Introduction

Chronic obstructive pulmonary disease is a prevalent chronic condition among the elderly, characterised by high morbidity and mortality rates worldwide1. It accelerates lung damage and ageing through abnormal repair mechanisms driven by oxidative stress2. Respiratory dysfunction is closely linked to frailty, therefore, elderly patients with COPD are more likely to develop frailty3. Studies have reported an average frailty prevalence of 47.0% among hospitalised elderly patients with COPD in other countries4, while in hospitalised elderly individuals in China, the prevalence is 25.0%5. Frailty is further associated with advanced age, muscle loss, the coexistence of multiple chronic conditions, and a decline in the ability to perform activities of daily living6,7,8. When frailty and COPD coexist, they exacerbate negative outcomes such as sarcopenia, falls, disability, hospitalisation, and mortality9. However, most previous studies have not fully elucidated the interactions among high-risk factors for frailty, the relative strength of their effects, or the presence of direct or indirect mediating factors in hospitalised elderly patients with COPD.

The cycle of frailty theory10 and the integrated conceptual model of frailty11 are widely accepted frameworks for explaining frailty. The cycle of frailty theory describes a vicious cycle in which the coexistence of multiple diseases increases energy expenditure, leading to decreased muscle strength and reduced walking speeds12. Frailty is recognised as a dynamic and continuous process influenced by physical, psychological, and social factors13. Potential indirect effects may arise along pathways involving decreased muscle strength and the coexistence of multiple chronic diseases, two major geriatric syndromes that are strongly associated with the onset of frailty and exhibit bidirectional relationships. Recent evidence suggests that elderly patients with COPD with below-normal muscle strength are significantly more likely to develop frailty14. Additionally, the progressive decline in muscle mass, bone density, and cardiorespiratory fitness in elderly individuals substantially increases frailty risk15.

The simultaneous occurrence of muscle strength decline and multimorbidity is common in late life, particularly in population-based studies. Most patients with COPD have at least one comorbid condition16, and those with multimorbidity experience significantly higher frailty and mortality rates than those without17. Sarcopenia is also prevalent among older COPD patients, with reported rates of 24% in Asia18, and 15% in Europe19. Both low muscle strength and multimorbidity are associated with adverse health outcomes, including falls, hospitalisation, institutionalisation, and death20. These conditions also contribute to poor quality of life, often reflected in impaired CAT scores21,22,23. Their combined effects exacerbate these adverse outcomes, highlighting the need for further research into their shared pathological mechanisms, potential interactions, and predictive roles. Such investigations could provide valuable insights into the prevention and management of frailty in elderly patients with COPD.

In summary, this study aims to investigate the interrelationships among key high-risk factors closely associated with frailty—namely grip strength, multimorbidity, CAT score, GOLD stage, and age—and to examine their impact on frailty in hospitalised elderly patients with COPD. The findings aim to provide a scientific basis for advancing clinical understanding of the factors influencing frailty.

Methods

Conceptual framework of the study

In this study, on the basis of the cycle of frailty theory and the integrated conceptual model of frailty, we integrated its aggregated factors—grip strength, multimorbidity, CAT score, GOLD stage and age to construct a theoretical hypothesis model. This model was used to analyse the interactions among these factors, as well as their direct and indirect effects on frailty in hospitalised elderly patients with COPD. In summary, this study proposes the following theoretical hypotheses (Fig. 1):

Hypothesis 1

Age directly affects frailty.

Hypothesis 2

Grip strength directly affects frailty.

Hypothesis 3

The CAT score directly affects frailty.

Hypothesis 4

Multimorbidity directly affects frailty.

Hypothesis 5

GOLD stage directly affects frailty.

Hypothesis 6

Age indirectly affects frailty through multimorbidity.

Hypothesis 7

Age indirectly affects frailty through grip strength.

Hypothesis 8

Grip strength indirectly affects frailty through multimorbidity.

Hypothesis 9

Multimorbidity indirectly affects frailty through the CAT score.

Fig. 1
figure 1

Theoretical hypothetical model.

