Abstract
Objective
To investigate the relationships among economic toxicity, family function, self-management ability, and stroke-specific quality of life in stroke patients during the rehabilitation period, and to construct and test a moderated mediation model with self-management as the mediator and family function as the moderator, so as to provide empirical evidence for improving the quality of life of stroke patients.
Methods
A cross-sectional survey was conducted from May to June 2025 across 22 medical and elderly care institutions in Chongqing, China. A total of 557 stroke patients in the rehabilitation period completed questionnaires measuring sociodemographics, economic toxicity (COST-PROM), family function (Family APGAR Scale), self-management (Stroke Self-Management Scale), and stroke-specific QoL (SS-QOL Scale). The moderated mediation model was tested using the PROCESS macro in SAS 9.4.
Results
Economic toxicity was significantly negatively correlated with quality of life (r = -0.424, P < 0.001); self-management played a partial mediating role between economic toxicity and quality of life, with the mediating effect accounting for 50.56%; family function played a significant buffering moderating role between economic toxicity and self-management (β = 0.209, P < 0.001), and high family function could significantly weaken the negative impact of economic toxicity on self-management.
Conclusion
Economic toxicity impairs QoL both directly and indirectly through reduced self-management. Strong family function buffers the detrimental effect of economic toxicity on self-management. Interventions targeting economic toxicity screening, self-management support, and family empowerment may help improve QoL in stroke survivors and warrant prospective evaluation.
Keywords: Stroke, Economic toxicity, Self-management, Family function, Quality of life, Moderated mediation model
Introduction
Stroke is a major chronic neurological disease with high incidence, disability, and recurrence, imposing substantial economic burden worldwide. Data from the Global Burden of Disease Study 2019 showed that global stroke incidence, prevalence, and years lived with disability remained high from 1990 to 2019, concentrated in low- and middle-income countries with a rising proportion of young and middle-aged patients, presenting long-term challenges to health systems [1]. The recent GBD 2023 study further confirmed that stroke remains a core cause of healthy life loss globally [2].
With population aging and improved treatment technologies, direct costs including hospitalization, rehabilitation, and assistive devices, as well as indirect costs such as lost productivity and caregiver absence, have imposed severe economic burden [3]. At the individual level, this burden manifests as economic toxicity, a multidimensional construct encompassing not only tangible financial hardship but also persistent psychological distress. Describing disease-related financial strain and its harmful impacts on psychology, health behaviors, and prognosis, this construct has become an emerging research focus in stroke care. Economic toxicity impairs mental health, limits rehabilitation access, and weakens health management, thereby hindering functional recovery and quality of life (QoL) care [4, 5]. Despite its recognized importance, the underlying mechanism by which economic toxicity affects post-stroke QoL remains unclear. Health-related quality of life has become a core patient‑reported outcome in stroke rehabilitation. Compared with traditional physical function scores, it emphasizes subjective experience, social participation, and overall well-being, and is widely used to evaluate rehabilitation efficacy [6, 7]. Therefore, clarifying how economic toxicity translates into impaired rehabilitation behaviors is essential for understanding its effect on post-stroke QoL.
Given that economic toxicity may translate into impaired rehabilitation behaviors, self-management may represent a key behavioral pathway linking economic toxicity to post-stroke QoL. It is essential for long-term stroke rehabilitation, including symptom monitoring, medication adherence, rehabilitation persistence, healthy lifestyle, and stress coping, which directly determine prognosis and quality of life [8].A one-year follow-up systematic review indicated that self‑management and rehabilitation adherence are independent protective factors for long‑term quality of life, as good self-management mitigates the harm of functional impairment and facilitates social re-integration [8]. Economic stress may reduce health control perception, lower rehabilitation motivation, and restrict resources for self-management, while stable economic status helps maintain consistent self-management. Few studies have integrated economic toxicity and self-management into a single framework, and the mediating role of self-management between economic toxicity and quality of life remains empirically unexamined.
Beyond individual self-management, the family environment may further shape how patients cope with economic stress during stroke rehabilitation. Family function is the most important source of social support for stroke patients, including emotional communication, caregiving, and collaborative coping, which plays an irreplaceable protective role in chronic disease rehabilitation, captures relational and emotional dynamics within the family, reflects the quality of family support rather than economic resources or individual behaviors [9, 10]. In Asian culture, stroke care and rehabilitation strongly rely on family support, and family function directly affects psychological state, rehabilitation confidence, and medical decision-making [10]. Evidence indicates that family function improves health behaviors and quality of life by regulating resilience and coping styles, and a supportive family environment alleviates disease-related stress [9, 11].
The Stress Buffering Model provides a classic framework for understanding stress, support, and health outcomes. It states that stressors impair health behaviors, mental state, and quality of life, while strong social support and family function buffer these negative impacts and block adverse pathways [12, 13]. Studies have verified that social support alleviates the harmful effects of stress on emotions, behaviors, and long-term health [13, 14]. For stroke patients, economic toxicity is a typical chronic stressor that damages self-management and quality of life, and family function may buffer its negative effect. However, no study has verified the moderating role of family function in the pathway from economic toxicity through self-management to quality of life based on the Stress Buffering Model.
Currently, several research gaps remain. First, most stroke burden studies focus on macroeconomic levels rather than the micro-impact of economic toxicity on rehabilitation behaviors and quality of life. Second, studies on post-stroke quality of life mainly focus on physical function and demographic characteristics, lacking exploration of financial-psychological mechanisms and the mediating role of self-management. Third, although family function is protective, its buffering effect against economic toxicity has not been confirmed in a moderated mediation model. Fourth, evidence among Chinese populations under the context of local culture, healthcare, and family structure remains limited.
