Higher Sense of Coherence Is Associated With Lower Frailty Among Community-Dwelling Older Adults: A Cross-Sectional Mediation Study

Abstract

Introduction

This cross-sectional study examined the association between sense of coherence and frailty and explored the mediating roles of social support and self-rated health among community-dwelling older adults.

Methods

Data from 218 older adults were analyzed. Descriptive statistics were used to summarize participant characteristics and study variables, and structural equation modeling with maximum likelihood estimation was performed to test the hypothesized pathways. Indirect effects were evaluated using bias-corrected bootstrap 95% confidence intervals.

Results

Sense of coherence exerted a direct negative effect on frailty (β = −0.257, p = 0.005) and an indirect effect through self-rated health, as well as through social support and self-rated health in sequence (β = −0.136, p = 0.001). Although the direct path from social support to frailty was not statistically significant (p = 0.197), social support was indirectly associated with frailty through self-rated health. The final model showed moderate fit (χ²/df = 2.188, SRMR = 0.070, RMSEA = 0.074, CFI = 0.895, TLI = 0.858) and accounted for 35.8% of the variance in frailty.

Conclusion

These findings suggest that a higher sense of coherence is associated with a lower level of frailty, partly through psychosocial factors. Further longitudinal and interventional studies are warranted to examine whether interventions targeting sense of coherence, social support, and self-rated health may help mitigate frailty and promote healthy aging.

Keywords

sense of coherence frailty social support self-rated health older adults

1. Introduction

Frailty, a medical and geriatric condition frequently associated with aging, describes a state of heightened vulnerability to adverse health outcomes resulting from the dysregulation of multiple, complex physiological systems and diminished resilience in response to stressors.¹ Frail individuals face elevated risks of disability, falls, institutionalization, hospitalization, and mortality compared to their non-frail counterparts.^2,3 As a significant public health concern linked to global population aging,⁴ frailty has been reported to affect between 4% and 59% of older adults in previous studies.⁵ Its prevalence also increases with advancing age.⁶ Nevertheless, frailty is not static; an individual’s frailty status can fluctuate over time, potentially improving, remaining stable, or worsening.⁴ Consequently, identifying modifiable factors and underlying mechanisms associated with frailty progression is crucial for developing effective interventions for older adults.

The sense of coherence (SOC) is a core concept developed by sociologist Aaron Antonovsky in the late 1970s as part of his salutogenic model.⁷ It refers to an individual’s capacity to perceive and make sense of life experiences, particularly stressful or challenging situations, and is determined by the extent to which one views their internal and external environments as comprehensible, manageable, and meaningful.⁸ Within the salutogenic framework, Antonovsky proposed a wellness continuum ranging from complete health to severe illness.⁷ According to his theory, SOC serves as a critical salutary factor: individuals with a strong SOC are more likely to maintain or improve their health because they can better find meaning in adversity, cope effectively, and engage in health-promoting behaviors. Empirical studies have shown that a strong SOC is associated with lower levels of anxiety and depression, higher quality of life, and reduced physical functional decline among older adults.⁹

According to the salutogenic model, a strong SOC is associated with more effective stress-coping mechanisms, whereas stress generally compromises health status and increases vulnerability to somatic or psychosomatic conditions.¹⁰ Thus, an individual’s SOC may influence their vulnerability. Given that frailty is defined as a state of increased vulnerability to adverse health outcomes, it is reasonable to infer that SOC may exert a protective effect against frailty. However, research specifically examining the relationship between SOC and frailty remains limited. Considering that SOC is malleable and can be enhanced through appropriate interventions in older adults,¹¹ strategies aimed at improving SOC may represent a promising approach to mitigating frailty. Therefore, investigating the association between SOC and frailty is warranted.

Generalized Resistance Resources (GRRs) constitute another key concept in the salutogenic model. GRRs encompass internal and external factors and resources that individuals can draw upon to effectively manage stressors and maintain or improve their health and well-being.⁷ These resources play a vital role in strengthening an individual’s SOC and their capacity to navigate life’s challenges in a health-promoting manner.¹² GRRs manifest at three levels: individual, family, and community.¹² At the individual level, they include a relatively broader array of resources, such as emotional closeness and attachment relationships, personal characteristics (e.g., childhood living conditions, marital status, and self-rated health), social support, and genetic factors.¹² The availability and utilization of GRRs vary across individuals and communities. In Antonovsky’s model, SOC and GRRs share a reciprocal and dynamic relationship: sufficient GRRs and their successful application contribute to a strong SOC, while a high level of SOC, in turn, facilitates the mobilization of GRRs to enhance stress management.¹²

