Background

Stroke is a major global health burden characterized by high incidence, recurrence rates, and costly treatment1,2. According to a 2024 report from The Lancet, up to one in five people in high-income countries will experience a stroke in their lifetime, while in low-income countries, this proportion rises to nearly one in two individuals. Globally, stroke is the second leading cause of death3. Recent studies indicate that the proportion of stroke survivors aged 40–64 in China has increased to 52.65%, with 65% experiencing their first stroke before the age of 65, highlighting the trend of rising stroke incidence among young and middle-aged individuals4. Stroke research suggests that effective treatment, robust self-management support, and regular follow-up are key factors in ensuring successful rehabilitation for stroke survivors5.

Self-management refers to behaviors in which survivors actively maintain and improve their health, reduce the impact of disease on their social functions, emotions, and interpersonal relationships, and consistently manage their illness6,7. Studies have found that positive self-management interventions can equip survivors with essential self-management skills such as goal-setting, decision-making, and self-monitoring6. Even after the intervention ends, survivors who have acquired these skills are more likely to maintain a healthy lifestyle compared to those lacking self-management skills. Self-management interventions have been shown to improve outcomes in chronic disease management, but their effectiveness in stroke rehabilitation remains understudied. Watkins et al. found that motivational interviewing has been demonstrated to reduce emotional distress and promote better adjustment after stroke8. Similarly, the Take Charge intervention, which focuses on self-directed rehabilitation, has shown significant improvements in quality of life and functional outcomes9. These studies highlighted the potential of structured self-management approaches in stroke rehabilitation.

However, the levels of self-management among stroke survivors are generally low, with over half of the survivors not knowing how to conduct standardized self-behavior management10. A cross-sectional study showed that the self-management behavior of Chinese stroke patients is at a moderate level, and the main risk factors include self-management efficacy, discharge health education, age, and education level11. Lindsay Bright et al. surveyed 736 survivors with mild stroke, finding that up to 10% reported deficiencies in self-care behaviors12. Sohyune et al. analyzed 146 stroke survivors in Korea, revealing that regardless of risk factors, stroke survivors exhibited low levels of self-care behavior, which correlated with lower health-related quality of life13. Magwood et al. found that in the past 12 months, less than two-thirds of individuals sought knowledge about stroke self-management behaviors14. Therefore, effectively raising self-management awareness and cultivating the self-management skills of young and middle-aged stroke survivors are pressing issues.

To address the challenges faced by young and middle-aged stroke survivor in self-management education, Health Empowerment Theory (HET) has been widely applied15,16. This theory emphasizes patient-centered care, advocating for collaboration between healthcare providers and patients to jointly promote recovery. By offering personalized advice, healthcare providers assist patients in evaluating the costs and benefits of treatment plans, setting achievable goals, and encouraging active participation. This approach significantly enhances patients’ self-management awareness and engagement. Compared to the traditional passive acceptance of health education, HET more effectively facilitates lifestyle modifications in patients.

However, maintaining health behaviors during post-discharge rehabilitation for stroke survivors is a gradual, iterative process. Thus, establishing an objective and practical assessment tool to identify patients’ current behavioral stages and tailor stage-specific interventions is critical. The Trans-theoretical Model (TTM), designed to help patients consciously adopt and sustain healthy behaviors, is commonly used to study individual behavior change17,18. While TTM focuses on behavioral stages and their characteristics, it lacks explicit guidance on stage-matched interventions19,20. Integrating HET effectively addresses this limitation. Specifically, TTM serves to assess patients’ behavioral change stages, while HET guides the intervention process, enabling their synergistic application. This dual-theory integration not only optimizes personalized intervention plans but also promotes health behavior changes more effectively, thereby supporting patients’ long-term self-management.

Previous randomized controlled trials have often employed single-mode theoretical guidance, failing to integrate the process and methods of stroke survivor interventions with the characteristics of behavioral stage transitions21,22. Additionally, the study populations and forms of application have been relatively broad, necessitating further exploration specifically for young and middle-aged stroke populations. Thus, this study aims to evaluate the distribution of self-management behavior stages in young and middle-aged stroke survivor using the TTM and compare the clinical efficacy of Trans-theoretical model-based Empowerment education versus conventional health education in enhancing self-management behaviors, thereby establishing the applicability and value of integrated Empowerment strategies for optimizing behavioral interventions in this population.

Materials and methods

Study design

This is a two-arm, single-blind, randomized controlled trial (Clinical Trial Registration Number: ChiCTR2300070976).