Based on the Stress Buffering Model, this study enrolled stroke patients in the rehabilitation stage to establish a moderated mediation model. We aimed to investigate the levels of economic toxicity, family function, self-management, and health-related quality of life; examine the direct effect of economic toxicity on quality of life; test the mediating role of self-management; and verify the moderating effect of family function on the association between economic toxicity and self-management. The results will enrich the understanding of quality-of-life mechanisms from financial-psychological and family support perspectives, support clinicians in identifying high‑risk patients, and provide evidence for developing integrated interventions including psychological support, family guidance, and self-management training to reduce family burden and improve long-term outcomes in stroke patients.
Participants and methods
Study design and setting
A cross-sectional, questionnaire-based observational study was conducted to examine the associations among economic toxicity, family function, self-management ability, and stroke-specific quality of life (QoL), and to test a moderated mediation model with stroke-specific QoL as the primary outcome variable.
This study was carried out in 22 medical and elderly care institutions in Chongqing, China, including 11 tertiary hospitals, 6 secondary hospitals, 3 community health service centers, and 2 nursing homes, covering urban and suburban areas with diversified health service resources.
Chongqing is a municipality in western China with both densely populated urban districts and less-developed suburban or rural counties. Stroke patients in this region are mainly covered by basic public medical insurance. However, reimbursement levels, accessibility to rehabilitation services, community-based rehabilitation resources, and long-term care support may vary across districts and counties. In the Chinese cultural context, family members remain the major providers of long-term care and rehabilitation support after stroke. Therefore, the association between economic toxicity, self-management, family function, and quality of life should be interpreted within this healthcare and sociocultural context.
Participants
Convenience sampling was used to recruit patients with stroke from May 12 to June 23, 2025, in Chongqing, China.
Inclusion Criteria: Diagnosed with stroke (cerebral infarction and/or cerebral hemorrhage) confirmed by cranial CT or MRI; In the stable recovery stage;
Clear consciousness and stable vital signs; Able to complete the questionnaires independently or with assistance; Agreed to participate in this study.
Exclusion Criteria: Severe aphasia, cognitive impairment, or communication barriers; A history of mental disorders; Severe concurrent organic diseases that severely affect quality of life.
To ensure data quality and eligibility, a strict pre-screening mechanism was employed prior to questionnaire distribution. Only pre-screened eligible patients were approached. The participant selection process is illustrated in Fig. 1.
Fig. 1.
Participant selection flow diagram
Measurements
Sociodemographic and clinical characteristics
A self-designed questionnaire was used to collect sociodemographic and clinical variables (e.g., gender, age, educational level, marital status, stroke type, stroke duration, comorbidities, etc.).
Chinese version of the comprehensive score for financial toxicity based on patient-reported outcomes (COST-PROM)[15, 16]
The Chinese version of the COST-PROM scale was employed in this study. This instrument comprises 11 items assessing economic expenditure, economic resources, and social implications. Responses were recorded on a 5-point Likert scale ranging from 0 to 4. The total score ranges from 0 to 44, for ease of interpretation, the COST-PROM total score was reverse-coded to generate an economic toxicity severity score, with higher scores indicating greater economic toxicity. In the present study, the scale demonstrated good internal consistency, with a Cronbach’s α coefficient of 0.865.
Family adaptability, partnership, growth, affection, resolve (APGAR) Scale [17]
Family function was assessed using the Family APGAR scale, which comprises five items: adaptation, partnership, growth, affection, and resolve. Each item is scored on a scale ranging from 0 to 2, yielding a total score between 0 and 10, with higher scores indicative of superior family function. The internal consistency of the scale was excellent, as demonstrated by a Cronbach’s α coefficient of 0.928.
Stroke self-management ability scale [18, 19]
The validated Stroke Self-Management Scale, comprising 23 items across six dimensions (communication confidence, professional guidance perception, decision-making ability, stroke impact, perceived confidence, and health management), was employed. Responses were measured on a 6-point Likert scale ranging from 1 (completely incorrect) to 6 (completely correct), with 13 items reverse-scored. The total scores ranged from 23 to 138, with higher scores indicating superior self-management ability. The scale demonstrated excellent internal consistency, with a Cronbach’s α coefficient of 0.900.
Conceptual distinction of key variables
In summary, these three instruments assess a patient’s subjective economic distress (COST-PROM), relational and emotional family system (Family APGAR), and individual health behavior agency (Stroke Self-Management Scale), respectively. They represent three distinct and non-overlapping dimensions of the post-stroke experience. Specifically, variance Inflation Factors (VIFs) were calculated in the multivariable regression models to assess multicollinearity among the key variables.
Stroke-specific quality of life (SS-QOL) Scale [20]
The Chinese version of the Stroke-Specific Quality of Life (SS-QOL) scale was employed in this study. This instrument comprises 49 items distributed across 12 dimensions: physical fitness, family role, language, mobility, mood, personality, self-care ability, social role, thinking, upper limb function, vision, and work productivity. Responses were assessed using a 5-point Likert scale (ranging from 1 to 5). The total score ranges from 49 to 245, with higher scores indicating a better quality of life. The Cronbach’s α coefficient for the scale was 0.954.
Descriptive statistics and internal consistency reliability for all measurement scales used in this study are presented in Table 1.
Table 1.