Research has shown that certain GRRs, such as social support and self-rated health, are closely associated with frailty among older adults. Social support plays a significant role in the development, prevention, and management of frailty. As Liu et al¹³ noted, interventions providing social support may mitigate frailty in older patients with hypertension and diabetes. Similarly, Anantapong et al⁶ found that each 1-point increase in the social support scale score was associated with a 1% reduction in the odds of developing pre-frailty or frailty. Poor self-rated health can serve as an early indicator of frailty; individuals rating their health as fair or poor are at higher risk of future frailty.¹⁴ In a study conducted among 603 older adults with cancer in the United States, self-rated health was identified as a viable screening tool for detecting frailty.¹⁴ Another study further demonstrated that poorer self-rated health was associated with concurrent frailty.¹⁵

Conversely, prior studies have also established that social support significantly influences self-rated health.^16,17 For instance, among 656 migrant older adults who relocated to new cities in China to live with their children, those residing with family were more likely to report higher self-rated health.¹⁶ Individuals receiving robust social support tend to perceive their health more positively, experience better mental health, and adopt healthier behaviors.

Moreover, evidence regarding the association between SOC and frailty, as well as the underlying psychosocial mechanisms among community-dwelling older adults, remains limited. Elucidating these pathways may inform the community-based identification of vulnerable older adults and guide future preventive strategies.

Based on theoretical and empirical evidence, we proposed that SOC was associated with frailty in older adults, and that social support and self-rated health acted as mediating variables. The following hypotheses were formulated and tested (Figure 1).

Hypothesis 1: SOC directly and negatively affects frailty among older adults.

Hypothesis 2: SOC has an indirect negative association with frailty mediated by social support among older adults.

Hypothesis 3: SOC has an indirect negative association with frailty mediated by self-rated health among older adults.

Hypothesis 4: Social support and self-rated health play a chain mediating role between SOC and frailty among older adults.

Figure 1.

Theoretical Model and Hypotheses

2. Methods

2.1. Study Design and Participants

This cross-sectional descriptive study was conducted in two communities within a district in Nanjing City, Mainland China. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for cross-sectional studies.¹⁸ Participants were recruited via convenience sampling. Inclusion criteria were: (1) age ≥ 60 years, and (2) willingness to participate. Older adults with cognitive impairment, as indicated by Mini-Mental State Examination (MMSE) scores (illiteracy ≤ 17, primary school ≤ 20, junior high school and above ≤ 24), were excluded.

A total of 247 community-dwelling older adults were initially enrolled. However, 21 were excluded due to cognitive impairment, and 8 questionnaires were incomplete. Thus, 218 participants were included in the final analysis, yielding a response rate of 88.3%. Sample size adequacy was evaluated for structural equation modeling (SEM). Common recommendations suggest a sample size of at least 200 or 5–10 times the number of observed variables to ensure stable parameter estimates and model fit.¹⁹ The model included 12 observed variables, implying a minimum requirement of 60–120 participants based on the variable-to-case ratio. The final sample of 218 participants satisfied both criteria, exceeding the recommended threshold of 200 and the upper bound of the variable-based requirement.

2.2. Instruments

2.2.1. Sense of Coherence Scale

The 13-item short-form Sense of Coherence Scale (SOC-13), developed by Antonovsky, was used to assess SOC.^7,8 It comprises three subscales: comprehensibility, manageability, and meaningfulness. Each item is rated on a 7-point Likert scale, yielding a total score ranging from 13 to 91. Five items are negatively worded and were reverse-scored. Higher scores indicate stronger SOC. Scores of 13–63 denote low SOC, 64–79 medium SOC, and 80–91 high SOC. The Chinese version of the SOC scale has been widely used in China, with a Cronbach’s α coefficient of 0.76.²⁰ Exploratory factor analysis supports the expected three-factor structure, explaining 47% of the variance (factor loadings: 0.41–0.69).²⁰ Criterion-related validity is evidenced by expected correlations with related constructs (e.g., negative correlations with stress and positive correlations with life satisfaction).²⁰ In this study, the SOC-13 demonstrated good internal consistency (Cronbach’s α = 0.849).

2.2.2. Fried’s Frailty Phenotype

Frailty was assessed using Fried’s frailty phenotype, which includes five criteria aligned with Chinese expert consensus guidelines.^2,21 Participants were classified based on the following cutoffs: (1) unintentional weight loss: >4.5 kg in the past year; (2) exhaustion: self-reported exhaustion, defined as a response of “3–4 days” or “most of the time” to either of two items from the Center for Epidemiologic Studies Depression Scale (e.g., “I felt that everything I did was an effort” or “I could not get going”) in the last week; (3) slow walking speed (4.57 m): ≥7 seconds for men ≤173 cm or women ≤159 cm, or ≥6 seconds for taller individuals; (4) low grip strength (dominant hand, average of three trials): ≤29–32 kg for men and ≤17–21 kg for women, stratified by BMI; and (5) low physical activity (International Physical Activity Questionnaire-Short Form): energy expenditure <383 kcal/week for men and <270 kcal/week for women. Frailty was defined as meeting ≥3 criteria, pre-frailty as 1–2 criteria, and non-frailty as none. The Fried phenotype has demonstrated concurrent and predictive validity, independently predicting adverse outcomes such as incident falls, worsening mobility, disability, hospitalization, and mortality.² As it comprises heterogeneous criteria rather than homogeneous scale items, internal consistency measures (e.g., Cronbach’s α) are not applicable.

