Healthy aging in frail older adults: Active aging project of a national survey

Abstract

Background

Frailty is a growing concern among the aging population. Due to their vulnerability and reduced physiological reserves, frail older adults face an increased risk of functional decline and loss of independence. In frail older adults, the concept of healthy aging emphasizes maintaining independence, functionality, and a high quality of life despite frailty and other age-related challenges.

Objective

Grounded in the multidimensional model of healthy aging and aging-related resilience theory, we aimed to propose and test a hypothesized model. The model included exogenous variables (cognitive and physical function, perceived health) and mediators (resilience, social support, and daily activity) as influencing factors of healthy aging among frail older adults.

Methods

The cross-sectional study was conducted in South Korea using a national survey of 505 frail older adults living in the community. The sample was representative across age groups, sex, and regional distributions. A structural equation modeling was employed to test a hypothesized model, examining both direct and indirect effects of influencing factors on healthy aging. Data were collected between October and December 2023 and analyzed using IBM SPSS 26.0 and AMOS 26.0.

Results

A total of 505 frail older adults with an average age of 74 participated in the study. The hypothesized model demonstrated a good fit with the data. The exogenous variables and mediators accounted for 43 % of the variance in healthy aging. Resilience, cognitive function, and perceived health had significant direct and indirect effects on healthy aging. Daily activity also had a significant direct effect on healthy aging. Social support, while not directly affecting healthy aging, exerted a significant indirect effect through daily activity. Similarly, physical function influenced healthy aging indirectly via daily activity. The model was tested after controlling for age, sex (ref=male), economic status, and years of education. Among these confounding variables, economic status was a significant influencing factor in healthy aging (β = 0.14, p = 0.016).

Conclusion

Cognitive and physical function, alongside resilience and social support as mediators, significantly associated with healthy aging in frail older adults through daily activity. Health promotion strategies could focus on resilience and social support to empower individuals and strengthen social connectedness, thereby supporting independence in daily life and healthy aging. Further, we need to explore diverse social and cultural activities associated with healthy aging in this population to expand the definition of active aging.

What is already known

• •Healthy aging involves optimizing physical, mental, and social well-being to maintain independence, functionality, and a high quality of life, even in the presence of frailty and age-related conditions.
• •Social isolation and loneliness are significant challenges for frail older adults, often worsened by limited social support, which negatively affects their healthy aging.
• •Resilience, defined as the ability to adapt and recover after experiencing stress or adversity, is a fundamental component of healthy aging in older adults.

What this paper adds

• •Maintaining independence in daily life was critical for achieving healthy aging in frail older adults.
• •Resilience played a pivotal role as a mediator, linking cognitive function and engagement in daily activities to healthy aging.
• •Cognitive and physical function, mediated by resilience and social support, accounted for 43 % of the variance in healthy aging of frail older adults through the engagement in daily activities.

1. Introduction

Frailty is a growing concern among the aging population. Among community-dwelling older adults worldwide, the prevalence of frailty varies widely, ranging from 12 % for physical frailty to 24 % for a frailty index, and is expected to increase with aging (O'Caoimh et al., 2021). Frailty is associated with increased risks of disability, hospitalization, and mortality (Hoogendijk et al., 2019), as well as significant impacts on the quality of life and overall health of older adults (Vermeiren et al., 2016). Due to their vulnerability and reduced physiological reserves, frail older adults face heightened risks of functional decline and loss of independence in daily life (Ludlow et al., 2023). Frail older adults often experience interrelated health problems that can adversely affect their quality of life and impede healthy aging (Allison et al., 2021; Kojima et al., 2019).

In frail older adults, healthy aging refers to maintaining independence, function, and a high quality of life despite frailty and other age-related conditions (Liu et al., 2023). Achieving healthy aging requires proactive measures to manage chronic diseases, enhance physical and cognitive functions, and foster social connections and emotional well-being (Ludlow et al., 2023). By focusing on these factors, frail older adults can overcome age-related challenges, maintain independence in daily activities, and achieve healthy aging. Active aging aligns with the concept of healthy aging, embracing a nontraditional paradigm of human aging. Rather than accepting functional decline and chronic diseases as inevitable consequences of aging, active aging emphasizes delaying functional decline and postponing morbidity through engagement in diverse activities that promote independence (Fries, 2012).

