Healthcare Costs and Utilization Associated with Participation in the National Fitness Award Program: Analysis using Linked NHIS Data
Article information
Abstract
PURPOSE
This study examined the influence of participation in the National Fitness Award (NFA) program on healthcare costs and utilization by analyzing linked data from the NFA database and the National Health Insurance Service (NHIS).
METHODS
A total of 721 individuals who participated in the NFA program between 2013 and 2017 and consented to data linkage were included in the study. A control group comprising 1,426 individuals was selected from the NHIS big data through 1:2 propensity score matching based on age, sex, and healthcare utilization in the year before the index date. This study was conducted as part of an official NHIS research project. Changes in healthcare utilization and costs before and after NFA participation were analyzed and compared with the control group to assess the effectiveness of the program.
RESULTS
From the year before the intervention to the year following NFA participation, total healthcare costs increased by approximately 70,000 KRW in the participating group and by approximately 470,000 KRW in the control group. After adjusting covariates and applying the difference-in-differences (DiD) methodology, the NFA participants demonstrated the following reductions relative to the control group: 0.1 fewer total number of inpatient visits, approximately 402,000 KRW lower total healthcare costs, 1.9 fewer inpatient visit days, 2.4 fewer inpatient treatment days, and a reduction of approximately 339,000 KRW in inpatient healthcare costs.
CONCLUSIONS
The findings suggest that participation in the NFA program was associated with reductions in overall healthcare costs and utilization. These results highlight the effectiveness of fitness programs as a cost-efficient strategy for managing rising healthcare costs.
INTRODUCTION
In contemporary societies, the prevalence of noncommunicable diseases is rapidly increasing due to rapid aging and lifestyle modifications, thereby posing significant public health challenges and rising healthcare costs [1,2]. Chronic conditions such as obesity, hypertension, and diabetes not only diminish individual quality of life but also impose a substantial financial burden on national healthcare systems [3-5]. Consequently, there is a global shift in health policy from treatment-centered approaches to preventive healthcare strategies. In this context, state-led initiatives promoting physical activity and implementing health promotion programs based on fitness assessments are increasingly recognized as effective public health measures.
Physical inactivity ranks as the fourth leading cause of mortality worldwide [6] and is a risk factor for various chronic diseases, including coronary artery disease (6%), type 2 diabetes (7%), breast and colon cancers (10% each), and premature death (9%) [7]. In addition to its detri-mental effects on individual health, physical inactivity incurs substantial indirect costs, including increased healthcare costs and reduced produc-tivity, thereby exacerbating the socioeconomic burden [8]. Conversely, engagement in regular physical activity provides a range of health benefits, such as decreased risk of premature mortality, reduced incidence of cardiovascular disease and cancer, enhanced cognitive function, and improved mental well-being. These benefits collectively contribute to lower healthcare costs [9]. The World Health Organization highlights that the systematic monitoring of physical activity levels can enhance physical activity engagement and aid in the prevention of cardiovascular disease [10]. Globally, physical inactivity has been estimated to account for 1.2 to 2.5% of total annual healthcare costs [11]. For example, in Canada, the ParticipACTION campaign estimated that a 10% increase in physical activity participation could yield a reduction of approximately CAD 629 million annually in healthcare costs associated with heart disease and type 2 diabetes [12], suggesting that increased physical activity can make a substantial contribution to improving population health and reducing national healthcare costs.
Previous studies have established associations between physical activity, cardiorespiratory fitness, and healthcare costs. Davidson et al. reported that physical activity plays a critical role in improving cardiorespiratory fitness; however, exercise specifically designed to improve cardiorespiratory fitness, rather than merely increasing activity levels, may be more effective in promoting long-term health outcomes [13]. Furthermore, individuals with higher levels of cardiorespiratory fitness in middle age incurred significantly lower annual healthcare costs in old age compared to those with lower fitness levels. Notably, a 1 MET increase in cardiorespiratory fitness was associated with a 6.8% reduction in healthcare costs among men and a 6.7% reduction among women [14]. Similarly, a Korean study reported that high-risk individuals with low cardiorespiratory fitness spent approximately 320,000 KRW more per year on medical expenses than low-risk individuals with high cardiorespiratory fitness [15]. Collectively, these studies underscore cardiorespiratory fitness as a key determinant of both health status and healthcare costs.