2.2.3. Social Support Scale

The Social Support Rating Scale (SSRS) was used to measure social support.²² It contains 10 items across three dimensions: subjective support (4 items), objective support (3 items), and utilization of social support (3 items). Subjective support reflects perceived emotional support, objective support refers to tangible assistance, and utilization indicates the degree to which individuals make use of available support. Total scores range from 12 to 66, with higher scores indicating greater support. Scores of 12–22, 23–44, and 45–66 correspond to low, medium, and high social support, respectively. The SSRS has demonstrated good reliability and validity in Chinese populations. In the original report, 2-month test-retest reliability for the total score was R = 0.92 (p < 0.01), with item-level coefficients ranging from 0.89 to 0.94.²² Predictive validity is supported by associations with mental and physical health outcomes.²² In this study, the SSRS showed acceptable internal consistency (Cronbach’s α = 0.807).

2.2.4. Self-Rated Health

Self-rated health was assessed using a single self-designed item: “How do you think about your health status?” Responses were rated on a 3-point Likert scale (1 = poor, 2 = normal, 3 = good). Single-item self-rated health measures are widely used and have demonstrated good construct and predictive validity in older adults.²³ As a single-item measure, internal consistency indices (e.g., Cronbach’s α) are not applicable.

2.3. Data Collection

Trained data collectors conducted face-to-face interviews with participants who provided written informed consent. After completing the questionnaire, participants received a small gift (approximately two USD) as appreciation. Sociodemographic information was collected via a structured questionnaire, including age (years), gender (male/female), educational level (illiterate/primary school/junior high school/senior high school or above), marital status (married/widowed/divorced/never married), economic status (monthly income: <2,000 RMB; 2,000–3,999 RMB; ≥4,000 RMB), and number of chronic illnesses (count of self-reported physician-diagnosed conditions). Data collection occurred from March 2025 to June 2025.

2.4. Statistical Analysis

Descriptive analyses were performed using SPSS for Windows version 24.0. Sociodemographic characteristics and variable scores were summarized using frequencies, percentages, means, and standard deviations as appropriate. To examine the mediating effects of social support and self-rated health on the relationship between SOC and frailty, structural equation modeling (SEM) was conducted using AMOS 23.0 with maximum likelihood estimation. Given that self-rated health was measured on a 3-point ordinal scale and several frailty indicators were binary or ordinal, bias-corrected bootstrap confidence intervals for indirect effects were used to reduce reliance on normality assumptions.

Model fit was evaluated using multiple indices: chi-square/degrees of freedom ratio (χ²/df), standardized root mean square residual (SRMR), root mean square error of approximation (RMSEA), goodness-of-fit index (GFI), adjusted goodness-of-fit index (AGFI), normed fit index (NFI), comparative fit index (CFI), and Tucker-Lewis Index (TLI). Following established guidelines, acceptable fit was indicated by χ²/df < 3.00, SRMR < 0.08, and RMSEA < 0.08, while incremental fit indices (e.g., CFI, TLI, NFI) ≥ 0.90 were considered acceptable and ≥ 0.95 indicative of good fit. These thresholds are commonly used and should be interpreted cautiously.^19,24

To estimate indirect effects and obtain bias-corrected 95% confidence intervals (BC 95% CI), we employed the bootstrap method with 1,000 bootstrap resamples. Initially, the bootstrap resampling factor (bootfactor) was set to 1, as conventionally recommended. However, estimation errors occurred under this setting, even though the model converged when either the “Standardized estimates” or “Perform bootstrap” option was disabled. Closer inspection revealed no missing data, outliers, or abnormal variance estimates; thus, the estimation failure was attributed to instability in resampling. This instability likely arose because some bootstrap resamples yielded improper solutions (e.g., non-positive variance estimates), which precluded computation of standardized bootstrap estimates. To enhance computational stability, the bootfactor was increased to 2, a strategy adopted in prior studies and deemed acceptable for models with moderate sample sizes and elevated structural complexity.²⁴ As a sensitivity analysis, we re-estimated the model using frailty as an observed summed Fried phenotype score (0–5) rather than a latent factor and evaluated indirect effects using bias-corrected bootstrap confidence intervals. As an additional robustness check, we re-estimated the primary latent frailty model while adjusting for age, gender, educational level, economic status, number of chronic illnesses, and marital status (dummy-coded), and assessed indirect effects using 1,000 bias-corrected bootstrap resamples.