According to the multidimensional model of healthy aging (Rivadeneira et al., 2021), this process is influenced by intrinsic capacities, such as cognitive and physical function, and interactions with the social environment. Physical and cognitive functions are key components of intrinsic capacity, playing a crucial role in maintaining healthy aging. These capacities significantly impact older adults' ability to perform daily activities and sustain their quality of life (WHO, 2020). Interactions with the social environment, including relationships with family and friends, provide critical social support and connectedness (Rivadeneira et al., 2021). When frail older adults lose social support systems, they are at increased risk of social isolation and loneliness, which negatively affect their healthy aging process (Cacioppo and Cacioppo, 2014). Social support coming from family, friends, community groups, and formal services can mitigate these risks, offering emotional, informational, and practical assistance.

Resilience is another key pillar of healthy aging (Merchant et al., 2022). It is a dynamic process involving positive adaptation and recovery after experiencing stress or adversity (Windle, 2011). Frail older adults exhibit diverse responses to stressors, depending on the assets and resources available to them for navigating and managing challenges (Hale et al., 2019). Resilience has been associated with improved psychological and functional outcomes, fostering better coping mechanisms, and encouraging active engagement in daily life (MacLeod et al., 2016). Similarly, social support provides emotional, informational, and practical assistance, enhancing motivation, reducing stress, and enabling older adults to perform daily tasks independently (Taylor, 2011). Research indicates that resilience and social support often work synergistically, promoting connectedness critical for sustaining independence in frail older adults (Windle, 2011). Nursing strategies that focus on enhancing resilience and providing social support can significantly improve independence, thereby contributing to healthy aging among frail older adults facing age-related changes or health conditions (Beard et al., 2022).

Based on the multidimensional model of healthy aging (Rivadeneira et al., 2021) and prior research, we proposed the following assumptions for a conceptual framework for healthy aging: (1) Engagement in daily activities is associated with healthy aging in frail older adults; (2) Maintaining physical and cognitive function, along with perceived health, supports independence in daily activities; and (3) Resilience and social support act as mediators, influencing independence in daily activities and healthy aging (Fig. 1-A).

Fig. 1.

Hypothesized model of the study

Note. A. H=hypothesis; H₁₈ = indirect effect of physical function on healthy aging through daily activity; B. Dotted lines refer to non-significant coefficients.

The model included three exogenous variables and three mediating variables. The hypotheses posited in the study are structured as follows.

1. Resilience as an endogenous variable

• H₁. Cognitive function has a significant direct effect on resilience.

• H₂. Perceived health has a significant direct effect on resilience.

• H₃. Physical function has a significant direct effect on resilience.

2. Social support as an endogenous variable

• H₄. Cognitive function has a significant direct effect on social support.

• H₅. Perceived health has a significant direct effect on social support.

• H₆. Physical function has a significant direct effect on social support.

• H₇. Resilience has a significant direct effect on social support.

3. Daily activity as an endogenous variable

• H₈. Resilience has a significant direct effect on daily activity.

• H₉. Cognitive function has a significant direct and indirect effect on daily activity.

• H₁₀. Perceived health has a significant direct and indirect effect on daily activity.

• H₁₁. Physical function has a significant direct and indirect effect on daily activity.

• H₁₂. Social support has a significant direct effect on daily activity.

4. Healthy aging as an endogenous variable

• H₁₃. Resilience has a significant direct effect on healthy aging.

• H₁₄. Cognitive function has significant direct and indirect effects on healthy aging.

• H₁₅. Perceived health has significant direct and indirect effects on healthy aging.

• H₁₆. Daily activity has a significant direct effect on healthy aging.

• H₁₇. Social support has a significant direct and indirect effect on healthy aging.

• H₁₈. Physical function has a significant indirect effect on healthy aging.

2. Methods

2.1. Study design and setting

This cross-sectional study involved 505 frail older adults living in the community in South Korea, ensuring a representative sample across age groups, sex, and regions.