In response to these findings, several countries have implemented various programs aimed at improving physical fitness and increasing participation in physical activity. In the United States, the FitnessGram test serves as a standardized national fitness assessment program for adoles-cents, evaluating multiple domains including cardiorespiratory endurance, strength, flexibility, and body composition. The program facilitates individual fitness profiling and personalized exercise guidance based on assessment results [16]. In Germany, the Deutsches Sportabzeichen program evaluates endurance, strength, agility, and coordination across various sporting activities, which is accessible to citizens of all ages and physical abilities, and awards an official medal of honor from the Federal Republic of Germany to individuals who meet prescribed performance criteria. The primary objectives are to promote public health and en-courage lifelong participation in sports [17]. These programs represent exemplary models of national programs that enable systematic assessment and management of fitness levels in youth and adults.
Aligned with this national trend, Korea has implemented the National Fitness Award (NFA), a program that offers free fitness assessments and individualized exercise prescriptions across the lifespan from infants to the elderly on a nationwide scale [18]. By evaluating various fitness fac-tors such as cardiorespiratory fitness and muscle strength, and providing visualized feedback along with personalized exercise recommendations (including intensity, frequency, and modality), the program enhances both exercise self-awareness and perceived feasibility of sustained participation. In addition, the program aims to improve population health through various behavior reinforcement strategies, such as physical fitness classes, physical activity monitoring, and exercise participation events. Empirical data from 2021 demonstrated a notable difference in annual medical expenses between participants in the National Physical Fitness 100 program (318) and non-participants (627), with participants incurring approximately 240,000 KRW less in total healthcare costs [19]. This finding may serve as preliminary evidence suggesting that participation in fitness assessments can yield economic benefits such as promoting health behaviors and reducing healthcare costs. Previous studies have also shown that receiving individualized fitness assessment results positively influences exercise intention, likely due to shifts in health perception and improvements in self-efficacy [20]. Furthermore, individualized exercise prescriptions have been identified as an effective means of converting exercise intentions into actions by refining exercise action plans [21]. However, existing studies are limited by small sample sizes and the narrow scope of healthcare utilization predictors, notably the re-striction to prior-year healthcare costs [19]. Therefore, the interpretation and generalization of the results should be approached with caution.
This study aims to more precisely analyze the effectiveness of the NFA program, addressing limitations identified in existing studies. To achieve this, we expanded the sample size and analyzed healthcare costs and utilization behaviors while accounting for multiple covariates, including gender, age, and income level. Furthermore, we employed the difference-in-differences (DiD) method, a causal effect analysis technique, to more accurately estimate the effect of program participation on healthcare cost reduction.
METHODS
1. Participants
This study utilized data linked internally by the National Health Insurance Service (NHIS) (2013-2017) for individuals who participated in research related to the NFA program and consented to the use of their personal information for data linkage. The dataset for participants in the NFA program was constructed by linking medical claims data from the NHIS using personal identification codes. A control group was selected through 1:2 random matching based on demographic variables, such as sex and age (as of the year before participation), using claims data from the Korea Health Insurance Corporation. The variables used for control group selection are summarized in Table 1. In total, participants were categorized into two groups: those who participated in the NFA Program (n=721) and a matched control group (n=1,426). The study received ethical approval from the Institutional Review Board of the Korea Institute of Sports Science in 2019 (IRB No. 20036-06-12).
Matching variables for control group extraction
2. Data analysis
This study linked physical fitness data from the NFA Program to medical claims data from the NHIS. Personal identification codes of participants were used to extract individual-level data, and the medical history database from the NHIS claims records (2013 to 2017) was utilized for analysis. Baseline personal information for both the participating and control groups was determined based on the year preceding participation in the program. To evaluate healthcare utilization and healthcare costs, data from the year before and the year after participation were analyzed. This approach allowed for an understanding of the impact of the project while accounting for temporal trends in healthcare usage, such as variations in the number of medical procedures.