2.5. Ethical Considerations

The study protocol was approved by the human ethics research committee of Nanjing University of Chinese Medicine (Ethics Approval No. 2025NZY-5-01). Permission was also obtained from the community committees. Older adults were informed about the study aims and procedures and assured of the confidentiality of all data.

3. Results

3.1. Sociodemographic Characteristics of the Participants

The sociodemographic characteristics of the older adults are presented in Table 1. The mean age was 71.25 years (SD = 7.20), ranging from 60 to 101 years. Among the participants, 110 (50.5%) were female and 176 (80.7%) were currently married. A total of 73 (33.5%) participants had no formal education, and 87 (39.9%) had completed primary education. One hundred and thirty-four (61.5%) participants reported monthly incomes below 2,000 RMB.

Table 1.

Sociodemographic Characteristics of the Participants (n=218)

Variables		n	%
Age		71.25 (Mean)	7.20 (SD)
Gender	Male	108	49.5
Gender	Female	110	50.5
Educational level	Illiterate	73	33.5
	Primary school	87	39.9
	Junior high school	32	14.7
	≥ Senior high school	26	11.9
Marital status	Married	176	80.7
	Widowed	40	18.3
	Divorced	2	0.9
Monthly income (RMB)	< 2,000	134	61.5
	2,000∼3,999	46	21.1
	≥ 4,000	38	17.4
Number of chronic illnesses	0	39	17.9
	1	65	29.8
	≥ 2	114	52.3

3.2. Scores of Different Measuring Variables

The mean total scores for SOC and social support were 73.52 (SD = 10.20) and 40.92 (SD = 8.19), respectively, both at moderate levels. The mean self-rated health score was 2.25 (SD = 0.63). Specifically, 77 (35.3%) older adults rated their health as good, 118 (54.1%) as moderate, and 23 (10.6%) as poor. The mean frailty score was 0.88 (SD = 0.96); 91 (41.7%) participants exhibited no frailty, 113 (51.8%) were pre-frail, and 14 (6.4%) were frail. Detailed descriptive statistics for SOC, social support, self-rated health, and frailty are provided in Table 2.

Table 2.

The Scores of SOC, Social Support, Self-Rated Health, and Frailty

Variables	Mean	SD
SOC	73.52	10.20
Comprehensibility	28.40	4.42
Manageability	24.26	3.84
Meaningfulness	20.85	3.49
Social support	40.92	8.19
Subjective support	23.72	4.81
Objective support	10.51	2.93
Utilization of social support	6.68	2.36
Self-rated health (1–3)	2.25	0.63
Frailty (0–5)	0.88	0.96

Note. SOC = sense of coherence.

3.3. Study Hypothesis Verification

Among the direct paths in the primary model, the path from social support to frailty was not statistically significant (p > 0.05; Table 3). Nevertheless, social support exerted a statistically significant indirect effect on frailty through self-rated health, resulting in a significant total effect (Table 4). Despite its non-significance, this direct path was retained in the final model due to its theoretical relevance, as supported by prior literature. The final model is depicted in Figure 2.

Table 3.

Structural Equation Model Direct Path Estimates (Unstandardized)

Direct path (from → to)	B	SE	CR	p-value
SOC → Social support	0.077	0.032	2.444	0.015
Social support → Self-rated health	0.110	0.046	2.406	0.016
SOC → Self-rated health	0.042	0.014	2.919	0.004
SOC → Frailty	−0.009	0.004	−2.363	0.018
Self-rated health → Frailty	−0.072	0.020	−3.586	<.001
Social support → Frailty	−0.014	0.011	−1.291	0.197

Note. SE indicates standard errors of unstandardized coefficients (B). SOC = sense of coherence.

Table 4.

Total, Direct, and Indirect Effects in the Structural Model

Effect (from → to)	β	Bootstrap SE	BC 95% CI
Effect (from → to)	β	Bootstrap SE	Lower	Upper
Total effects
Self-rated health → Frailty	−0.414	0.095	−0.543	−0.284
Social support → Frailty	−0.227	0.133	−0.452	−0.069
SOC → Frailty	−0.393	0.117	−0.554	−0.222
Direct effects
SOC → Social support	0.214	0.117	0.036	0.361
SOC → Self-rated health	0.210	0.079	0.102	0.312
SOC → Frailty	−0.257	0.119	−0.416	−0.082
Social support → Self-rated health	0.197	0.071	0.096	0.291
Social support → Frailty	−0.145	0.133	−0.366	0.012
Self-rated health → Frailty	−0.414	0.095	−0.543	−0.284
Indirect effects
SOC → Frailty	−0.136	0.046	−0.215	−0.083
Social support → Frailty	−0.082	0.035	−0.138	−0.039

Note. Total effect = Direct effect + Indirect effect(s). SOC = sense of coherence.