Reliability and descriptive statistics of study scales
| Scale | Items | Cronbach’s coefficient | Total score Mean ± SD |
Average score of items | Minimum | Maximum |
|---|---|---|---|---|---|---|
| Economic toxicity | ||||||
| Economic expenditure | 4 | 0.835 | 6.57 ± 4.57 | 1.64 ± 1.14 | 0 | 16 |
| Economic resources | 5 | 0.676 | 10.93 ± 4.64 | 2.19 ± 0.93 | 0 | 20 |
| Patient social reaction | 2 | 0.715 | 3.06 ± 2.51 | 1.53 ± 1.26 | 0 | 8 |
| Overall | 11 | 0.865 | 20.55 ± 10.10 | 1.87 ± 0.92 | 0 | 44 |
| Family function | ||||||
| Overall | 5 | 0.928 | 7.77 ± 3.01 | 1.55 ± 0.60 | 0 | 10 |
| Self-management ability | ||||||
| Communication confidence ability | 5 | 0.877 | 21.31 ± 6.02 | 4.26 ± 1.20 | 5 | 30 |
| Professional guidance perception ability | 4 | 0.719 | 18.88 ± 3.65 | 4.72 ± 0.91 | 4 | 24 |
| Self-management decision-making ability | 4 | 0.810 | 16.26 ± 4.81 | 4.06 ± 1.20 | 4 | 24 |
| Impact of stroke | 3 | 0.725 | 9.64 ± 3.60 | 3.21 ± 1.20 | 3 | 18 |
| Perceived self-confidence | 3 | 0.715 | 9.64 ± 3.61 | 3.21 ± 1.20 | 3 | 18 |
| Health management ability | 4 | 0.561 | 15.40 ± 3.82 | 3.85 ± 0.95 | 4 | 24 |
| Overall | 23 | 0.900 | 91.12 ± 18.65 | 3.96 ± 0.81 | 30 | 138 |
| Stroke-specific quality of life | ||||||
| Physical fitness | 3 | 0.734 | 8.04 ± 3.24 | 2.68 ± 1.08 | 3 | 15 |
| Family role | 3 | 0.806 | 9.22 ± 3.67 | 3.07 ± 1.22 | 3 | 15 |
| Language | 5 | 0.973 | 17.58 ± 7.64 | 3.52 ± 1.53 | 5 | 25 |
| Mobility | 6 | 0.949 | 16.10 ± 8.30 | 2.68 ± 1.38 | 6 | 30 |
| Mood | 5 | 0.887 | 16.82 ± 5.79 | 3.36 ± 1.16 | 5 | 25 |
| Personality | 3 | 0.887 | 10.56 ± 3.67 | 3.52 ± 1.22 | 3 | 15 |
| Self-care ability | 5 | 0.814 | 15.26 ± 5.89 | 3.05 ± 1.18 | 5 | 25 |
| Social role | 5 | 0.898 | 12.07 ± 5.91 | 2.41 ± 1.18 | 5 | 25 |
| Thinking | 3 | 0.876 | 8.78 ± 3.83 | 2.93 ± 1.28 | 3 | 15 |
| Upper limb function | 5 | 0.948 | 14.66 ± 7.49 | 2.93 ± 1.50 | 5 | 25 |
| Vision | 3 | 0.948 | 11.78 ± 3.54 | 3.93 ± 1.18 | 3 | 15 |
| Work productivity | 3 | 0.949 | 5.79 ± 3.69 | 1.93 ± 1.23 | 3 | 15 |
| Overall | 49 | 0.954 | 146.66 ± 40.57 | 2.99 ± 0.83 | 50 | 245 |
Note: Total score (Mean ± SD) = raw sum score of the scale/dimension; Average score of items = total score divided by the number of items in the corresponding dimension
Statistical methods
Data were collected via Questionnaire Star Online. Statistical analyses were performed using SAS 9.4. Normally distributed continuous variables were expressed as mean ± standard deviation. Group comparisons were conducted using independent-sample t-tests or one-way ANOVA, followed by SNK-q tests for pairwise comparisons. Categorical variables were described as frequencies and percentages. Correlations were examined using Pearson correlation analysis. Variables with P < 0.05 in univariate analysis were entered into a multivariable linear regression model using the stepwise method (entry criterion: P < 0.05; removal criterion: P ≥ 0.05). Basic demographic confounders (e.g., age, educational level) were adjusted in the model. Model 4 (mediation) and Model 7 (moderated mediation) from the PROCESS macro were used. Bootstrap with 5,000 resamples generated 95% confidence intervals for effect estimates. A two-sided α level of 0.05 indicated statistical significance. Harman’s single-factor test assessed common method bias, yielding 17 factors with eigenvalues > 1, with the first factor explaining 25.50% of total variance (below the 40% threshold), suggesting no serious common method bias.
Results
Participant characteristics and univariate analysis
A total of 557 stroke patients were included, with 354 males (63.55%) and 203 females (36.45%). Baseline characteristics and univariate comparisons of stroke-specific quality of life are shown in Table 2. Significant differences were found for gender, age, education, retirement, commercial insurance, primary caregiver, stroke type, recurrent stroke, and stroke duration (all P < 0.05). No significant differences were observed for marital status, medical insurance type, hypertension, or diabetes (all P > 0.05).
Table 2.