Healthcare utilization and costs were assessed using the following indicators: number of visits, visit days, treatment days, prescription days, and total healthcare costs. Healthcare utilization data were categorized into three distinct groups: total (including both inpatient and outpatient services), inpatient only, and outpatient only. The “ number of visits” was defined as the number of claims submitted by healthcare institutions to the Health Insurance Review and Assessment Service (HIRA) for review. This definition reflects the current HIRA classification (i.e., number of claims submitted), which was previously based on the number of medical services provided. Medical institutions refer to facilities dedicated to delivering healthcare services and encompass a wide range of healthcare providers. These include tertiary and general hospitals, nursing hospitals, clinics, dental hospitals and clinics, maternity clinics, public health cen-ters, health stations, health clinics, oriental medicine hospitals and clinics, and pharmacies. “ Visit days” represent the number of days on which insured individuals (including medical aid recipients) received services or were hospitalized at healthcare institutions. “ Treatment days” indicate the number of days a patient physically attended a medical institution and received medication. Notably, days involving both visitation and medication administration were counted as a single treatment day. For prescriptions dispensed at pharmacies, the “ number of days” refers to the duration over which the medication was administered. The “ prescription days” metric denotes the cumulative number of days for which prescriptions were provided and medication administration occurred. “ Healthcare costs” represent the total expenditures incurred by healthcare institutions for treating insured patients. These include both the contribution of the insurer (i.e., national health insurance) and the out-of-pocket expenses of the patient, excluding non-covered services. Specifically, these costs refer to the amounts claimed by medical institutions approved through the review process of HIRA.
3. Statistical analysis
This study aimed to evaluate the effects of participation in the NFA program after adjusting for changes in healthcare utilization and costs. For each participant, the year of enrollment was defined as the baseline, with the preceding year and the subsequent year designated as the pre- and post-participation years, respectively. The characteristics of participants were analyzed using descriptive statistics, including distributions of hospitalization and outpatient costs, individual and outpatient utilization rates based on the year of program participation. Paired sample t-tests were performed to compare healthcare utilization and costs between the pre- and post-participation years in the participating and control groups. To account for potential confounding, analyses were adjust-ed for insurance premium deciles, baseline medical utilization, sex, and age. Differences-in-differences (DiD) analysis was further employed to assess changes in healthcare utilization and costs between groups. All statistical analyses were conducted using SAS software, version 9.4 (SAS Institute Inc., Cary, NC), with statistical significance was set at α=0.05
RESULTS
1. General characteristics of the participating and control groups
This study was conducted as part of a project implemented within the National Health Insurance Security Analysis System. Within this system, data access and analysis are restricted to a designated analysis environment, and only pre-approved results are permitted for download. Consequently, the study was limited in terms of organizing or conducting additional analyses on the demographic characteristics of the study participants. However, to enhance the validity of group comparisons, 1:2 matching was conducted based on gender, age, and prior healthcare utilization, thereby maximizing balance and internal validity.
The general characteristics of the participating and control groups are summarized in Table 2. The mean age of the participating group was 53.71±16.43 years, while that of the control group was 53.92±16.29 years. The 20th percentile value of insurance premiums was 12.83±6.29 in the participating group and 11.52±6.16 in the control group.
Characteristics of the participating and control groups
2. Comparison of healthcare utilization behaviors before and after the year of participation between the program participating and control groups
This study compared medical service utilization in the pre- and post-participation years for the program participating and control groups. Table 3 presents the results of this comparison. In the participating group, statistically significant increases were observed in several key metrics when comparing the pre- and post-participation years. These included the total number of outpatient visits (p =.0141), total number of prescription days (p <.0001), and total healthcare costs (p =.0025). Specifically, within the outpatient care, significant increases were observed in outpatient visit days (p =.0147), outpatient treatment days (p =.0270), outpatient prescription days (p <.0001), and outpatient healthcare costs (p<.0001).