2.2. Participants

The Korean FRAIL scale (Jung et al., 2021) was used as an initial screening tool for determining eligibility for the study, including those who responded 'yes' to two (pre-frail) or more items (frail on three or more). We invited older adults aged 65 and older residing in the community and willing to participate in the interview to complete a standardized questionnaire. Institutionalized or hospitalized individuals who had difficulty communicating and who had limitations in daily activities due to dementia or other diseases were excluded from the survey. A total of 505 older adults completed the questionnaire and were included in the analysis.

2.3. Sample size calculation

Although there is no consensus in the literature regarding the appropriate sample size for a study using structural equation modeling (SEM), a conventional sample size of 200 has been recommended when using the maximum-likelihood method in a structural model (Hoogland and Boomsma, 1998). In addition, a widely accepted rule of thumb requires a minimum of 10 participants per parameter when conducting structural equation modeling. With 28 parameters and seven observed variables in our study, a sample size of 280 was required. The actual sample size of 505 for the national survey met both requirements.

2.4. Data collection procedure

The sample was selected using a convenience sampling method from small-medium-large cities and rural areas across South Korea. In each region, data were collected through face-to-face interviews at apartment complexes, senior centers, shopping malls, and bus terminals. Interviewers with face-to-face survey training were carefully selected and familiarized with the questionnaire prior to conducting data collection across multiple regions. Participants were interviewed in a quiet environment where they felt comfortable and able to communicate without interruptions. The interviews generally took between 40 and 50 min, and the participants were permitted to take a break if necessary before continuing with the interview.

2.5. Ethical consideration

We conducted the study in accordance with the Declaration of Helsinki. Prior to data collection, this study protocol was approved by the Institutional Review Board at Chungnam National University (No. 202210-1-SB-139-01). All study participants were informed of the purpose and procedure of the study. We obtained informed consent in writing and explained to the subjects their right to withdraw from the study at any time, as well as the assurance that their information would remain confidential.

2.6. Measures

2.6.1. Healthy aging

Healthy aging was assessed using the Healthy Aging Scale for frail older adults developed and validated in frail Korean older adults (Seo et al., 2023). The scale consists of 12 items measured in Likert type from 1′not at all’ to 4′very much so’. A higher score indicates better healthy aging. Content validity, criterion validity, and confirmatory factor analysis were reported with satisfactory goodness-of-fit parameters (Seo et al., 2023). Cronbach's alpha for the development study was 0.81, and it was 0.80 for the present study.

2.6.2. Resilience

Resilience was assessed using a scale specifically developed for frail older adults in this study. This scale was designed based on a literature review (Resnick and Inguito, 2011; Wagnild and Young, 1993) and in-depth interviews to reflect the cultural characteristics of frail older adults in Korea. Initially, 40 items were identified through the literature, and 15 additional items were derived from interviews with eight frail older adults. The items were reviewed to eliminate redundancies, resulting in a preliminary scale of 10 items representing resilience. A panel of five experts in nursing and gerontology evaluated the items for content validity. Following their review, the final scale included eight items, each with a content validity index of 80 % or higher. Two items were excluded due to their negative content, such as "When I'm worried, I get somatic symptoms" and "It takes me a long time to get over failure or disappointment." The final version of the resilience scale consisted of eight items measured on a 4-point Likert scale, ranging from 'not at all' (1 point) to 'very much so' (4 points), with higher scores indicating greater resilience. In this study, the scale demonstrated a Cronbach's α of 0.66, which could be considered acceptable for a newly developed instrument and short scales with fewer than 10 items (Taber, 2018).

2.6.3. Physical function

Physical function was assessed using the physical summary score of 12-item Short Form Health Survey (Ware et al., 1996). The scale was used with the permission from QualityMetric, and the physical summary score was calculated by the standardized scoring program with a range of 0 to 100. The Cronbach's α of the scale was 0.89 in the present study.

2.6.4. Cognitive function

Cognitive function was evaluated using the mild cognitive impairment questionnaire, which was originally designed to assess cognition-related quality of life in individuals with cognitive impairment (Dean et al., 2014). For this study, the validated Korean version of the scale (Song et al., 2021) was used with a Likert scale ranging from 1 'almost never' to 4 'always'. The scale was reverse-coded, so higher scores indicated better cognitive function among frail older adults. The Cronbach's alpha for the validation study was 0.93 (Song et al., 2021), and it was 0.88 for the present study.