Comparisons of stroke-specific quality of life across patient subgroups
| Variable | Number of cases (%) | Hierarchical labels | Stroke-specific quality of life | t/F | P |
|---|---|---|---|---|---|
| General characteristics | |||||
| Gender | |||||
| Male | 354(63.55) | 149.40 ± 40.53 | 2.112 | 0.035 | |
| Female | 203(36.45) | 141.88 ± 40.29 | |||
| Age | |||||
| < 60 years old | 190(34.11) | a | 153.24 ± 34. 03c | 8.610 | < 0.001 |
| 60 ~ 80 years old | 241(43.27) | b | 147.88 ± 42. 67c | ||
| ≥ 80 years old | 126(22.62) | c | 134.38 ± 43. 05ab | ||
| Educational level | |||||
| Junior high school and below | 293(52.60) | a | 140.43 ± 41. 37bc | 8.233 | < 0.001 |
| High school or junior college | 183(32.85) | b | 151.52 ± 39. 52a | ||
| Bachelor degree and above | 81(14.54) | c | 158.17 ± 36. 17a | ||
| Marital status | |||||
| Married | 421(75.58) | 148.18 ± 40.24 | 1.561 | 0.119 | |
| Single or divorced or widowed | 136(24.42) | 141.94 ± 41.38 | |||
| Retired | |||||
| No | 289(51.89) | 150.52 ± 37.19 | 2.332 | 0.020 | |
| Yes | 268(48.11) | 142.49 ± 43.61 | |||
| Buy commercial insurance | |||||
| Yes | 258(46.32) | 151.95 ± 40.83 | 2.880 | 0.004 | |
| No | 299(53.68) | 142.09 ± 39.85 | |||
| Medical insurance type | |||||
| Employee medical insurance | 378(67.86) | a | 146.19 ± 41.70 | 0.119 | 0.888 |
| Resident medical insurance | 94(16.88) | b | 146.83 ± 39.11 | ||
| New Rural Cooperative Medical Insurance | 85(15.26) | c | 148.55 ± 37.33 | ||
| Primary caregiver | |||||
| Spouse | 170(30.52) | a | 155.26 ± 36. 79bcd | 18.796 | < 0.001 |
| Child | 96(17.24) | b | 141.58 ± 37. 56ad | ||
| Nanny or other | 245(43.99) | c | 136.76 ± 39. 99ad | ||
| No caregiver | 46(8.26) | d | 178.15 ± 41. 50abc | ||
| Disease related | |||||
| Isolated cerebral infarction | |||||
| No | 153(27.47) | 136.69 ± 35.61 | -3.870 | < 0.001 | |
| Yes | 404(72.53) | 150.43 ± 41.72 | |||
| Hemorrhage combined with infarction | |||||
| No | 507(91.02) | 147.64 ± 41.47 | 2.467 | 0.016 | |
| Yes | 50(8.98) | 136.72 ± 28.45 | |||
| First stroke | |||||
| No | 210(37.70) | 148.66 ± 40.60 | 0.908 | 0.364 | |
| Yes | 347(62.30) | 145.44 ± 40.56 | |||
| Recurrent stroke | |||||
| No | 453(81.33) | 148.42 ± 40.81 | 2.151 | 0.032 | |
| Yes | 104(18.67) | 138.96 ± 38.77 | |||
| Stroke duration | |||||
| < 1 year | 341(61.22) | a | 152.62 ± 39. 20bc | 12.480 | < 0.001 |
| 1 ~ 3 years | 95(17.06) | b | 130.29 ± 36. 65ac | ||
| > 3 years | 121(21.72) | c | 142.69 ± 43. 49ab | ||
| Diabetes | |||||
| No | 332(59.61) | 147.07 ± 41.42 | 0.290 | 0.772 | |
| Yes | 225(40.39) | 146.05 ± 39.36 | |||
| Hypertension | |||||
| No | 222(39.86) | 148.36 ± 38.83 | 0.809 | 0.419 | |
| Yes | 335(60.14) | 145.52 ± 41.70 |
Note: SNK-q test is used for pairwise comparison. a: compared with layer a, P < 0.05; b: compared with layer b, P < 0.05; c: compared with layer c,
P < 0.05; d: compared with layer d, P < 0.05
Correlation analysis
Correlation results are summarized in Table 3. Economic toxicity was negatively correlated with family function, self-management ability, and quality of life (all P < 0.05). Family function, self-management ability, and quality of life were positively correlated with each other (all P < 0.001).
Table 3.
Correlations among study variables
| Variables | Economic toxicity | Family function | Self-management ability | Quality of life |
|---|---|---|---|---|
| Economic toxicity | 1 | |||
| Family function | -0.145** | 1 | ||
| Self-management ability | -0.360*** | 0.441*** | 1 | |
| Quality of life | -0.424*** | 0.277*** | 0.679*** | 1 |
Note: ***P < 0.001. **P < 0.05
Multiple linear regression for quality of life
Independent predictors of quality of life are displayed in Table 4. Better quality of life was predicted by higher education, being cared for by a spouse, having no caregiver, cerebral infarction diagnosis, shorter stroke duration, lower economic toxicity, and higher self-management ability (all P < 0.05). The model was significant (F = 66.98, P < 0.001), with R² = 0.5509. Importantly, the Variance Inflation Factor (VIFs) for all predictors ranged from 1.028 to 1.253 (all < 10), indicating no severe multicollinearity among the independent variables (Table 4).
Table 4.