Analysis of healthcare utilization before and after the participation year of the participating and the control groups using the paired t-test
The control group demonstrated statistically significant increases in the total number of inpatient visits (p =.0002), the total number of outpatient visits (p =.0332), total visit days (p <.0001), total treatment days (p <.0001), total prescription days (p <.0001), and total healthcare costs (p <.0001) compared to the pre-participation year. In the inpatient setting, inpatient visit days (p =.0005), inpatient treatment days (p =.0001), and inpatient healthcare costs (p <.0001) increased significantly in the control group. In the outpatient setting, statistically significant increases were observed in outpatient visit days (p =.0346), outpatient treatment days (p =.0042), outpatient prescription days (p <.0001), and outpatient healthcare costs (p <.0001). In the control group, all variables increased significantly, with the exception of inpatient prescription days.
Regarding total healthcare cost, both groups demonstrated a statistically significant increase in healthcare costs between the pre- and post-participation periods (p<.0001). When comparing the changes in annual per capita medical expenses, the participating group experienced an increase of approximately 70,000 KRW following program participation, while the control group experienced a smaller increase of approximately 47,000 KRW during the same period. This difference was statistically significant. A breakdown by medical expenses component showed no significant change in hospitalization costs for the participating group, while the control group showed a significant increase of 340,000 KRW (p <.0001). Conversely, outpatient expenses increased significantly in both groups (each p <.0001).
3. Difference-in-difference Analysis of healthcare utilization before and after the participation year of the participating and the control groups
Table 4 presents the results of the DiD analysis comparing healthcare utilization before and after the participation year between the participating and control groups. To control for trends in healthcare utilization, such as annual changes in healthcare costs, we selected insurance premium deciles, pre-participation healthcare utilization levels, gender, and age as covariates. The DiD results revealed that, compared to the control group, the participating group experienced a statistically significant reduction of approximately 0.1 in the total number of inpatient visits and a decrease of approximately 402,000 KRW in total healthcare costs. In the inpatient care category, the participating group recorded statistically significant reductions of approximately 1.9 inpatient visit days, 2.4 inpatient treatment days, and 330,000 KRW in inpatient healthcare costs relative to the control group (p <.05).
Comparison of healthcare utilization before and after the participation year by group using the difference-in-difference (DiD) method
DISCUSSION
This study aimed to examine the impacts of participation in the physical fitness assessment program of NFA on the healthcare costs and healthcare utilization of individuals. Through this analysis, we quantitatively evaluated the influence of physical fitness-oriented public policies on healthcare costs. The findings showed that participants in the NFA program experienced a comparatively attenuated increase in healthcare costs compared to the control group. Notably, there were significant reductions in inpatient-related healthcare utilization metrics, including the number of inpatient days and associated inpatient costs. These results suggest that engagement in the NFA program may promote healthier behaviors, ulti-mately contributing to a reduction in healthcare costs.
Previous studies examined the effects of program participation by adjusting for pre-participation healthcare utilization using a relatively small sample size (318 participants and 627 controls), and reported an annual difference of approximately 240,000 KRW in healthcare costs [18]. Conversely, this study employed a larger sample size (721 participants and 1,426 controls), controlled for confounding variables such as gender, age, and income level, and conducted a DiD analysis, resulting in an estimated annual healthcare cost reduction of approximately 400,000 KRW. This suggests a pronounced cost-saving effect than that reported in the previous study. Given the scalability of the NFA program, it is estimated that an annual medical cost savings of approximately 1.2 trillion KRW could be achieved if 300,000 individuals (as of 2019) participated in the program.
The core components of the NFA program are the physical fitness assessment and the provision of personalized exercise prescriptions. This integrated approach serves as a foundational intervention aimed at promoting the awareness of the health status of individuals and potentially leading to future exercise practices. Previous studies have shown that participants who received physical fitness assessment reported improvements in exercise intention, self-efficacy, and health status perception [20], all of which are considered critical antecedents of health behavior change. Personalized exercise prescriptions are an effective strategy for promoting behavioral changes as they offer specific, actionable guidance tailored to the fitness profiles of participants and health-related goals [21].