2.6.5. Perceived health

Perceived health was assessed using a single item of self-reported health, 'How would you rate your health?' on a 10-point numerical rating scale ranging from 1 ' poor' to 10 ' excellent’. The self-perceived health rating was validated as a simple tool strongly associated with physical health, mental health, and cognitive function (Caramenti and Castiglioni, 2022).

2.6.6. Daily activity

The engagement of daily activity was measured using the Korean version of the Falls Efficacy Scale-International (Park et al., 2010) to assess confidence in performing various daily activities without falling. In this study, we utilized a Likert scale ranging from 1 'not at all confident' to 4 'very confident' to assess the level of confidence in engaging in both indoor and outdoor activities. Higher scores indicate greater confidence in participating in such activities. The Falls Efficacy Scale-International has been validated in Korean populations with a Cronbach's alpha of 0.97 (Park et al., 2010), and it was 0.90 for the present study.

2.6.7. Social support

Social support was assessed using a 10-item scale representing emotional, informational, and instrumental support (Kim, 2024). This scale was developed on the basis of the Social Support Survey of Medical Outcomes Study (Sherbourne and Stewart, 1991) and revised to reflect the cultural context of community-dwelling frail older adults in Korea. Social support was assessed on a Likert scale from 1 'not at all' to 5 'very much', with higher scores indicating greater social support. The Cronbach's α of the validation study was reported as 0.81 (Kim, 2024) and 0.82 in the present study.

2.7. Data analysis

A SEM analysis was employed to test a hypothesized model positing that both direct and indirect effects of influencing factors on healthy aging among frail older adults. This approach enabled us to assess complex relationships and mediating effects among variables (Kline, 2015). An analysis of the research problem and hypothesis verification was conducted using IBM SPSS 26.0 and AMOS 26.0 programs with a significance level of <0.05. Descriptive statistics were used to analyze sociodemographic characteristics of the participants, and Pearson's correlations were calculated to assess associations between the main study variables. Multivariate normality testing in the AMOS program was used to verify the normality of the sample. Mean, standard deviation, skewness, and kurtosis were used as parameters to determine the normality of the sample. Multicollinearity among measurement variables was evaluated using correlation coefficients, tolerance values, and the Variance Inflation Factor (VIF). We estimated the fit of the hypothesis model using the Maximum Likelihood Estimation Method (ML) and evaluated it using absolute fit indices, incremental fit indices, and parsimonious fit indices. The significance of the paths in the structural model was verified using unstandardized coefficients (B), standardized coefficients (β), standard error (SE), critical ratio (CR), and p-values. The explanatory power for endogenous variables was analyzed using the Squared Multiple Correlation (SMC, R²). The statistical significance of the direct, indirect, and total effects of the model was verified using bootstrapping.

3. Results

3.1. Sociodemographic characteristics of the participants

Over 500 frail older adults, with an average age of 74 years (ranging from 65 to 89), participated in the study. Approximately half of the participants had no education or had completed only primary education. The sex distribution was balanced, and most participants described their economic status as middle or low class. While the majority had not experienced falls in the past year, most were taking at least one medication for a chronic illness. Additionally, about one in five older adults lived alone (Table 1).

Table 1.

Sociodemographic characteristics of the participants (N = 505).

Categories	Variables	n (%)	Mean±SD
Age, years			74.07 ± 6.72
Education, years			8.55 ± 3.95
	No formal education	48 (9.5)
	Primary school	190 (37.6)
	Middle school	102 (20.2)
	High school	136 (26.9)
	College or above	29 (5.7)
Sex	Male	227 (45.0)
	Female	278 (55.0)
Economic status	High	8 (1.6)
	Middle	298 (59.0)
	Low	199 (39.4)
Fall history(last year)	Yes	69 (13.7)
	No	436 (86.3)
Medication due to chronic disease	Yes	416 (82.4)
	No	89 (17.6)
Living status	Live alone	108 (21.4)
	Live with spouse	356 (70.5)
	Live with family	41 (8.1)

Note. SD=standard deviation.

3.2. Preliminary analysis

3.2.1. Descriptive statistics

Table 2 presents the correlations, means, and standard deviations (SD) for main study variables. Healthy aging was significantly correlated with the predictor variables, with correlation strengths ranging from small (e.g., physical function) to moderate (e.g., resilience). Daily activity also showed significant correlations with other predictors, ranging from small (resilience) to moderate (physical function). Most predictors showed moderate correlations; however, physical function was not significantly correlated with either cognitive function or resilience. No potential multi-collinearity was suspected since the level of the variance inflation factors between predictors ranged from 1.15 – 1.99 with the level of tolerance greater than 0.50 (Field, 2018).

Table 2.

Mean and correlation coefficients among study variables.

Variables	Mean (SD)	Range	Skewness	Kurtosis	CF r	PH r	PF r	RE r	SS r	DA r
CF	33.1 (4.60)	16–39	-1.38	1.47	1
PH	5.45 (1.53)	2–10	0.36	-0.77	0.28⁎⁎	1
PF	46.4 (6.27)	29–62	0.10	-0.77	0.05	0.38⁎⁎	1
RE	24.4 (2.59)	13–31	-0.49	0.31	0.32⁎⁎	0.17⁎⁎	0.08	1
SS	37.2 (4.64)	19–48	-0.42	0.21	0.25⁎⁎	0.21⁎⁎	0.10*	0.26⁎⁎	1
DA	38.5 (6.31)	21–48	-0.10	-1.01	0.33⁎⁎	0.64⁎⁎	0.42⁎⁎	0.25⁎⁎	0.28⁎⁎	1
HA	36.9 (4.33)	22–46	-0.76	-0.08	0.47⁎⁎	0.43⁎⁎	0.11*	0.44⁎⁎	0.29⁎⁎	0.42⁎⁎
Multivariate normality Kurtosis = 4.9, CR = 4.9

^⁎⁎p < 0.01

Note. SD=standard deviation; CR=critical ratio; CF=cognitive function; PH=perceived health; PF=physical function; RE=resilience; SS=social support; DA=daily activity.

3.2.2. Normality

The normality of data for the main variables was confirmed by examining skewness and kurtosis levels. The distribution of skewness for the study variables was within the absolute value of 3, and the distribution of kurtosis was within the absolute value of 7. The results of the multivariate normality test satisfied the assumption of multivariate normality (Bentler, 2006).

3.2.3. Model identification

The Maximum Likelihood Estimation Method was used in SEM to structurally verify the relationship between 11 study variables (7 exogenous and 4 endogenous variables). The exogenous variables included cognitive function, perceived health, and physical function, along with four unobserved error terms. Endogenous variables included resilience, social support, daily activity, and healthy aging. The identification of the hypothesized model depends on the amount of information provided to determine the free parameters (Woo, 2022). This study with seven observed variables had 28 distinct sample moments and 26 distinct parameters to estimate, resulting in 2 degrees of freedom, which allowed for the identification of the model as an over-identified model. The hypothesized model provided a good fit for the analyzed data (Chi square = 5.54, p = 0.06, Goodness of fit index = 0.99, Standardized Root Mean square Residual = 0.01, Root Mean Square Error of Approximation = 0.05, Incremental Fit Index = 0.99, Turker-Lewis index = 0.95, and Comparative Fit Index = 0.99). The models explained 49 % of the variance in daily activity and 42 % of the variance in healthy aging. The coefficients (b) and standardized coefficients (β) of all the direct and indirect relationships in the final model can be found in Table 3, Table 4.

Table 3.

Standardized estimation of the hypothesized model (N = 505).

