts in Korea
GRDP=gross regional domestic product.

When we analyzed the spatial distribution and clustering characteristics of health behaviors in cities, counties, and districts, all indices were significant. The current smoking rate was concentrated in HH regions in Dangjin, Asan, Cheonan, Jincheon, Eumseong, Jungpyeong, Goesan, and northern Cheongju. Conversely, besides the HH regions in Cheonan and parts of Cheongju, some southern parts of the Chungcheong region, including Boryeong, Buyeo, Nonsan, and Yeongdong, showed LL trends for monthly alcohol consumption and high-risk drinking rate. Additionally, exercise performance rate, walking performance rate, and healthy lifestyle performance rate also exhibited spatial clustering with adjacent regions (Figure 2).

Figure 2. Spatial cluster characteristics of health behavior and health outcomes in 250 cities, counties, and districts in Korea

When analyzing spatial clustering of health behaviors and outcomes across cities, counties, and districts, significant associations were found for obesity rate, perceived stress, and subjective health status.

Obesity rates in the Chungcheong region displayed various patterns: an LH pattern in northwestern Cheonan-si, HL pattern in Gyeryong, HH pattern in Eumseong, Chungju, and Jecheon, as well as an LL trend in some parts of Daejeon. Perceived stress varied with an HL pattern in Boeun, HH pattern in Cheonan, and LL pattern in Nonsan and Yeongdong. Subjective health status showed an LH pattern in southeastern Cheonan, HL pattern in Chungju, and HH pattern in Gyeryong, Daejeon, and parts of Cheongju.

Depression experience rate, hypertension prevalence, and diabetes prevalence did not show significant associations, with Z-scores <1.96 and pseudo p-values ≥0.05 (Figure 2).

For strategic and efficient management of chronic diseases in the Chungcheong region (Chungcheongnam-do, Chungcheongbuk-do, Daejeon, Sejong), we conducted a detailed spatial analysis and visualization to identify health gaps by city, country, and district. Our analysis aimed to determine if there are any regional clusters related to health. Below are our findings:

First, upon analyzing major health indicators in 250 cities, counties, and districts, we found that the aging index in the Chungcheong region was relatively low, while the GRDP was high. The total number of hospitals was low, but there was a high number of essential medical clinics, and unmet healthcare needs were minimal. Rates of walking practice and healthy lifestyle practice were relatively low, while the rate of moderate or vigorous exercise practice, subjective health status, and obesity rate were high. However, smoking rate, drinking rate, perceived stress, hypertension prevalence, diabetes prevalence, and depression experience rate were higher compared to other regions.

These findings are in line with a previous study that reported similar results based on the Korea Disease Control Agency’s Community Health Survey in the Chungcheong region from 2018 to 2022. The study revealed that smoking rate, drinking rate, and early stroke and myocardial infarction symptoms worsened relative to the nation as a whole and compared to the previous year [11].

Second, when visualizing spatial clustering characteristics for sociodemographic variables in cities, counties, and districts nationwide, and investigating the significance of clustering in the Chungcheong region, we discovered a distinction between areas with concentrations of older adult (HH) or younger (LL) populations based on the aging index. We also identified a cluster in some parts of Chungcheongnam-do with a shortage (LL) of hospitals per 100,000 population. Given that the older adult population (≥65 years old) makes up around 20% of the total population in Chungcheongnam-do and 11.6% in Cheonan-si, while 40% of the populations in Seocheon-gun, Cheongyang-gun, and Buyeo-gun are older adults, we were able to pinpoint regional differences [5]. These regional differences in the extent of aging highlight the need for efficient management through the establishment of selective management strategies for chronic diseases.

Third, we depicted spatial clustering characteristics for health behavior indices in cities, counties, and districts nationwide, and explored the significance of clustering in the Chungcheong region. We noted a cluster with a high current smoking rate (HH) in northern Chungcheongnam-do, and a cluster with low monthly alcohol consumption and low high-risk drinking rates (LL) in various parts of the Chungcheong region. Notably, there were distinct HH (northern) and LL (southern) clusters in the Chungcheong region for high-risk drinking rate, indicating a strong regional clustering pattern compared to the nation as a whole.

Fourth, there was a pattern of low (LL) obesity rates, perceived stress, and subjective health status in some parts of Chungcheongbuk-do, while these were high (HH) in others. The Chungcheong region consists of four administrative zones (Daejeon, Sejong, Chungcheongnam-do, and Chungcheongbuk-do) with significant differences in aging, health behaviors, and chronic disease prevalence [11]. Our findings indicate heterogeneity among different parts of the Chungcheong region based on major health indicators. This highlights the importance of prioritizing areas for health management interventions and providing focused management in those areas. Furthermore, the lack of practical resources for local government, such as funding and personnel, along with shrinking populations, necessitate more diverse and efficient measures to ensure equality in health management intervention projects.

Recently, the Chungcheong Disease Response Center has worked on implementing tailored solutions for health issues through the “2023 Expert Forum to Develop Measures to Reduce Health Disparities in the Chungcheong Region,” which includes discussions on “Chronic Disease Health Disparities and Strategies in the Chungcheong Region” and “Development of Long-Term Strategies for Chronic Disease Management in the Chungcheong Region” [11].

Our study was a cross-sectional investigation limited to 2022 due to data constraints, preventing us from analyzing temporal clusters. Further research will be required to examine changes over time compared to other areas using spatiotemporal cluster analysis.

Unlike previous studies on regional health disparities, we focused on chronic disease-related indicators at a more detailed level including cities, counties, and districts. This highlights the importance of specialized chronic disease management plans that involve cooperation among regions with similar characteristics. Our findings can support collaborative efforts on intervention projects based on clustering patterns.

Ethics Statement: Not applicable.

Funding Source: This study was funded by the grant from the Korea Disease Control and Prevention Agency (2023-02).

Acknowledgments: None.

Conflict of Interest: The authors have no conflicts of interest to declare.

Author Contributions: Conceptualization: JHK, GML. Data curation: EAK. Formal analysis: MBP, EAK. Funding acquisition: SCL, SH Kang, SH Kim, ESK. Investigation: GML. Methodology: JHK, GML. Project administration: EAK. Resources: SH Kim. Software: GML. Supervision: JHK. Validation: GML, JHK. Visualization: GML. Writing – original draft: GML. Writing – review & editing: SCL, SH Kang.

Supplementary data are available online.