Figure 2.

Final Model with Standardized Path Coefficients

The model demonstrated moderate fit to the data: χ² (df = 49, n = 218) = 107.201, p < 0.001, χ²/df = 2.188, SRMR = 0.070, RMSEA = 0.074, GFI = 0.928, AGFI = 0.886, NFI = 0.827, CFI = 0.895, and TLI = 0.858. The model accounted for 35.8% of the variance in frailty (R² = 0.358).

According to the overall mediation model (Figure 2), SOC exerted a direct negative effect on frailty (β = −0.257, p = 0.005), supporting hypothesis 1. SOC also demonstrated a significant indirect effect on frailty (β = −0.136, p = 0.001) mediated by self-rated health, as well as by social support and self-rated health in sequence, thereby supporting hypotheses 3 and 4. Although the direct path from social support to frailty was retained for theoretical reasons, it was not statistically significant (p = 0.197); thus, hypothesis 2 was not supported. Table 4 presents the total, direct, and indirect standardized effects, while Table 4 complements it by reporting the corresponding unstandardized direct path estimates (B), standard errors (SE), critical ratios (CR), and p-values. Additionally, Table 5 reports the measurement model loadings, including both standardized and unstandardized estimates, along with standard errors and critical ratios for each observed variable. Notably, two frailty indicators exhibited weak standardized loadings (walk time β = 0.163; weight loss β = 0.139; Table 5).

Table 5.

Measurement Model Loadings

Latent variable	Observed variable	Unstandardized loading	SE	CR	Standardized loading
SOC	Manageability	1.0			0.821
SOC	Comprehensibility	1.280	0.105	12.228	0.913
SOC	Meaningfulness	0.723	0.072	9.989	0.653
Social support	Utilization of social support	1.0			0.482
Social support	Subjective support	3.120	0.538	5.795	0.737
Social support	Objective support	2.024	0.357	5.676	0.785
Frailty	Exhaustion	1.0			0.463
Frailty	Physical activity	1.440	0.370	3.896	0.505
Frailty	Grip strength	1.941	0.513	3.783	0.467
Frailty	Walk time	0.707	0.403	1.754	0.163
Frailty	Weight loss	0.320	0.211	1.516	0.139

Note. SOC = sense of coherence.

In the sensitivity analysis treating frailty as an observed 0–5 score, the key indirect effects remained statistically significant based on bias-corrected bootstrap confidence intervals (SOC → frailty: standardized indirect effect, β = −0.104, BC 95% CI −0.163 to −0.061; social support → frailty through self-rated health: standardized indirect effect, β = −0.052, BC 95% CI −0.091 to −0.026; Supplementary Table S1). The total effects of SOC and social support on frailty also remained significant (SOC: standardized total effect, β = −0.199, BC 95% CI −0.310 to −0.096; social support: standardized total effect, β = −0.224, BC 95% CI −0.323 to −0.108), supporting the robustness of the key mediation findings to an alternative operationalization of frailty.

In an additional covariate-adjusted robustness check (adjusting for age, gender, educational level, economic status, number of chronic illnesses, and marital status dummy variables), the unstandardized indirect effect of social support on frailty through self-rated health remained statistically significant (B = −0.009, BC 95% CI −0.028 to −0.001). The unstandardized indirect effect of SOC on frailty was attenuated after covariate adjustment (B = −0.002, BC 95% CI −0.007 to 0.000).

4. Discussion

This study examined the relationships among sense of coherence (SOC), frailty, social support, and self-rated health in community-dwelling older adults. Grounded in Antonovsky’s salutogenic model and supported by empirical evidence from prior research, a comprehensive hypothesis model was developed and tested to elucidate the specific effects among these four variables.

The study found that 51.8% of older adults were pre-frail and 6.4% were frail. These prevalence rates align closely with previous Chinese studies using the same frailty measurement (51.2% pre-frail, 7.0% frail).²⁵ Any discrepancies may reflect differences in age distribution and participant characteristics across studies. As noted earlier, frailty prevalence increases with advancing age.⁶ Prior reports based on the China Health and Retirement Longitudinal Study (CHARLS) analyzed data from 2011, 2013, and 2015 and found that the prevalence rates of frailty among the same group of Chinese older adults were 18.7% (in 2011), 20.6% (in 2013), and 28.4% (in 2015), respectively, indicating that frailty became more common with age.²⁶

In the final model, SOC, social support, and self-rated health were all directly or indirectly associated with frailty. First, SOC was directly and positively associated with both social support and self-rated health, consistent with prior findings.²⁷ According to the salutogenic model, social support and self-rated health constitute key generalized resistance resources (GRRs). Individuals with higher SOC are more likely to perceive social support as a valuable resource for problem-solving, stress management, and well-being maintenance. When confronted with stressors, they tend to seek and utilize social support more actively and effectively. Moreover, a strong SOC is linked to more positive emotions and an optimistic life outlook,⁷ which can enhance self-rated health.