Factors such as goal setting, incentives, feedback, and social support have been identified as effective strategies for promoting sustained exercise behavior [22,23]. Evidence from prior studies indicates that even brief consultations conducted in routine clinical settings can significantly improve physical activity levels [24]. This theoretical foundation aligns structurally with the counseling and exercise prescription services provided by professional exercise specialists in the NFA program, suggesting that the program functions as a behavior-change-oriented health promotion initiative, rather than merely a physical fitness measurement tool.
Additionally, the NFA program reflects life cycle characteristics by conducting comprehensive physical fitness assessments across all age groups from preschoolers to older adults. These assessments include both health-related fitness components (cardiorespiratory endurance, muscle strength, and flexibility) and skill-related components (coordination and agility), and are evaluated using gender and age-specific standards [25]. The measurement results were provided in visualized reports, and personalized exercise prescriptions, including exercise intensity, frequency, and modality, were provided by exercise prescription specialists [26].
This study holds significant academic and policy implications, as it analyzes the effectiveness of the physical fitness assessment program using actual medical claims data from the NHIS, rather than relying on estimated social value indicators. By identifying the trend in which participation in the physical fitness certification program is associated with reduced medical expenses and improved healthcare utilization patterns, the study highlights the potential of the NFA program to serve as a preventive health management policy.
However, this study has the following limitations. First, the dataset used in this study was derived from the initial phase of the NFA program (2013 to 2017), which may not adequately capture recent changes in health behaviors, advancements in the healthcare system, or the ongoing effects of population aging. Second, because this study was conducted within the secure analysis environment of the NHIS, access to data was restricted, and only pre-approved results could be externally reported. Consequently, a detailed analysis of demographic characteristics was not feasible. Third, the scope of the analysis was limited to participants in the physical fitness measurement (certification) program of the NFA project; individuals engaged in other intervention types, such as long-term exercise programs, were not included.
Future studies should collect and analyze data on changes in physical activity levels, adherence to exercise programs, and health-related behav-ior following participation in the NFA project in order to more precisely identify the indirect effects of engagement in physical fitness measurement. Additionally, long-term follow-up studies and randomized controlled trials are required to more clearly establish causal relationships. These should be accompanied by detailed impact analyses targeting specific disease groups, such as cardiovascular disease and metabolic syndrome, and diverse population groups. Furthermore, to maximize policy effectiveness, it is essential to establish a foundation by introducing a sport activity certification system that complements the existing physical fitness certification. An integrated data system linking physical fitness records with medical and health big data should be developed. Finally, sustainable models for physical fitness enhancement programs should be explored through community-based initiatives and public-private part-nerships to achieve public value through the reduction of medical expenses at both individual and societal levels.
CONCLUSION
This study empirically confirmed that participation in the NFA program has a beneficial effect in suppressing increases in individual healthcare costs and reducing inpatient-related healthcare utilization. Compared to the control group, the participating group showed a smaller increase in overall healthcare costs, along with significant reductions in the number of inpatient days and associated inpatient healthcare costs. These findings suggest that physical fitness assessment programs may assessment to chronic disease management and medical cost reduction by supporting preventive health strategies. Notably, the physical fitness assessment program demonstrates the potential to enhance the efficien-cy of the healthcare system and alleviate pressure on the national healthcare budget, extending its value beyond health promotion. These results emphasize the importance of expanding physical fitness certification programs at the public policy level and support the development and implementation of systematic preventive healthcare management programs.
Future studies should investigate the long-term effects of the physical fitness assessment program and perform detailed analyses by population group and disease type to strengthen the evidence base and inform more targeted policy decisions. Establishing the program as a central component of preventive health management may yield measurable health and economic benefits for both the population and the national healthcare system.
Notes
CONFLICT OF INTEREST
This article was written using data from research funded by the National Sports Promotion Fund.
AUTHOR CONTRIBUTIONS
Conceptualization: S Park; Data curation: M Lee, SH Lee; Formal analysis: IH Oh; Funding acquisition: S Park; Methodology: IH Oh; Project administration: S Park; Writing-original draft: M Lee, SH Lee; Writing-review & editing: S Park, SH Lee.