Results of multiple linear regression model for quality of life
| Variables | β | Standard error | Standardized regression coefficient | t | P | VIF |
|---|---|---|---|---|---|---|
| Constant | 6.440 | 6.507 | — | 0.990 | 0.323 | 0.000 |
| Educational level | ||||||
| Junior high school and below | 0.0(Reference) | |||||
| High school or junior college | 7.940 | 2.679 | 0.092 | 2.964 | 0.003 | 1.171 |
| Bachelor degree and above | 11.342 | 3.540 | 0.099 | 3.204 | 0.001 | 1.152 |
| Primary caregiver | ||||||
| Spouse | 8.493 | 2.826 | 0.096 | 3.006 | 0.003 | 1.253 |
| Child | 6.111 | 3.391 | 0.057 | 1.802 | 0.072 | 1.214 |
| Nanny or other | 0.0(Reference) | |||||
| No caregiver | 19.439 | 4.561 | 0.132 | 4.262 | < 0.001 | 1.166 |
| Cerebral infarction | ||||||
| No | 0.0(Reference) | |||||
| Yes | 10.458 | 2.641 | 0.115 | 3.961 | < 0.001 | 1.028 |
| Stroke duration | ||||||
| < 1 year | 0.0(Reference) | |||||
| 1 ~ 3 years | -10.964 | 3.255 | -0.102 | -3.369 | 0.001 | 1.109 |
| > 3 years | -6.403 | 2.967 | -0.065 | -2.158 | 0.031 | 1.107 |
| Economic toxicity | -0.726 | 0.125 | 0.181 | -5.789 | < 0.001 | 1.184 |
| Self-management ability | 1.200 | 0.069 | 0.552 | 17.304 | < 0.001 | 1.236 |
Note: Model F = 66.98 (df = 10, 546, P < 0.001), R2 = 0.5509, adjusted R2 = 0.5427
Mediating effect of self-management ability
To examine whether self-management ability mediated the effect of economic toxicity on stroke-specific quality of life, mediation analysis was conducted using PROCESS Model 4 (Table 5).
Table 5.
Mediating effect of self-management ability in the association between economic toxicity and quality of life
| Variable | Model 1: Quality of life | Model 2: Self-management ability | Model 3: Quality of life | ||||||
|---|---|---|---|---|---|---|---|---|---|
| β | t | P | β | t | P | β | t | P | |
| Constant | 0.832 | 2.923 | 0.004 | 0.809 | 2.644 | 0.008 | 0.388 | 1.675 | 0.095 |
| Economic toxicity | -0.360 | -9.754 | < 0.001 | -0.331 | -8.353 | < 0.001 | -0.178 | -5.617 | < 0.001 |
| Self-management ability | — | — | — | — | — | — | 0.549 | 16.926 | < 0.001 |
| R 2 | 0.316 | 0.210 | 0.554 | ||||||
| F | 13.075 | 7.499 | 33.349 | ||||||
| P | < 0.001 | < 0.001 | < 0.001 | ||||||
Economic toxicity had a significant negative total effect on quality of life (total effect = − 0.360, 95% CI: −0.433 to − 0.288, P < 0.001). Additionally, economic toxicity significantly negatively predicted self-management ability (β = −0.331, P < 0.001).
After controlling for self-management ability, the direct effect of economic toxicity on quality of life remained significant (direct effect = − 0.178, 95% CI: −0.240 to − 0.116, P < 0.001). The indirect effect through self-management ability was − 0.182 (95% CI: −0.234 to − 0.134), accounting for 50.56% of the total effect, confirming a partial mediation.
Moderating effect of family function
To test whether family function buffered the negative impact of economic toxicity, moderation analysis was performed using PROCESS Model 7 (Table 6).
Table 6.
Moderating effect of family function on the association between economic toxicity and self-management ability
| Variable | Model dependent variable: self-management ability | ||
|---|---|---|---|
| β | t | P | |
| Constant | 0.704 | 2.583 | 0.010 |
| Economic toxicity | -0.281 | -7.947 | < 0.001 |
| Family function | 0.344 | 9.711 | < 0.001 |
| Economic toxicity * Family function | 0.209 | 6.086 | < 0.001 |
| R 2 | 0.383 | ||
| F | 15.808 | ||
| P | < 0.001 | ||
The interaction term between economic toxicity and family function was statistically significant (β = 0.209, P < 0.001), indicating that family function significantly moderated the pathway from economic toxicity to self-management ability. The moderated mediation model is illustrated in Fig. 2.
Fig. 2.
Moderated mediation model of quality of life. Note: ***P < 0.001
Simple slope analysis was further conducted (Table 7) and plotted in Fig. 3.
Table 7.
Effects of economic toxicity on self-management ability by levels of family function
| Family function | β (95%CI) | Standard error | t | P |
|---|---|---|---|---|
| Low family function | -0.49(-0.586,-0.395) | 0.048 | -10.126 | < 0.001 |
| Moderate family functionality | -0.281(-0.351,-0.212) | 0.035 | -7.947 | < 0.001 |
| High family function | -0.126(-0.213,-0.039) | 0.044 | -2.846 | 0.005 |
Fig. 3.
Simple slope analysis of the moderating effect of family function on the association between economic toxicity and self-management ability
In the low family function group, economic toxicity had a strong negative effect on self-management ability (β = −0.490, P < 0.001). In the high family function group, this negative effect was significantly weakened (β = −0.126, P = 0.005).
The contrast of conditional indirect effects generated by PROCESS Model 7 yielded a difference of 0.200 (95% CI: 0.135 – 0.266, P < 0.001), indicating that higher family function buffered the adverse pathway from economic toxicity to reduced self-management and poorer quality of life.
Discussion
Univariate and multivariable analyses showed that factors such as gender, age, education level, primary caregiver, stroke type (cerebral infarction), disease duration, economic toxicity, and self-management ability, among others, were independent predictors of stroke-specific quality of life (Tables 2 and 4). This pattern is consistent with the broader epidemiological context in China, where demographic transition and socioeconomic disparity are increasingly recognized as major contributors to stroke burden [21]. Building on these findings, we tested a moderated mediation model to examine the joint roles of economic toxicity, self-management, and family function during rehabilitation. The model showed no evidence of problematic multicollinearity (all VIFs = 1.028 – 1.253), supporting the conceptual distinctiveness of the three core constructs and the interpretability of their independent effects. Importantly, economic toxicity was negatively correlated with quality of life (r = − 0.424, P < 0.001) and remained an independent negative predictor after adjustment for demographic and clinical covariates (Table 4; Fig. 2). This is in line with prior evidence that financial strain can worsen long-term rehabilitation outcomes through persistent psychological stress, reduced treatment adherence, and poorer health behaviors [5, 22]. The population-level relevance of this pathway is further supported by recent evidence showing a substantial national economic burden of stroke in China [23]. Our additional analyses on financial protection factors also provide useful context. Commercial insurance was associated with better quality of life (P = 0.004), whereas basic medical insurance type was not (P = 0.888). This finding is compatible with Wang et al. [24], who reported that Diagnosis-Related Groups (DRG) payment reform reduced hospitalization costs and length of stay in western China but mainly affected acute inpatient care, with limited influence on rehabilitation-phase expenses. Together, these results suggest that basic coverage alone may be insufficient to offset economic toxicity across the full recovery trajectory, while supplementary protection may be more relevant for preserving quality of life. From a practice perspective, routine screening for economic toxicity and targeted financial support could be integrated into stroke rehabilitation management.
This study further confirmed the partial mediating role of self-management ability in the relationship between economic toxicity and quality of life. The indirect effect through self-management was − 0.182 (95% CI: − 0.234 to − 0.134), accounting for 50.56% of the total effect. This indicates that economic toxicity not only directly erodes quality of life but also operates indirectly by weakening patients’ capacity to engage in rehabilitation-oriented self-management behaviors. Earlier research has demonstrated that self-management ability improves disease cognition, health beliefs, and functional outcomes in stroke populations [25, 26]. These findings align with the scoping review by Zhou et al. [27], which also identified financial strain as a key barrier to post-stroke self-management among young survivors. Critically, our study advances prior descriptive evidence by formally testing a moderated mediation model that positions economic toxicity as an upstream barrier to self-management in stroke rehabilitation: financial strain reduces patients’ sense of health control, limits access to rehabilitation resources, and diminishes investment in active health behaviors. Importantly, post-stroke mental health conditions may further amplify this pathway. Unmeasured post-stroke psychological sequelae constitute an important confounding factor. Up to 43% of mild stroke survivors develop anxiety, depressive symptoms and cognitive dysfunction within one year after onset; these hidden emotional impairments independently reduce rehabilitation engagement and subjective quality of life, and may amplify perceived financial distress [28].These converging lines of evidence suggest that psychological distress may not only impair self-management capacity but also heighten perceived economic toxicity, potentially creating a reinforcing cycle of poor health and financial strain. Therefore, integrating psychological screening and support into self-management interventions—particularly for economically disadvantaged patients—may be essential to maximize rehabilitation gains. This is especially critical given that quality of life is a multidimensional construct, and as highlighted in recent reviews, standard measures may not fully capture the complex interplay of emotional and social nuances that act as residual confounders [29].
The most notable finding of this study is the significant buffering role of family function in the pathway from economic toxicity to self-management ability (interaction β = 0.209, P < 0.001). Simple slope analysis revealed that among patients with low family function, economic toxicity exerted a strong negative effect on self-management ability (β = − 0.490, P < 0.001), whereas in the high family function group, this negative effect was markedly attenuated (β = − 0.126, P = 0.005). The difference in conditional indirect effects was 0.200 (95% CI: 0.135 – 0.266, P < 0.001), confirming that higher family function buffered the harmful chain from economic toxicity to reduced self-management and poorer quality of life. As discussed above, the current healthcare system provides limited financial coverage for the post-acute rehabilitation phase, making stroke survivors particularly dependent on family-based support. This contextual reality may explain why the buffering effect of family function against economic toxicity was especially pronounced in our study, as families serve as the primary safety net when institutional support diminishes.
This pattern is consistent with the core premise of the stress-buffering model, which posits that adequate social support can mitigate the harmful impact of negative stressors on health behaviors and outcomes [12, 13]. Our findings also resonate with previous studies showing that good family function enhances coping efficacy, disease adaptation and quality of life in stroke patients [30, 31]. In Eastern cultural contexts, where the family typically serves as the primary source of emotional care and daily support, a well-functioning family environment may effectively alleviate the psychological burden of economic pressure, stabilize rehabilitation motivation, and sustain self-management behaviors. It is important to interpret this buffering effect within the broader healthcare context. In settings with generous public rehabilitation services and comprehensive insurance coverage, the protective role of family function may be partially attenuated by systemic support. Conversely, when rehabilitation resources are unevenly distributed and out-of-pocket costs remain high—as is the case in many regions of China—patients rely more heavily on family support, making the buffering role particularly salient. Therefore, family function should not be viewed as a substitute for adequate healthcare system support, but rather as a critical complementary protective factor within the current care environment. Specifically, a supportive family atmosphere can buffer the psychological toll of economic toxicity, thereby sustaining patients’ self-management behaviors when systemic resources are limited.
These findings carry several clinical implications. First, routine economic toxicity screening using validated tools such as the COST-PROM scale could be integrated into standard rehabilitation assessments, treating financial distress as a clinical vital sign. Second, for high-risk subgroups—including patients covered only by basic medical insurance and those with severe functional deficits—a combined family-support and self-management layered nursing protocol is recommended, directly leveraging the mediating (50.56%) and buffering (β = 0.209) effects identified in this study. Given the high prevalence of undetected mental health comorbidities in stroke populations [28, 32], psychological screening and referral pathways could be embedded within rehabilitation follow-up protocols, particularly for patients reporting high economic toxicity. Third, at the community-policy linkage level, a hospital-community-family collaborative management model could be established to disseminate information on critical illness insurance and rehabilitation subsidy policies, thereby preventing the cycle of functional decline and economic strain.