Second, frailty was associated with self-rated health. Self-rated health reflects an individual’s subjective assessment of their physical, mental, and social well-being. Nielsen²⁸ posits that individuals incorporate both current health status and anticipated future outcomes when rating their health. This perception can influence health-related behaviors and attitudes, thereby affecting frailty risk. Older adults who perceive themselves as healthy and wish to maintain that status may be more likely to adopt preventive health behaviors that delay or mitigate frailty.²⁹

In contrast to some prior studies,^17,30 social support did not exhibit a direct effect on frailty in our analysis. This discrepancy may be because previous studies did not include self-rated health in the pathway between social support and frailty, thereby obscuring its potential mediating role. As a known predictor of self-rated health,³¹ social support was indirectly associated with frailty through this mediator.

Although the direct path from social support to frailty was not statistically significant, it was retained in the final model based on theoretical considerations and empirical support from prior literature. This decision preserves the conceptual integrity of the hypothesized model and facilitates a more comprehensive examination of mediating mechanisms.

Furthermore, SOC was directly and indirectly associated with frailty, with the indirect effects operating through social support and self-rated health. After covariate adjustment, however, the total indirect effect of SOC on frailty was attenuated and the bias-corrected 95% confidence interval approached zero, suggesting that part of this indirect association may overlap with demographic and health-related factors, such as age and chronic illness burden, thereby reducing the independent contribution of this pathway. Within the salutogenic framework, SOC enhances individuals’ capacity to manage declines in intrinsic capacity.³² Older adults with higher SOC often possess greater psychological resilience and employ more adaptive coping strategies, which may bolster their ability to navigate stressors contributing to frailty.⁵ Additionally, a strong SOC is frequently accompanied by a heightened sense of meaning and purpose in life,³³ which may motivate engagement in meaningful activities, foster positivity, and reduce risks of depression and anxiety, conditions associated with frailty.⁹ Individuals with higher SOC are also more likely to adopt health-promoting behaviors, such as regular exercise and balanced nutrition, which are associated with reduced frailty and improved overall health.³⁴

These findings suggest that enhancing SOC, social support, and self-rated health may serve as effective strategies to mitigate frailty in older adults. Interventions tailored to older adults, such as dialogues centered on aging experiences, health education courses covering physical and mental health topics, and activities designed to strengthen social capital and life meaning, could improve SOC.³⁵ Encouraging group physical activities, mutual visits, and shared social engagements (e.g., games, interest-based interactions, or cognitive training) may also bolster self-rated health and social support.³⁶ Successful implementation will require collaboration among healthcare professionals, family members, home care aides, and broader support networks.

5. Limitations

This study has several limitations. First, participants were recruited exclusively from an urban setting, limiting generalizability; findings should be interpreted with caution. Second, reliance on self-reported instruments may introduce response bias, as participants might tailor answers to perceived researcher expectations rather than reflecting their true circumstances. Third, the cross-sectional design precludes causal inference among the study variables. Fourth, the use of maximum likelihood (ML) estimation with a single-item, 3-point ordinal self-rated health measure and binary/ordinal frailty indicators may be suboptimal and could affect parameter estimates; results should therefore be interpreted cautiously. Future studies should consider estimators specifically designed for categorical outcomes (e.g., WLSMV or robust estimators). Fifth, two frailty indicators (walk time and weight loss) exhibited weak standardized loadings in the measurement model, suggesting the latent frailty construct may not be equally represented by all five phenotype indicators; thus, findings involving the latent frailty factor warrant cautious interpretation. Nevertheless, a sensitivity analysis using the observed summed frailty score (0–5) yielded consistent conclusions regarding key indirect and total effects. Although we conducted a covariate-adjusted robustness check, residual confounding may persist due to unmeasured or imperfectly measured covariates.

6. Conclusion

A higher level of sense of coherence (SOC) is associated with lower frailty among older adults, partly through social support and self-rated health. Interventions such as intergenerational or peer dialogues, comprehensive health education programs addressing physical and mental well-being, and promotion of group physical and social activities may be beneficial. Further research is needed to design, implement, and evaluate interventions aimed at enhancing SOC, social support, and self-rated health to reduce frailty in older adult populations.

Supplemental Material

Supplemental material - Higher Sense of Coherence Is Associated With Lower Frailty Among Community-Dwelling Older Adults: A Cross-Sectional Mediation Study

Supplemental Material for Higher Sense of Coherence Is Associated With Lower Frailty Among Community-Dwelling Older Adults: A Cross-Sectional Mediation Study by Libin Gu, Mingming Yu, Qiuling Wang and Zhiling Sun in INQUIRY: The Journal of Health Care Organization, Provision, and Financing.