Endogenous variables	Exogenous variables	B	β	SE	CR(p)	SMC
Resilience	Cognitive function	0.17	0.29	0.02	6.82(0.018)	0.11
	Perceived health	0.13	0.07	0.08	1.56(0.073)
	Physical function	0.01	0.03	0.01	0.75(0.575)
Social support	Cognitive function	0.15	0.46	0.04	3.31(0.011)	0.11
	Perceived health	0.35	0.12	0.14	2.46(0.028)
	Physical function	0.03	0.03	0.03	0.80(0.448)
	Resilience	0.35	0.19	0.08	4.33(0.008)
Daily activity	Resilience	0.19	0.07	0.08	2.27(0.023)	0.49
	Cognitive function	0.18	0.13	0.04	3.77(0.006)
	Perceived health	2.02	0.49	0.14	13.66(0.010)
	Physical function	0.21	0.21	0.03	6.17(0.009)
	Social support	0.14	0.11	0.04	3.14(0.026)
Healthy aging	Cognitive function	0.02	0.27	0.003	7.21(0.011)	0.43
	Resilience	0.04	0.27	0.005	7.22(0.010)
	Perceived health	0.05	0.23	0.010	5.18(0.003)
	Social support	0.01	0.08	0.003	2.19(0.184)
	Daily activity	0.01	0.09	0.003	2.01(0.020)

Note. B=unstandardized coefficient; β=standardized coefficient; SE=standard error; CR=critical ratio; SMC=squared multiple correlations.

Table 4.

Standardized direct, indirect, and total effects of the Hypothesized Model (N = 505).

Endogenous variables	Exogenous variables	Standardized direct effect	Standardized indirect effect	Standardized total effect
		β (p)	β (p)	β (p)
Resilience	Cognitive function	0.30(0.012)	–	0.30(0.012)
	Perceived health	0.07(0.077)	–	0.07(0.077)
	physical function	0.03(0.592)	–	0.03(0.592)
Social support	Cognitive function	0.15(0.008)	0.06(0.012)	0.21(0.009)
	Perceived health	0.11(0.030)	0.02(0.050)	0.13(0.020)
	Physical function	0.04(0.424)	0.01(0.509)	0.05(0.440)
	Resilience	0.19(0.008)	–	0.19(0.008)
Daily activity	Resilience	0.08(0.023)	0.02(0.016)	0.10(0.014)
	Cognitive function	0.13(0.006)	0.04(0.015)	0.17(0.006)
	Perceived health	0.49(0.010)	0.02(0.007)	0.51(0.009)
	Physical function	0.21(0.010)	0.01(0.324)	0.22(0.019)
	Social support	0.11(0.019)	–	0.11(0.019)
Healthy aging	Resilience	0.28(0.010)	0.02(0.010)	0.30(0.015)
	Cognitive function	0.28(0.012)	0.12(0.012)	0.40(0.015)
	Perceived health	0.19(0.003)	0.08(0.009)	0.28(0.004)
	Daily activity	0.12(0.045)	–	0.12(0.028)
	Social support	0.05(0.202)	0.01(0.013)	0.06(0.146)
	Physical function	–	0.04(0.045)	0.04(0.045)

Note. β=standardized coefficient; All values are standardized effects after controlling for age, gender (ref=male), economic status, and education years.

3.3. Testing of the hypothesized model

3.3.1. Hypothesis with resilience as an endogenous variable

Cognitive function (H₁), perceived health (H₂), and physical function (H₃) together explained 11 % of the variance in resilience, but only cognitive function showed a statistically significant direct effect on resilience (β = 0.30, p = 0.012), accepting the hypothesis (H₁).

3.3.2. Hypothesis with social support as an endogenous variable

Cognitive function (H₄), perceived health (H₅), and resilience (H₇) showed statistically significant direct effects on social support, explaining 11 % of the variance in social support. However, the hypothesis (H₆) was rejected because physical function did not show significant direct effect on social support.

3.3.3. Hypothesis with daily activity as an endogenous variable

Cognitive function (H₉), perceived health (H₁₀), and physical function (H₁₁), with resilience (H₈) and social support (H₁₂) as mediators, explained 49 % of the variance in daily activity. Social support had a significant direct effect, while resilience, cognitive function, and perceived health had statistically significant direct and indirect effects on daily activity. Thus, hypotheses 8, 9, 10 and 12 were accepted. However, hypothesis 11 was partially accepted since physical function (H₁₁) directly affected daily activity but not through resilience or social support.