Strengths and limitations
This study was conducted based on a large-sample (557 cases), multi-institutional cross-sectional survey, utilizing internationally recognized measurement scales with good reliability and validity, and constructing a moderated mediation model for mechanism analysis, thereby possessing certain advantages in methodology and research content. Firstly, the sample size of this study is sufficient, with data sources covering tertiary hospitals, secondary hospitals, community health service centers, and elderly care institutions; the research subjects are relatively representative, and the results possess a certain degree of generalizability. Secondly, this study adopted internationally accepted instruments such as the COST-PROM, the family APGAR Scale, the Stroke Self-Management Scale, and the SS-QOL Scale, ensuring standardized measurement and reliable results. Thirdly, this study provides a novel examination, within a Chinese stroke population, of a comprehensive model in which economic toxicity affects quality of life through self-management, a pathway buffered by family function; this model is grounded in sufficient theoretical basis and has clear clinical implications, providing new insights for stroke rehabilitation nursing and health management.
Several limitations should be acknowledged. First, the cross-sectional design precludes causal inferences regarding the observed relationships between economic toxicity, self-management, family function, and quality of life. Second, the convenience sampling method from a single urban-rural medical region in Chongqing, China—with its specific healthcare reimbursement policies, rehabilitation resource distribution, and family-based caregiving culture—may introduce selection bias and limit the generalizability of our findings to other healthcare settings or populations. Third, despite statistical adjustments for age and education, residual confounding from unmeasured variables—such as post-stroke functional severity, pain, depression, pre-stroke employment type, and specific financial mitigation interventions could not be fully ruled out. Fourth, the reliance on self-reported questionnaires may be subject to social desirability and recall bias, particularly among participants who required assistance with completion.
Conclusion
This study confirms that economic toxicity in stroke patients during the rehabilitation phase is a significant negative risk factor for quality of life; it not only directly reduces quality of life but also indirectly impairs it by weakening self-management abilities. Concurrently, family function plays a significant buffering and protective role between economic toxicity and self-management, as favorable family function can effectively mitigate the adverse impacts stemming from financial pressure. Future research should conduct prospective cohort studies and intervention studies to further verify the causal relationships among variables, and construct a comprehensive rehabilitation model centered on economic risk intervention, self-management improvement, and family support enhancement, so as to provide higher-level evidence for improving the long-term prognosis of stroke patients.
Acknowledgements
The authors would like to thank all the stroke patients who participated in this study. They also acknowledge the support of the staff at the participating medical and elderly care institutions in Chongqing for their assistance in data collection.
Author contributions
All authors have made substantial contributions to the work, which are as follows. Lili Yao, Shan He, Ling Xiao and Ling Liu contributed to Conceptualization, methodology, formal analysis, writing – original draft, writing–review & editing. Yetao Luo contributed to Formal analysis, software, data curation. Yi Duan, Qing Liu and Yanhui Du contributed to Investigation, data collection. All authors read and approved the final manuscript.
Funding
2026 Central Subsidy Fund for Enhancing Medical Services and Guarantee Capabilities (Healthcare Institution Capacity Building): National Clinical Key Specialty Construction Project - Rehabilitation Medicine [Grant No. 0203 (2026) No. 10-20265602].
Data availability
The datasets used and analyzed during this study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and informed consent
This study was conducted in accordance with the Declaration of Helsinki.The protocol was approved by the Ethics Committee of The First Affiliated Hospital of Chongqing Medical University (approval number: 2025-263-01). Informed consent was obtained from all participants.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ling Xiao and Ling Liu contributed equally as corresponding authors.
Contributor Information
Ling Xiao, Email: 710452902@qq.com.
Ling Liu, Email: liuling20000403@163.com.
References
- 1.Collaborators GBDS. Global, regional, and national burden of stroke and its risk factors, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Neurol. 2021;20(10):795–820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Disease GBD, Injury, Risk Factor C. Burden of 375 diseases and injuries, risk-attributable burden of 88 risk factors, and healthy life expectancy in 204 countries and territories, including 660 subnational locations, 1990–2023: a systematic analysis for the Global Burden of Disease Study 2023. Lancet. 2025;406(10513):1873–922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kazi DS, Elkind MSV, Deutsch A. Forecasting the economic burden of cardiovascular disease and stroke in the United States Through 2050: a presidential advisory from the American Heart Association. Circulation. 2024;150(4):e89–101. [DOI] [PubMed] [Google Scholar]
- 4.Jin L, Yin X, Wang J. Financial toxicity in stroke patients: current status, influencing factors, and coping strategies: a review. Top Stroke Rehabil. 2026. [DOI] [PubMed]
- 5.Xu L, Dong Q, Jin A. Experience of financial toxicity and coping strategies in young and middle-aged patients with stroke: a qualitative study. BMC Health Serv Res. 2024;24(1):94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gurková E, Štureková L, Mandysová P. Factors affecting the quality of life after ischemic stroke in young adults: a scoping review. Health Qual Life Outcomes. 2023;21(1):4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sun YA, Kalpakavadi S, Prior S. Socioeconomic status and health-related quality of life after stroke: a systematic review and meta-analysis. Health Qual Life Outcomes. 2023;21(1):115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Silva AC, Menezes KKP, Scianni AA. Predictors of health-related quality of life one year after stroke: a systematic review with meta-analysis. Int J Rehabil Res. 2024;47(2):53–63. [DOI] [PubMed] [Google Scholar]
- 9.Li Y, Ying J, Zhang X. Coping strategies mediate the association between family functioning and posttraumatic growth in family caregivers of people with dementia. Aging Ment Health. 2021;25(9):1684–91. [DOI] [PubMed] [Google Scholar]
- 10.Tyagi S, Luo N, Tan CS. Support system diversity among family caregivers of stroke survivors: a qualitative study exploring Asian perspectives. BMC Geriatr. 2021;21(1):594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wang H, Yue H, Ren M. Dyadic effects of family functioning and resilience on quality of life in advanced lung cancer patients and caregivers: an actor partner interdependence mediation model. Eur J Oncol Nurs. 2021;52:101963. [DOI] [PubMed] [Google Scholar]
- 12.Bekiros S, Jahanshahi H, Munoz-Pacheco JM. A new buffering theory of social support and psychological stress. PLoS ONE. 2022;17(10):e0275364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Delfin D, Wallace J, Baez S. Social support, stress, and mental health: examining the stress-buffering hypothesis in adolescent football athletes. J Athl Train. 2024;59(5):499–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jiang Y, Zilioli S, Stowe RP. Perceived social support and latent herpesvirus reactivation: testing main and stress-buffering effects in an ethnically diverse sample of adults. Psychosom Med. 2021;83(7):767–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.de Souza JA, Yap BJ, Hlubocky FJ, Wroblewski K, Ratain MJ, Cella D, et al. The development of a financial toxicity patient-reported outcome in cancer: The COST measure. Cancer. 2014;120(20):3245–53. [DOI] [PubMed] [Google Scholar]
- 16.Luo J, Sinicization. validation, and application of the comprehensive score for financial toxicity based on patient-reported outcomes. university of electronic science and technology of China; 2023.
- 17.Li XM, Zheng Y. Research progress of family function assessment tools. Chin Gen Pract Nurs. 2022;20(20):2765–8. [Google Scholar]
- 18.Boger EJ, Hankins M, Demain SH, Latter SM. Development and psychometric evaluation of a new patient-reported outcome measure for stroke self-management: The Southampton Stroke Self-Management Questionnaire (SSSMQ). Health Qual Life Outcomes. 2015;13:165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wang S, Zhou J, Wu QD. Reliability and validity of the Chinese version of Stroke Self-Management Scale. Chin Nurs Res. 2021;35(15):2682–6. [Google Scholar]
- 20.Wang Y, Ma J, Li J. Preliminary study on the reliability, validity and sensitivity of the Chinese version of the Stroke-Specific Quality of Life Scale. Chin J Geriatric Heart Brain Vessel Dis. 2003;5(6):4. [Google Scholar]
- 21.Zhao Y, Hua X, Ren X. Increasing burden of stroke in China: A systematic review and meta-analysis of prevalence, incidence, mortality, and case fatality. Int J Stroke. 2023;18(3):259–67. [DOI] [PubMed] [Google Scholar]
- 22.Scott JW, Knowlton LM, Murphy P. Financial toxicity after trauma and acute care surgery: From understanding to action. J Trauma Acute Care Surg. 2023;95(5):800–5. [DOI] [PubMed] [Google Scholar]
- 23.Li Y, Wang Y, Liu X. Estimating the economic burden of stroke in China: a cost-of-illness study. BMJ Open. 2024;14(3):e080634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wang J, Chen M, Yang J. Impact of DRG reform policies on hospitalization costs of stroke patients in western China: wisdom from traditional Chinese medicine. Risk Management and Healthcare Policy; 2025. [DOI] [PMC free article] [PubMed]
- 25.Lynch EA, Nesbitt K, Gulyani A. Do self-management interventions improve self-efficacy and health-related quality of life after stroke? A systematic review. Int J Stroke. 2025;20(7):786–800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Xing L, Wei J. The effect of self-management on the knowledge, beliefs, behavior and subjective well-being in stroke patients during the rehabilitation phase. Am J Translational Res. 2021;13(7):8337–43. [PMC free article] [PubMed] [Google Scholar]
- 27.Zhou Z, Fang X, Huang Y. A scoping review of factors associated with self-management in young adults with stroke. Patient Educ Couns. 2024;125:108308. [DOI] [PubMed] [Google Scholar]
- 28.Vlachos G, Ihle-Hansen H, Bruun Wyller T, Brækhus A, Mangset M, Hamre C, et al. Cognitive and emotional symptoms in patients with first-ever mild stroke: The syndrome of hidden impairments. J Rehabil Med. 2021;53(1):jrm00135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.George CM, Dabbagh A, Unger J. Quality of life measures for people following stroke: a structured content review. Qual Life Res. 2025;34(4):893–912. [DOI] [PubMed] [Google Scholar]
- 30.Li CC, Tsai SJ, Tai JC. Association Between Family Functioning and Health-related Quality of Life in Stroke Survivor-Informal Family Caregiver Dyads. Asian Nurs Res. 2024;18(2):141–7. [DOI] [PubMed] [Google Scholar]
- 31.Zhang W, Gao YJ, Ye MM. Post-stroke family resilience is correlated with family functioning among stroke survivors: The mediating role of patient’s coping and self-efficacy. Nurs Open. 2024;11(7):e2230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ignacio KHD, Muir RT, Diestro JDB, Singh N, Yu MHLL, Omari OE, et al. Prevalence of depression and anxiety symptoms after stroke in young adults: A systematic review and meta-analysis. J Stroke Cerebrovasc Dis. 2024;33(7):107732. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and analyzed during this study are available from the corresponding author on reasonable request.