Footnotes

Acknowledgments

The authors acknowledge the support and coordination of the staffs from the two community committees as well as all the participants in the study.

ORCID iDs

Libin Gu

Mingming Yu

Zhiling Sun

Qiuling Wang

Ethical Considerations

The study protocol was approved by the human ethics research committee of Nanjing University of Chinese Medicine (Ethics Approval No. 2025NZY-5-01). Permission was also obtained from the community committees. Older adults were informed about the study aims and process and assured of the confidentiality of all data. Written informed consent was obtained from all participants prior to participation.

Author Contribution

Libin Gu: Conceptualization, Methodology, Project Administration, Resources, Investigation, Formal Analysis, Writing-Original Draft, Writing-Review and Editing. Mingming Yu: Methodology, Supervision, Data Curation, Writing-Original Draft, Writing-Review and Editing. Qiuling Wang: Project Administration, Investigation, Data Curation, Formal Analysis. Zhiling Sun: Conceptualization, Methodology, Supervision, Resources, Writing-Original Draft, Writing-Review and Editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Philosophy and Social Science Research Project of Jiangsu Universities, China (Grant number: 2022SJYB0312); the Open Project of the Nursing Discipline at Nanjing University of Chinese Medicine under the Priority Academic Program Development of Jiangsu Higher Education Institutions (Phase IV) (Grant number: YSHL202533); and the Key Discipline of Traditional Chinese Medicine Nursing, National Administration of Traditional Chinese Medicine. The study funder was not involved in the study design, execution, data analysis, and writing for publication.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

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

Supplemental Material

Supplemental material for this article is available online.

References

Roe

Normand

Wren

Browne

O’Halloran

. The impact of frailty on healthcare utilisation in Ireland: evidence from the Irish longitudinal study on ageing. BMC Geriatr. 2017;17(1):203. doi:10.1186/s12877-017-0579-0.

Fried

Tangen

Walston

, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146-M156. doi:10.1093/gerona/56.3.m146.

Liu

Qin

Qiu

, et al. Multidimensional frailty as a predictor of mortality among older adults: a systematic review and meta-analysis. BMC Geriatr. 2024;24:793. doi:10.1186/s12877-024-05377-4.

Kojima

Taniguchi

Iliffe

Jivraj

Walters

. Transitions between frailty states among community-dwelling older people: a systematic review and meta-analysis. Ageing Res Rev. 2019;50:81-88. doi:10.1016/j.arr.2019.01.010.

Rohrmann

. Epidemiology of frailty in older people. Adv Exp Med Biol. 2020;1216:21-27. doi:10.1007/978-3-030-33330-0_3.

Anantapong

Wiwattanaworaset

Sriplung

. Association between social support and frailty among older people with depressive disorders. Clin Gerontol. 2020;43(4):400-410. doi:10.1080/07317115.2020.1728002.

Antonovsky

. Unraveling the Mystery of Health: How People Manage Stress and Stay Well. San Francisco, CA: Jossey-Bass; 1987.

Antonovsky

. The structure and properties of the sense of coherence scale. Soc Sci Med. 1993;36(6):725-733. doi:10.1016/0277-9536(93)90033-Z.

Qian

Deng

. The association between the sense of coherence and health outcomes among community-dwelling older adults: a systematic review. Interdiscip Nurs Res. 2023;2(1):46-57. doi:10.1097/NR9.0000000000000016.

10.

Betke

Basińska

Andruszkiewicz

. Nurses sense of coherence and stress management strategies against the types of health status. Nurs Open. 2021;8(6):3403-3410. doi:10.1002/nop2.886.

11.

Liu

Zhang

Montayre

Koduah

Leung

AYM

. Non-pharmacological interventions targeting sense of coherence among older adults and adults with chronic conditions: a meta-analysis with trial sequential analysis. J Adv Nurs. 2025;81(4):2165-2198. doi:10.1111/jan.16558.

12.

Idan

Eriksson

Al-Yagon

. Generalized Resistance Resources in the Salutogenic Model of Health. In: Mittelmark

Bauer

Vaandrager

, eds. The Handbook of Salutogenesis. 2nd ed. Cham, Switzerland: Springer; 2022::93–106. doi:10.1007/978-3-030-79515-3_12

13.

Liu

Meng

Liu

. The relationship between health literacy, social support, depression, and frailty among community-dwelling older patients with hypertension and diabetes in China. Front Public Health. 2020;8:280. doi:10.3389/fpubh.2020.00280.

14.

Giri

Mir

Al-Obaidi

, et al. Use of single-item self-rated health measure to identify frailty and geriatric assessment–identified impairments among older adults with cancer. Oncologist. 2022;27(1):e45-e52. doi:10.1093/oncolo/oyab020.

15.