Random processes and allocation concealment

A random number sequence was generated using Stata 18.0. Participants were stratified based on key baseline characteristics: age, gender, and stroke severity. The randomized allocation sequence, stratified by these factors, was concealed within sequentially numbered, opaque, sealed envelopes. Upon participant enrollment, after obtaining informed consent and completing baseline assessments, an independent data analyst blinded to randomization sequence generation and envelope preparation opened the envelope corresponding to the participant’s enrollment order. This process revealed the participant’s group allocation.

Blinding

Independent assessors, with no involvement in study design, implementation, or participant recruitment, were appointed as blinded outcome assessors. These blinded assessors conducted follow-up assessments at 1-, 3- and 6-months post-intervention. During assessments, blinded assessors utilized only predefined instruments. They refrained from actively inquiring or attempting to deduce group allocation. Assessments were directly reported to the data analyst, with blinded assessors excluded from subsequent data entry or analysis stages. To maintain blinding and protect participant confidentiality, all identifying information (e.g., names, contact details) was anonymized at the data collection stage. Participants were identified solely by a unique coded number. Paper-based records were securely stored in locked cabinets, and electronic data was encrypted and stored on password-protected, secure devices. It is acknowledged that blinding of researchers and participants was not feasible due to the nature of the intervention.

Participants recruitment

We recruited eligible stroke survivors from four neurology departments at Shaanxi Provincial People’s Hospital in Xi’an, Shaanxi Province, China. This hospital is a Grade III A public hospital in Northwest China and serves as a model hospital for stroke screening and prevention. Stroke survivors were considered eligible if they met the following criteria: Meeting the diagnostic criteria for stroke, confirmed by head Computed Tomography (CT) or Magnetic Resonance Imaging (MRI); Age between 18 and 59 years; Voluntarily agreed to participate; Owning a smartphone with WeChat (a mobile instant messaging application). Exclusion criteria included: History of mental illness; Presence of other major diseases or systemic infections; Severe limb pain or spasticity, or recent untreated seizures; Aphasia, hearing loss, or consciousness disorders making cooperation impossible; Expected survival less than 3 months.

Sample size

The sample size for this study was calculated using the following formula for comparing two independent groups:

$$n={\left( {\frac{{Z\alpha /2+Z\beta }}{{\delta /\sigma }}} \right)^2}$$

where: n = Sample size per group; \(Z\alpha /2\) = Standard normal deviate corresponding to the two-sided significance level (α) = 1.96 (for α = 0.05); \(Z\beta\) = Standard normal deviate corresponding to the desired power (1 - β) = 1.28 (for β = 0.10, 80% power); σ = Estimated standard deviation of self-management behavior scales in the intervention group = 9.24 (based on previous studies); δ = Mean difference of self-management behavior scale scores before and after intervention = 63.30.

Based on the formula, the calculated sample size per group was approximately 38.73. To account for a potential 20% attrition rate during the study, we inflated the sample size by dividing by (1 - attrition rate):

$${n_{{\text{adjusted}}}}=n/{\text{ }}\left( {{\text{1 }} - {\text{ attrition rate}}} \right)\,=\,{\text{38}}.{\text{73 }}/{\text{ }}\left( {{\text{1}}-0.{\text{2}}0} \right)\, \approx \,{\text{48}}.{\text{41}}$$

Rounding up to the nearest whole number, each group would require approximately 49 participants, resulting in a total of 98 participants for the study. To further ensure the reliability of the study results, minimize random error, and enhance internal validity, we ultimately recruited a total of 104 participants (52 per group).

Intervention group

Participants in the intervention group attended a health education program based on TTM-based Empowerment education, conducted every 1–2 weeks by a multidisciplinary team including doctors, nurses, health managers, family members, rehabilitation therapists, pharmacists, nutritionists, and psychologists. Each session lasted 15–20 min and continued for 24 weeks. Before the intervention, researchers assessed participants’ self-management behaviors stages using a self-management behavior stage assessment questionnaire. Based on these stages, appropriate health education topics were selected. After each session, which was tailored to the current behavior stage, participants’ self-management behavior stages were reassessed, and the content of the health education program was adjusted accordingly. Since behavior change often involves relapses, we continually reassessed behavior stages and provide stage-matched educational content. The health education program was based on evidence-based knowledge and consultations with Delphi experts. Detailed strategies are outlined in Table 1.