3.3.4. Hypothesis with healthy aging as an endogenous variable

The predictor variables and three mediators (resilience, social support, and daily activity) explained 42 % of the variance in healthy aging. Resilience (H₁₃), cognitive function (H₁₄), and perceived health (H₁₅) had significant direct and indirect effects on healthy aging. Daily activity (H₁₆) had a statistically significant direct effect on healthy aging. While most hypotheses were accepted to explain healthy aging, hypothesis 17 was partially accepted since social support (H₁₇) showed a significant effect on healthy aging only indirectly through daily activity. In addition, physical function (H₁₈) had a significant indirect effect on healthy aging through daily activity, accepting the hypothesis (H₁₈).

3.3.5. Controlling for demographic variables

The demographic variables of age, sex (ref=male), economic status, and education years were included in the hypothesized model as confounding variables. The hypothesized model was tested after controlling for the demographic variables (Fig. 1-B). Among confounding variables, economic status was a statistically significant influencing factor in healthy aging (β = 0.14, p = 0.016).

4. Discussion

We examined the explanatory model of healthy aging based on the assumptions derived from the multidimensional model of healthy aging and aging-related resilience theory. Our basic assumption was that daily activity was the key factor associated with healthy aging in this population. We found that intrinsic factors (physical and cognitive functions) and mediators (resilience and social support) were significantly associated with daily activity engagement, which may play an important role in supporting healthy aging of frail older adults living in the community.

4.1. Physical and cognitive functions for healthy aging

The physical and cognitive functions were included in the model as intrinsic factors contributing to healthy aging. In the proposed model, the healthy aging of frail older adults could be led by promoting and maintaining the functional capacity to engage in daily activities. Some changes in cognitive tasks are normal with aging, but preserving cognitive function is crucial for maintaining independence in daily life and promoting healthy aging. Cognitive function directly impacts an individual's ability to manage daily activities, maintain social interactions, and enhance well-being in older adults (Kelly et al., 2017). This association was supported by the study findings that indicated that cognitive function had significant direct and indirect effects on all other influencing factors, as well as healthy aging. Physical function, on the other hand, demonstrated significant direct effects on daily activity and healthy aging and indirect effects on healthy aging through daily activity in our study. Previous researchers have supported the role of physical function for performing daily tasks independently, preventing mobility disability, and sustaining independence in older adults (DiPietro, 2019).

4.2. Daily activity for healthy aging

In this study, the concept of active aging referred to maintaining functional independence in engaging in daily activities to achieve a good quality of life. We found that all of the influencing factors had direct effects on daily activity and significant indirect effects on healthy aging through daily activities. A crucial part of healthy aging is maintaining independence in daily activities, which allows older adults to remain active and engaged in activities that provide them with a sense of meaning and fulfillment (Taylor et al., 2023). Moreover, independence fosters a sense of individual identity and self-worth, which are important factors in healthy aging (Bar-Tur, 2021). While physical function and social support were not directly associated with healthy aging, we found significant indirect effects on healthy aging through daily activity.

4.3. Role of resilience and social support as mediators

In the proposed model, resilience and social support were included as mediators to healthy aging. Based on the study findings, we supported the mediating effects of resilience and social support between cognitive function and daily activity, as well as healthy aging. Resilient individuals maximize and recover function in response to stressors, injury, or age-related physical decline (Peters, 2020). Being resilient would be especially useful for frail older adults when they wish to adapt to the decline in their abilities due to frailty while maintaining independence in daily life and quality of life. Many frail older adults maintain healthy aging despite functional limitations by utilizing specific domains in which they can still function (Peeters et al., 2023).

A relationship between resilience and social support has also been documented in the literature. Those with high resilience would seek assistance to overcome the crisis (Feliciano et al., 2022), and social connections provide emotional and practical resources for coping with challenges more effectively and maintaining independence in daily life (Li et al., 2021). Researchers have shown that social support provides health benefits for older adults. According to the United States Centers for Disease Control and Prevention, social isolation poses a significant health risks to this population (Bruss et al., 2024), emphasizing the need for social and emotional support for mental health and well-being. In a cross-sectional study with 1280 community-residing older adults, researchers found that social connectedness through providing and receiving social support was closely associated with well-being in this population (Zanjari et al., 2022). Cognitive function has been clearly associated with activity engagement and healthy aging, and social support was the key factor in connecting this relationship (Hwang et al., 2018).

From the results of this study, however, we indicated that physical function was not associated with resilience. The relationship between physical function and resilience in aging is complex and multifaceted. While some researchers have reported that resilience leads to better physical outcomes, others have found no significant association, likely due to variations in the operational definitions or measurement methods for resilience (Peters, 2020). In this study, a resilience scale was specifically developed to reflect the cultural context of Korean frail older adults. However, further evaluation of this scale is needed due to its low reliability. Future researchers should aim to refine the definition of resilience among frail older adults with diverse health conditions and investigate its relationship with key factors influencing healthy aging.

5. Limitations

The findings of this study should be interpreted with caution due to the cross-sectional design. Although we employed structural equation modeling, which allows for the examination of complex relationships, including direct and indirect effects among multiple variables while managing measurement error (Beran and Violato, 2010), these results should not be understood as establishing causality. While significant influencing factors and mediators were identified within the proposed model, further experimental studies are needed to confirm whether these factors can promote healthy aging in this population.

The study participants were community-residing older adults who were not restricted in their ability to engage in activities of daily living. Individuals with cognitive impairments that hindered their ability to understand and participate in face-to-face interviews were excluded. Consequently, the findings are not generalizable to individuals with severe cognitive impairment or limited capacity to perform daily activities.

We assessed active aging by evaluating participants' confidence in engaging in everyday activities. Drawing on the theory of active aging, "active" was defined not only as the ability to maintain physical activity but also as the capacity for continued participation in social, economic, cultural, spiritual, and civic dimensions of life (Rivadeneira et al., 2021). It is important to note, however, that participation in social or other activities varies, based on the socio-demographic characteristics of different populations, including variations in physical and cognitive abilities. For this reason, we opted to use a scale focused on daily activity—the most common form of activity for this population—while controlling for socio-demographic characteristics as confounding variables.

As a result, the findings may not fully generalize to a broader concept of active aging that encompasses a wider range of activities. For future studies, we should aim to refine the definition of active aging, particularly in terms of how frail older adults engage in diverse social, economic, or cultural activities of their choosing.

5.1. Implications for practice and research

Healthy aging is a central focus of health promotion for aging populations, especially among frail older adults. In this study, maintaining independence in daily activities was highlighted as a critical factor associated with healthy aging. Intrinsic factors such as perceived health, physical function, and cognitive function were found to influence engagement in daily activities, potentially through pathways involving resilience and social support. Various exercise programs tailored for frail older adults have the potential to enhance independence in daily activities, which may be linked to improved healthy aging outcomes in this population.

Resilience was identified as a key mediator in this study, showing a strong association with cognitive function. Resilience directly influenced engagement in daily activities and contributed to healthy aging, both directly and indirectly through its effects on daily activity. Intervention strategies could consider prioritizing resilience-enhancing approaches that incorporate social support as a core component to support healthy aging among frail older adults

6. Conclusion

We found that cognitive and physical function, along with resilience and social support as mediators, were associated with healthy aging in frail older adults through their engagement in daily activities. Health promotion strategies could consider emphasizing the development of resilience and fostering social support to empower individuals and strengthen their sense of social connectedness. These efforts may potentially support greater independence in daily life, which is associated with healthy aging. Further, we need to explore diverse social and cultural activities linked to healthy aging in frail older adults. Such studies could broaden the understanding and definition of active aging, offering new directions for tailored interventions for this population.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Funding

This work was supported by National Research Foundation of Korea (2022R1A2C2011502), and Chungnam National University. The funding bodies had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CRediT authorship contribution statement

Jisu Seo: Writing – review & editing, Writing – original draft, Methodology, Data curation, Conceptualization. Kyungok Joo: Writing – review & editing, Writing – original draft, Visualization, Software, Methodology, Formal analysis. Yuelin Li: Writing – review & editing, Writing – original draft, Software, Methodology, Investigation, Data curation. Nayoung Kim: Writing – review & editing, Writing – original draft, Validation, Resources, Methodology, Investigation. Eunna Oh: Writing – review & editing, Writing – original draft, Methodology, Investigation, Conceptualization. Lkhagvajav Gansukh: Writing – review & editing, Writing – original draft, Methodology, Investigation, Conceptualization. Rhayun Song: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.