Pilleron

Le Goff

Ajana

, et al. Self-rated health and frailty in older adults from the population-based Three-City Bordeaux cohort. Gerontology. 2022;68(7):755-762. doi:10.1159/000518864.

16.

Wei

Kong

. The effects of social support and morbidities on self-rated health among migrant elderly following children to Jinan, China. Healthcare (Basel). 2021;9(6):686. doi:10.3390/healthcare9060686.

17.

Yang

Zhang

Meng

Liu

Sun

. Mediating effect of social support on the associations between health literacy, productive aging, and self-rated health among elderly Chinese adults in a newly urbanized community. Medicine (Baltimore). 2019;98(16):e15162. doi:10.1097/MD.0000000000015162.

18.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med. 2007;4(10):e296. doi:10.1371/journal.pmed.0040296.

19.

Kline

. Principles and Practice of Structural Equation Modeling. 5th ed. New York, NY: Guilford Press; 2023.

20.

Bao

Liu

. The reliability and validity of Chinese version of SOC-13 [in Chinese]. Chin J Clin Psychol. 2005;13(4):399-401. doi:10.3969/j.issn.1005-3611.2005.04.007.

21.

Hao

Dong

, et al. Chinese expert consensus on assessment and intervention for elderly patients with frailty [in Chinese]. Chin J Geriatr. 2017;36(3):251-256. doi:10.3760/cma.j.issn.0254-9026.2017.03.007.

22.

Xiao

. Theoretical basis and research application of the Social Support Rating Scale [in Chinese]. Lin Chuang Jing Shen Yi Xue Za Zhi. 1994;4(2):98-100.

23.

Takahashi

Tanno

Yonekura

, et al. Poor self-rated health predicts the incidence of functional disability in elderly community dwellers in Japan: a prospective cohort study. BMC Geriatr. 2020;20:328. doi:10.1186/s12877-020-01743-0.

24.

Byrne

. Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming. 3rd ed. New York, NY: Routledge; 2016.

25.

Smit

Xue

Odden

. Prevalence and correlates of frailty among community-dwelling Chinese older adults: the China Health and Retirement Longitudinal Study. J Gerontol A Biol Sci Med Sci. 2017;73(1):102-108. doi:10.1093/gerona/glx098.

26.

Chao

Che

Wong

. Supportive environments moderate the relationship between frailty and life satisfaction in older adults: evidence from the China Health and Retirement Longitudinal Study (CHARLS). BMC Geriatr. 2025;25(1):1004. doi:10.1186/s12877-025-06696-w.

27.

Norlin

Faresjö

Falk

Jones

Walter

. Primary healthcare utilisation and self-rated health among patients with irritable bowel syndrome: what are the impacts of comorbidities, gastrointestinal symptom burden, sense of coherence and stress. J Psychosom Res. 2019;119:1-7. doi:10.1016/j.jpsychores.2019.01.015.

28.

Nielsen

. The relationship between self-rated health and hospital records. Health Econ. 2016;25(4):497-512. doi:10.1002/hec.3167.

29.

Calvey

McHugh Power

Maguire

. Expecting the best or fearing the worst: discrepancies between self-rated health and frailty in an ageing Irish population. Br J Health Psychol. 2022;27(3):971-989. doi:10.1111/bjhp.12585.

30.

Ermer

Proulx

. Associations between social connectedness, emotional well-being, and self-rated health among older adults: difference by relationship status. Res Aging. 2019;41(4):336-361. doi:10.1177/0164027518815260.

31.

Mitchell

Kowalsky

Christian

Belury

Cole

. Perceived social support predicts self-reported and objective health and health behaviors among pregnant women. J Behav Med. 2022;45(4):589-602. doi:10.1007/s10865-022-00306-5.

32.

Chen

, et al. Association between sense of coherence and frailty: a cross-sectional study in China. Front Psychiatry. 2022;13:844439. doi:10.3389/fpsyt.2022.844439.

33.

Winger

Adams

Mosher

. Relations of meaning in life and sense of coherence to distress in cancer patients: a meta-analysis. Psychooncology. 2016;25(1):2-10. doi:10.1002/pon.3798.

34.

Riera-Sampol

Bennasar-Veny

Tauler

Nafría

Colom

Aguilo

. Association between depression, lifestyles, sleep quality and sense of coherence in a population with cardiovascular risk. Nutrients. 2021;13(2):585. doi:10.3390/nu13020585.

35.

Guo

Deng

. Interventions based on salutogenesis for older adults: an integrative review. J Clin Nurs. 2024;33(7):2456-2475. doi:10.1111/jocn.17028.

36.

Koren

Leveille

You

. Tai Chi interventions promoting social support and interaction among older adults: a systematic review. Res Gerontol Nurs. 2021;14(3):126-137. doi:10.3928/19404921-20210325-02.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB