Gender-Specific Factors Associated with Health-Related Quality of Life in Obese Korean Older Adults: Evidence from the 2020 Korean National Health and Nutrition Examination Survey

Given the increasing aging population in South Korea, the quality of life of older adults must be ensured. This cross-sectional descriptive study investigated the gender-specific factors associated with health-related quality of life in obese older adults aged 65 years and above based on Korean National Health and Nutrition Examination Survey (KNHNES) 2020 data. In total, 507 obese Korean older adults participated in the 8th KNHNES. Chi-square tests and logistic regression analysis were performed to determine the variation in health-related quality of life according to socioeconomic and health-related factors and assess their inter-relationships. The influencing factors of health-related quality of life in obese Korean older adults were national health insurance (odds ratio (OR) = 1.02, 95% confidence interval (CI): 0.40–2.21), private health insurance (OR = 0.36, 95% CI: 0.28–0.75), arthritis (OR = 6.64, 95% CI: 2.57–17.14), and good dietary lifestyle (OR = 0.07, 95% CI: 0.05–0.93) in men; and private health insurance (OR = 2.66, 95% CI: 1.05–6.72), arthritis (OR = 2.81, 95% CI: 1.44–5.51), and physical activity (OR = 4.33, 95% CI: 1.71–10.94) affected health-related quality of life in women. The importance of health behaviors should be considered in the development of health programs and interventions for improving the quality of life of older adults.


Introduction
The rate of aging in South Korea has been the highest among the Organisation for Economic Co-operation and Development member countries in the past four decades. According to Statistics Korea, the percentage of Korean older adults aged 65 years and above was 14.9% in 2019 and was predicted to surpass Japan (36.7%), the most aged country globally, by 2045 [1]. With an increasing older adult population, the life expectancy has also markedly increased to 82.4 years, resulting in a lengthened average lifespan. Hence, the concept of health in senescence is not simply defined as life sustenance but directed at "healthy aging" or living a self-satisfactory healthy life. Interest in enhancing the quality of life beyond disease prevention and treatment has thus steadily increased [2]. Older adults are generally frail due to declining physical and mental capabilities, and therefore, health-related quality of life (HRQOL) is of major concern. However, it is possible to improve their quality of life (QOL) by investigating factors related to HRQOL.
Obesity is an important worldwide public health issue, which increases the risk of metabolic syndrome, diabetes, hypertension, dyslipidemia, and cardiovascular disease, irrespective of ethnicity, gender, and age [3,4]. Moreover, obese older adults have limited social activities due to impairments in activities of daily living and mobility that directly or indirectly affect their QOL [4]. Psychosocial factors also play a critical role in the treatment of diseases [5]. Notably, body mass index (BMI) in older adults is an important indicator of physical and psychological QOL in senescence [6,7] that can be used to evaluate the need to devise an appropriate weight management plan [8]. In particular, overweight older adults have lower HRQOL than older adults with average weight [8].
Previous studies have assessed socioeconomic factors such as education, economic status, residential area, current employment [9][10][11], smoking habit, and healthy lifestyle [10] that influence HRQOL, as well as health-related behaviors including regular exercise, sleep, frequency of alcohol drinking, muscle training, and walking [12][13][14]. There were also studies investigating HRQOL in older adults with a focus on individual health status, performance of daily activities, depression, and health-behaviors [10,15,16].
A comprehensive analysis of the variations across these factors may contribute to providing appropriate policies and resources for the individuals concerned. It is crucial for projects and policies regarding health promotion behaviors to extensively consider factors such as age, gender, socioeconomic status, and health-related and psychological characteristics. Hence, it is necessary to determine the subjective health perception in Korean older adults for health promotion. However, it is challenging to comprehensively investigate the entire population in South Korea. In this study, we specifically determined the HRQOL in male and female obese Korean older adults aged 65 years and above using the data of the 2020 Korean National Health and Nutrition Examination Survey (KNHNES), with verified reliability and representativeness regarding socioeconomic status and physical and mental health variables to provide the basic data for developing policies to improve health quality based on gender.

Study Design
This study involved a secondary analysis using data from the 8th KNHNES, targeting the nationwide population in South Korea in 2020. The correlations of HRQOL in obese Korean older adults with socioeconomic status and physical and mental health variables were analyzed based on gender.