Full size image

Study design and population

This study employed a cross-sectional design with convenience sampling to recruit participants. Data were collected from the First Affiliated Hospital of Chongqing Medical University, involving 283 patients between March and August 2024. Data were collected within 48 h after the attending physician issued the pre-discharge order, once patients met the hospital’s standardised discharge criteria for COPD, ensuring clinical stability and minimising symptom fluctuations during hospitalisation. Baseline data were collected using a general form. Frailty was assessed via the FRAIL scale, grip strength measured with the XiangShan EH101 device, CAT score used to evaluate disease impact, and GOLD stage determined by pulmonary function tests. Path and mediation analyses explored the relationships between these variables and frailty in hospitalised elderly COPD patients.

The study considered 11 parameters, with 20 observations for each parameter, yielding a minimum required sample size of 220. The final sample size included 283 cases, ensuring adequate statistical power24. Inclusion criteria were: (a) age ≥ 65 years; (b) diagnosis of COPD based on the GOLD guidelines (2024 edition); and (c) voluntary participation with informed consent. Exclusion criteria included: (a) speech disorders, cognitive impairment, or physical activity limitations (such as severe mobility restrictions, dependence on a wheelchair or bedridden status)25; and (b) inability to complete the FRAIL assessment.

All participants provided written informed consent. The study was conducted according to the Declaration of Helsinki and approved by the Ethics Committee of the Chongqing Medical University First Affiliated Hospital Clinical Research (No. 2024-224-01).

Measures

Demographic variables

A general data collection form, designed by the research team, was used to document demographic and clinical characteristics, including age, gender, marital status, education level, living arrangements, BMI (body mass index), lung function, CAT score, multimorbidity, grip strength, and frailty status. Patients provided personal demographic information, while clinical data were extracted from electronic medical records. Data collection was conducted in the ward by a postgraduate researcher and a clinical nurse, data collection forms were retrieved immediately after completion. All collected materials were strictly used for research purposes.

Chronic Obstructive Pulmonary Disease Assessment Test (CAT scale)

The CAT scale, introduced by Jones in 2009, quantifies the impact of COPD to facilitate health status assessment and improve communication between patients and clinicians14. This self-reported scale consists of eight items addressing cough, sputum production, chest tightness, dyspnoea, activity limitations, confidence in leaving the home, sleep quality, and energy levels. Each item is scored on a 0 to 5 scale, with a total score ranging from 0 to 40. Higher scores indicate greater disease impact. Assessment categories were as follows: ≤ 10 (mild impact), 11–20 (moderate impact), 21–30 (severe impact), and > 30 (very severe impact). The Chinese version of the CAT scale has demonstrated robust internal consistency and validity in the local population, with a Cronbach’s alpha coefficient of 0.80526.

Hand grip strength

Grip strength was assessed using a digital measuring device (XiangShan EH101, China). Patients were instructed to exert maximum grip strength with their dominant hand over 30 s to one minute per measurement. The highest value from three consecutive attempts was recorded for analysis27.

FRAIL scale

Frailty levels were evaluated using the FRAIL scale, a simple and practical self-assessment tool developed from the Fried frailty phenotype10. The scale comprises five components: fatigue, resistance, ambulation, illness, and weight loss. Each “yes” response scores 1 point, yielding a total score between 0 and 5. Scores are categorised as follows: 0 (robust), 1–2 (prefrail), and ≥ 3 (frail). The Chinese version of the FRAIL scale has shown satisfactory validity and reliability in community-dwelling older adults28.

GOLD stage

The GOLD stage are determined based on the patient’s pulmonary function test results. The classification is as follows: Grade 1 (Mild): FEV1 ≥ 80% of the predicted value; Grade 2 (Moderate ): 50% ≤ FEV1 ≤ 80% of the predicted value; Grade 3 (Severe): 30% ≤ FEV1 < 50% of the predicted value; Grade 4 (Very Severe): FEV1 < 30% of the predicted value29.

Statistical analyses

Statistical analyses were performed using IBM SPSS 26.0. Continuous variables conforming to a normal distribution were described as means and standard deviations (mean ± SD), whereas non-normally distributed variables were expressed as medians and interquartile ranges M(IQR). The Kolmogorov–Smirnov test was used to assess the normality of continuous variables. Pearson correlation analysis was used for variables with a normal distribution, while Spearman’s rank correlation was used for non-normally distributed variables to explore relationships between variables. Categorical variables were compared using the chi-square test, and continuous variables were compared using either the independent samples t-test or the Mann–Whitney U test according to their distribution. All statistical tests were two-tailed, and a p < 0.05 was considered statistically significant.

Path analysis was performed using IBM SPSS AMOS 26.0 to evaluate the hypothetical model, with maximum likelihood estimation used to estimate model parameters. Mediation models were applied to examine the mediating roles of key variables, and hierarchical regression analysis was employed to validate mediation effects.

Model fit was assessed using standard indices: χ2/df (< 3.0 acceptable), RMSEA (< 0.08 adequate; < 0.05 good) with 90% CI, CFI and TLI (≥ 0.90 acceptable; > 0.95 good), SRMR (< 0.08 acceptable), and GFI (> 0.90 acceptable). Model estimates were evaluated using critical ratios (CRs), p-values, and bootstrapping to test direct, indirect, and total effects. The study adhered to STROBE guidelines for cross-sectional research.

Results

Characteristics of the study population

A total of 283 participants were included in this study, with a mean age of 68 ± 6.46 years. There were 255 males (90.11%) and 28 females (9.89%). Among all the hospitalised elderly patients with COPD, the incidence of robust is 26.15% (74), prefrail is 39.93% (113), Frailty is 33.92% (96), respectively. (Table 1)

Table 1 The demographic and clinical characteristics of all participants (N = 283).
Full size table

Correlations between frailty variables

Spearman correlation analysis was employed to examine the correlations among variables related to frailty. The results revealed significant correlations among age, multimorbidity, the CAT score, grip strength, and the GOLD stage. Age, CAT score, multimorbidity, and GOLD stage all exhibited positive correlations with frailty, whereas grip strength demonstrated a negative correlation with frailty. Notably, the correlation with frailty was strongest for grip strength (Table 2).

Table 2 Correlation analysis between independent variables.
Full size table

Path analysis

In this study, the maximum likelihood method was used to estimate the parameter of the hypothetical model. The results of the hypothetical model employ standard estimates (β), standard errors (SEs), and critical ratios (CRs), which are the relative effect sizes for each pathway. The parameter estimation, SE, CR values and hypothesis testing of the model revealed that the eight paths were statistically significant (Fig. 2), indicating that the proposed hypotheses were fully supported and that the goodness of fit of the model was generally good: X2/df = 1.170, RMSEA = 0.027, RMSEA 90% CI = 0.024 to 0.085, CFI = 0.992, TLI = 0.982, SRMR = 0.051, GFI = 0.981, indicating that the model had good applicability to the sample data and analysis of this study.

Age (β = 0.173, p < 0.01), the CAT score (β = 0.221, p < 0.01), multimorbidity (β = 0.239, p < 0.01), GOLD stage (β = 0.161, p < 0.01) and grip strength (β = -0.352, p < 0.01) had direct impacts on frailty. Age, multimorbidity, GOLD stage, and CAT score were positively associated with frailty, indicating that increased age, a higher number of comorbidities, elevated CAT scores, and more advanced GOLD stage may exacerbate the severity of frailty. Grip strength has a negative effect on frailty, and the absolute value of the standardized path coefficient is the largest, indicating that grip strength has the greatest direct effect on frailty. (Table 3)

Fig. 2
figure 2

Standardized coefficients of path analyses model for factors influencing hospitalised elderly patients with COPD.

Full size image
Table 3 Parameter estimates of variables for the structural equation model (n = 283).
Full size table

Mediating effects analysis of factors influencing hospitalised elderly patients with COPD

The mediating effects analysis of the variables in the model are shown in Table 4. Age, multimorbidity, GOLD stage, and grip strength not only had direct effects on frailty but also exerted partial mediated effect on frailty. The total effects of age, multimorbidity status and grip strength on frailty were 0.098, 1.434 and − 0.108 respectively (P < 0.01). The mediating effect proportion of grip strength was 21.916%, multimorbidity was 17.106%, CAT score was 17.253%, multimorbidity was 12.846% and GOLD stage was 6.831%.

Table 4 The direct, indirect and total effects of variables in the structural equation model (n = 283).
Full size table

Hierarchical regression analysis of variables

The hierarchical regression results for age, multimorbidity, CAT score, GOLD stage, and grip strength in hospitalised elderly COPD patients are shown in Table 5. These variables jointly explained 46.9% of the variance in frailty. In Model 1, age alone explained 10.8% of the variance (R2 = 0.108). Adding multimorbidity in Model 2 increased R2 to 0.203 (∆R2 = 9.4%). In Model 3, the inclusion of CAT score raised R2 to 0.276 (∆R2 = 7.3%). Grip strength, added in Model 4, contributed the largest increase, raising R2 to 0.388 (∆R2 = 11.2%). Finally, adding GOLD stage in Model 5 increased R2 to 0.470, accounting for an additional 8.2% of the variance.

Table 5 Hierarchical regression analysis.
Full size table

Discussion

This study is the first to examine the independent and mediating effects of age, grip strength, CAT score, GOLD stage, and multimorbidity on frailty in hospitalised older adults with COPD. These factors jointly explained 46.9% of frailty variance, with grip strength having the strongest effect (11.2%) and the greatest negative direct impact (β = -0.352, p < 0.01). Although not included in the FRAIL scale, grip strength is a core component of Fried’s frailty phenotype and a validated marker of physical frailty10. Our findings move beyond identifying associations to quantifying explanatory power, underscoring grip strength’s central role in frailty. We also identified intricate interactions among risk factors. Age-related muscle loss contributes to grip strength decline and multimorbidity, both of which elevate frailty risk30,31,32. Low grip strength has been linked to poorer quality of life, multimorbidity, and higher mortality33,34. While age, multimorbidity, CAT score, and GOLD stage influence frailty, grip strength exerts a more substantial impact. Additionally, multimorbidity worsens respiratory symptoms, reflected in higher CAT scores, which are themselves strong predictors of frailty21,22.

For hospitalised elderly patients with COPD, disease severity is a critical factor influencing the development of frailty. These individuals are particularly vulnerable to frailty during exacerbations, which also contribute to elevated overall mortality rates5,35. GOLD stage directly affected frailty (β = 0.161, p < 0.01) and had a modest indirect effect via multimorbidity (6.831%, 95% CI: 0.004–0.066), suggesting additional latent mechanisms may be at play36,37. Despite controlling for key covariates, residual confounding cannot be ruled out. Longitudinal studies are needed to better delineate these pathways.

Our mediation analysis confirmed that grip strength, multimorbidity, GOLD stage and CAT score partially mediated the effects of risk factors on frailty, with grip strength showing the highest mediating proportion. These factors interact synergistically, and failure to manage them can significantly compromise quality of life and burden healthcare systems38,39, Fortunately, frailty is reversible40,41, and targeted interventions addressing modifiable risks can mitigate its impact.

Current frailty management often follows generic guidelines that lack stratification, limiting clinical applicability, particularly in patients with reduced exercise tolerance42,43. Our findings offer actionable insights into key frailty determinants—grip strength, CAT score, multimorbidity, GOLD stage, and age—supporting early risk identification and personalised intervention. Such strategies can reduce rehospitalisation, enhance life quality, and ease the burden on healthcare systems44,45, representing a meaningful advancement in the management of frailty in hospitalised elderly COPD patients.

Limitations and strengths

This study has several limitations. First, its cross-sectional design limits causal interpretations and cannot capture longitudinal changes in the variables associated with frailty. Second, the sample was predominantly male, with relatively few female participants. While this reflects the higher COPD prevalence among older men, the gender imbalance may introduce bias and limits generalisability to female patients. The small number of women also prevented gender-stratified analyses, potentially obscuring sex-specific associations with frailty.

To our knowledge, this is the first study to assess the relative contributions of age, grip strength, CAT score, multimorbidity, and GOLD stage to frailty in hospitalised elderly COPD patients. Our results offer valuable evidence to inform targeted frailty interventions aimed at improving clinical outcomes in this high-risk population.

Conclusion

This study demonstrates that hospitalised elderly patients with COPD often present with reduced grip strength, multimorbidity, elevated CAT scores, higher GOLD stages, and advanced age—factors significantly associated with frailty. Advancing age leads to decreased muscle strength and a higher risk of chronic diseases, thereby contributing to the development of frailty. Additionally, GOLD stage mediates the effect of multimorbidity, while multimorbidity mediates the effect of CAT score on frailty. Although path analysis confirmed these complex interrelationships, their temporal and causal dynamics require validation through longitudinal studies. Such research would support the design of targeted interventions to improve outcomes and ease the burden on healthcare systems.