Study Population
The number of respondents to the KNHNES 2020 was 7359, with 1712 older adults aged 65 years and above. After excluding those without BMI data or with BMI less than 25 (n = 1110) and those without EuroQol-5 dimension (EQ-5D) data (n = 95), the number of participants included in this study was 507 ( Figure 1). Height was measured to an accuracy of 0.1 cm while the participants were standing straight without shoes. Weight was measured to an accuracy of 0.1 kg while the participants wore a light garment. BMI was calculated by dividing the weight (kg) by height squared (m 2 ). Thus, the obese group comprised those with BMI ≥ 25 kg/m 2 .
2.3. Instruments 2.3.1. HRQOL HRQOL was measured using a tool developed by Euroqol [17]. EQ-5D scores were measured using the self-report questionnaire. The tool consists of five subcategories: mobility (M), self-care, usual activities, pain/discomfort, and anxiety/depression. Each category is rated as no problems, some problems, and severe problems. The European Quality of Life-5 Dimensions (EQ-5D) index as the overall HRQOL score was calculated using an equation with weighted values. The highest score of one was obtained if all five subcategories were rated as no problems. Higher scores indicated higher HRQOL [17]. For the tool's reliability, the intra-class correlation coefficient in the test-retest analysis was 0.65. The validity of the tool was confirmed based on previously known group validity and convergent and discriminant validity [18].

Demographic Factors
Demographic factors comprised age, marital status, and total household members. Age was divided into three groups: <70 years, ≥70 years and <75 years, and ≥75 years. Marital status was divided into married (cohabitation) and unmarried (separated, bereaved, divorced, or unmarried) groups. Household was categorized as one-person (no cohabitant) and multi-person (one or more cohabitants) based on the response to the question "How many cohabitants are there in your household?"

Socioeconomic Factors
Socioeconomic factors comprised family income, education, economic activity, home ownership, basic livelihood security, health insurance type, and private health insurance. Family income was categorized as high, middle-high, low-middle, and low based on the total mean monthly income quartiles. Education was categorized as elementary school, middle school, high school, and college graduates. Economic activity was categorized based on the response to the question "In the past week, have you worked for one or more hours to earn an income or been an unpaid family worker for 18 h or more?" as No ("Unemployed") and Yes ("Employed"). Home ownership was categorized based on ownership of a house under the name of a household member as No ("None") and Yes ("Have"). Basic livelihood security was categorized into Yes and No. Health insurance type included National Health Insurance and Health Care Coverage. Private health insurance was categorized based on the response to the question "Are you insured by any private health care insurance supporting the health care cost such as cancer insurance, cardiovascular disease insurance, and accident cover insurance?" as No ("None") and Yes ("Have"). Unmet medical needs refer to the cases that did not receive necessary health care (examination or treatment) in the past year. The categories were based on the response as "I have had an unmet medical need" ("Have") and "I have never needed a health care service" ("None").

Health-Related Factors
Health-related factors comprised alcohol drinking, smoking, hypertension, arthritis, inactivity, sleep, exercise, unmet medical needs, and dietary lifestyle. Alcohol drinking was divided into "High-risk drinkers", defined as individuals drinking alcohol twice a week or more based on the Korea Disease Control and Prevention Agency (KDCPA) criteria (seven glasses or more for males and five glasses or more for females), and "Non-high-risk drinkers", defined as individuals drinking with lower frequency and amounts than "highrisk drinkers". Smoking was divided into "smoker" (current smokers consuming ≥ 5 packs a year) and "non-smoker" (past smokers and non-smokers). Hypertension and arthritis were defined as current cases diagnosed by physicians. Sleep duration was categorized based on the mean daily hours of sleep into groups of <5 h or ≥9 h and ≥5 h and <9 h. Exercise was analyzed based on the information of "a physical activity of a moderate level that makes you feel a little more fatigue than usual or a little out of breath" being performed for ≥30 min per session on five days or more a week, and of "a walk from one location to another" being performed for ≥30 min per session on five days or more per week. Annual weight change was categorized into "None", "Decrease", and "Increase". Dietary lifestyle was categorized as "Good" if an adequate amount of food was available and as "Poor" if financial difficulties prevented an adequate supply of food. Suicidal ideation and mental health counseling were categorized as "None" and "Have" based on the corresponding questions. Stress was categorized based on the question on the level of stress on a typical day: "None" if the response was a negligible level and "Have" if the response was a very high, high, or low level.

Data Analysis
Collected data were analyzed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). The variations in socioeconomic status, health-related factors, and HRQOL in obese Korean older adults were analyzed according to gender using the χ 2 test. A multivariate analysis was performed to determine the correlations of HRQOL with socioeconomic status and health-related factors in obese Korean older adults according to gender. For the multivariate analysis, the logistic regression that could incorporate the survey characteristics was used.

Ethical Considerations
The KNHNES was approved by the Institutional Review Board (IRB) of the KDCPA (IRB No. 2019-01-03-2C-A). The KNHNES raw data were available on the KDCPA website, data acquisition was approved, and the data were downloaded on 4 March 2022. The present study was exempt from approval by the IRB of the present institute (IRB No. SMU-EX-2022-04-003).

HRQOL According to Demographic and Socioeconomic Factors of the Participants
A total of 224 men and 283 women were included to determine whether the HRQOL varied significantly according to the demographic and socioeconomic characteristics. Among the demographic factors, marital status had a significant effect on HRQOL in men (t = 5.11, p = 0.005). Meanwhile, among the socioeconomic factors, family income (F = 3.146, p = 0.026), health insurance type (t = 6.883, p = 0.007), private health insurance (t = 5.594, p = 0.022), and unmet medical needs (t = 3.678, p = 0.019) had a significant effect on HRQOL in men, whereas health insurance type (t = 2.515, p = 0.038) and private health insurance (t = 1.951, p = 0.003) had a significant effect on HRQOL in women (Table 1).

Correlation of HRQOL with Socioeconomic Status and Health-Related Factors
The significant differences in the result of logistic regression and regression coefficients were analyzed in two models (Table 3). In Model 1, the correlations of HRQOL with socioeconomic factors were analyzed via logistic regression. The odds ratio (OR) for family income in men was 3.43 (95% CI = 1.30-9.04) for "high", with "low" as the reference to indicate statistical significance. The OR for "have" private health insurance in men was 0.43 (95% CI = 0.22-0.82), with "none" as the reference to indicate significance. The OR for "have" private health insurance in women was 0.57 (95% CI = 0.34-0.95), with "none" as the reference to indicate significance.
In Model 2, the variables were controlled by adding health-related factors in Model 1. Regarding socioeconomic factors in men, the OR for national health insurance in health insurance type was 1.02 (95% CI = 0.40-2.21), with health care coverage as the reference to indicate statistical significance. Meanwhile, the OR for "have" private health insurance was 0.36 (95% CI = 0.28-0.75), with "none" as the reference to indicate significance. Regarding health-related factors in men, the OR for "have" arthritis was 6.64 (95% CI = 2.57-17.14), with "none" as the reference to indicate significance, and the OR for "good" dietary lifestyle was 0.07 (95% CI = 0.05-0.93), with "poor" as the reference to indicate significance. In socioeconomic factors for women, the OR for "middle-to-high" family income was 2.66 (95% CI = 1.05-6.72), with "low" as the reference to indicate significance, and the OR for "have" private health insurance was 0.49 (95% CI = 0.25-0.95), with "none" as the reference to indicate significance. In health-related factors for women, the OR for "have" arthritis was 2.81 (95% CI = 1.44-5.51), with "none" as the reference to indicate significance, and the OR for "have" inactivity was 4.33 (95% CI = 1.71-10.94), with "none" as the reference to indicate significance.

Discussion
This study determined the gender-specific differences in HRQOL in obese Korean older adults based on demographic, socioeconomic and health-related factors, as well as analyzed the correlation using the data from the KNHNES 2020.
The HRQOL of obese female Korean adults based on demographic factors was low in several cases, which corroborates the findings of Hong [19]. Compared with males, female older adults showed longer life expectancy and duration of living alone, for which an extended scope of social participation, training and opportunity of economic activity, and public health care policy seems necessary. Thus, promotion of a more optimistic gender-specific HRQOL should be performed.
Meanwhile, the HRQOL based on socioeconomic factors of obese male older adults varied according to family income, health insurance type, private health insurance, and unmet medical needs, whereas in women, the HRQOL varied solely based on private health insurance. The markedly high OR for having private health insurance leading to high HRQOL was observed in both males and females. This coincided with a previous study reporting higher HRQOL for those with higher family income [20]. Individuals with low family income had difficulties in accessing health information and services, preventing them from receiving health promotion and disease prevention services [20]. Economic strength in older adults had a positive effect on health, and a key influencing factor of HRQOL was private health insurance [21]. In addition, the level of unmet medical needs in older adults was high in those with a low socioeconomic status. This was similar to the findings of Hoebel et al., who reported that the burden of transportation and medical cost of using health care services resulted in the burden of addressing medical needs in older adults [22]. A low socioeconomic status signifies a deferred decision of using health care services due to insufficient allocation for health management. Such unmet medical needs could cause health inequality and, hence, it is necessary to develop measures to increase the accessibility of the health care system to older adults with a low socioeconomic status. Continuous self-management through health behaviors, such as periodic medical checkups, that improve the physical health of an individual could enhance the QOL and life expectancy in older adults, as well as reduce the caregiving burden on the family and the government [23]. Recent studies on the relationship between HRQOL and health status based on the socioeconomic status including ethnicity showed that various socioeconomic resources had a significant impact on health outcomes [24,25]. This is consistent with the present study, placing further emphasis on the health inequality in older adults with a low socioeconomic status as an important social issue given the rapidly growing aging population. This implies the need for basic and long-term measures for the health of the older adult population in the future.
Furthermore, while all older adults experience aging-related changes and constraints, the way obese older adults viewed their aging has been shown to affect self-satisfaction related to health and QOL [26]. Hence, the influence on the QOL may be attributed not to the practice of ineffective health behaviors but to whether an individual regarded his or her health-related problems as a natural part of aging or as a constraining disorder or disease in terms of comorbidity, inactivity, and physical activity. Considering that the recommended intake of proteins, vitamins, and minerals for older adults based on the KDRIs is not much lower than that for younger adults [27], the result of poor subjective perception on dietary lifestyle in older adults raises concerns. Community nurses should take particular care in providing interventions tailored to the individual needs of older adults to ensure regular and balanced nutritional intake. They should relay information on programs such as the delivery of meals and side dishes and free or covered services for food and restaurants for older adults, while assisting in nutritional diagnosis and counseling, as well as taking on economic and political roles in meal preparation. In addition, arthritis and inactivity were shown to affect the physical and psychological domains of QOL, respectively [28]. The QOL will be enhanced with programs that reinforce optimism or motivation in older adults to increase their positive self-perception of health and with various assistive tools to alleviate inactivity and information on social resources to increase mobility.
This study has certain limitations. First, this study combined and analyzed the data of the KNHNES 2020, which was conducted at the time of this study as a cross-sectional investigation of the relevant variables. As the data of a cross-sectional study restrict timedependent before and after relationships, only the correlations, but not the cause-effect relations, could be determined between subjective health status and the socioeconomic factors, health-related factors, and psychological factors.
Second, the small sample size of the current data source implicated the presence of variables that could not be addressed in this study. If an adequate sample was ensured, we could have performed subgroup analyses according to age for identifying which age was more heavily influenced by obese factors. In addition, the monthly income in this study was based on the total mean monthly family income, where there was a possibility of dishonest responses; the resulting limitation was mitigated by applying an additional variable of health insurance type (e.g., national health insurance, health care coverage, no coverage) that could indicate the level of income. Moreover, missing values were found for approximately 30% of socioeconomic factors. Lastly, the health survey data used in this study were obtained through a self-reporting questionnaire, and thus there could have been certain bias in personal characteristics or circumstances in addition to the recall bias.

Conclusions
Variations in health behaviors were found in accordance with the analyzed factors, suggesting the need to incorporate such differences in the development of public health care policies and programs for obese older adults. As HRQOL exerts a significant effect on the health and mortality of obese older adults, various social efforts should be taken to ensure continuous monitoring of HRQOL and to support health services including health behaviors to enhance the HRQOL in obese older adults. Moreover, for public health care projects to reflect the ever-changing society, studies on the health behaviors that affect the HRQOL in obese older adults should be periodically conducted. Funding: This work was supported by Namseoul University.

Institutional Review Board Statement:
The KNHNES was approved by the Institutional Review Board (IRB) of the KDCPA (IRB No. 2019-01-03-2C-A). The KNHNES raw data were available on the KDCPA website, data acquisition was approved, and the data were downloaded on 4 March 2022. The present study was exempt from approval by the IRB of the present institute (IRB No. SMU-EX-2022-04-003).