Table 1 Empowerment intervention program based on the Trans-Theoretical model (TTM).
Full size table

Control group

In the conventional health education program, stroke survivors are introduced to the hospital environment and admission procedures upon arrival. An admission assessment was conducted to evaluate their needs. During the hospital stay, stroke survivors received instruction on various aspects of care, including bed rest, dietary needs, medication management, in-bed hygiene, limb positioning, and swallowing function training. On the day before discharge, stroke survivors were provided with health education covering diet, medication, lifestyle routines, and the importance of regular follow-up visits. After discharge, a follow-up record was established for each survivor. Telephone follow-ups were conducted to guide stroke survivors on home care issues and to ensure they adhere to scheduled stroke survivor visits.

Ethics and data security

This study has been approved by the Shaanxi provincial people’s hospital ethics committee (Approval Number: R035). All methods were performed in accordance with the Declaration of Helsinki and relevant guidelines/regulations. All participants included in the study provided written informed consent. Participants were informed that their participation is voluntary, and they may withdraw from the study at any time without any negative impact on their future care. All paper documents and data were stored in a secure filing cabinet. Electronic data was protected by passwords and stored on a secure laptop. Documents containing personal identification information were kept separate from other research data and were identified by coded numbers. Access to files was restricted to the research personnel involved in the study. The final statistical analysis was conducted using de-identified data, with participants’ identification information replaced by unrelated numeric codes.

Outcomes and measuring instruments

Socio-Demographic profile

(1) General Information Survey Form

Based on extensive review of relevant literature, design a general information survey form including age, gender, educational level, marital status, average monthly family income, primary caregiver, number of coexisting chronic diseases, muscle strength, and healthcare payment method.

Outcomes

(2) Stroke Self-Management Behavior Assessment Scale

This scale was designed by Wang Yanjiao et al. for assessing self-management behaviors in stroke recovery survivors. It consists of 51 items across 7 dimensions, each rated on a 5-point scale. The total score ranges from 50 to 255, with higher scores indicating better self-management behaviors. The Cronbach’s α for this scale was 0.83523.

(3) Stroke Rehabilitation Self-Efficacy Scale

Developed by Jones and translated by Li Hongyan et al., this scale evaluates stroke survivors’ perceived rehabilitation self-efficacy. It includes 13 items, with a total score ranging from 0 to 130. Higher scores indicate better self-efficacy. The Cronbach’s α for this scale was 0.9724.

(4) Self-Management Behavior Stage Assessment Questionnaire

This questionnaire was based on the TTM and adapted from the Chinese version of the behavior stage assessment questionnaire developed by the American Cancer Prevention Research Center. It has been localized and adjusted for stroke self-management behavior over a 6-month period. A pre-study with 50 randomly selected survivors determined the test-retest reliability to be 0.83. The questionnaire contains 5 options to determine the stage of self-management behavior based on the stroke survivors’ responses (see Table 1).

(5) Barthel Index

The Barthel Index (BI) was used to assess stroke survivors’ daily living activities. It includes 10 items, with a total score ranging from 0 to 100. Higher scores reflect greater independence. The index has demonstrated good internal consistency, inter-rater reliability, and concurrent validity25,26.

Data collection

With the consent and support of the management of Shaanxi Provincial People’s Hospital, participants meeting the inclusion criteria were recruited from the Neurology Department. The study’s purpose, significance, and procedures were explained to the participants, and informed consent was obtained before inclusion in the study. Before completing the questionnaires, participants received standardized instructions from the researchers. During the completion process, any questions were addressed using uniform explanations. Questionnaires were checked on-site upon collection; if any items were missing, participants were asked to complete them after receiving clarification. Following the initial intervention, if survivors were discharged home, data collection was conducted via telephone follow-up at 1 month (T1), 3 months (T2), and 6 months (T3) post-intervention. If participants did not respond to follow-up attempts, they were contacted up to two additional times to inquire about their willingness to complete the survey. Specific measurement time points for each study indicator are detailed in Supplement 1.

Statistical analysis

Data were entered into the Epidata database, where a Check file was set up to define range limits and perform double entry with verification. After ensuring the accuracy of the data, it was imported into IBM SPSS Statistics 26.0 for statistical analysis. Categorical data was presented as frequencies and percentages, with inter-group comparisons made using chi-square tests. Continuous data, which meet normality assumptions, were reported as means ± standard deviations. Repeated measures data for two groups were analyzed using a 2 (Group) × 3 (Time) ANOVA to determine the effects of group factors, within-group differences over time, and interaction effects. When interaction effects between group factors and time factors were present, simple effects analysis was conducted. Data analysis followed the intention-to-treat principle, with a two-tailed p-value of < 0.05 considered statistically significant.

Retention strategies

(1) Throughout the study, stroke survivors and their families were treated with utmost respect, empathy, and encouragement. Follow-up personnel received psychological counseling training to ensure professional and compassionate interactions in all scenarios. (2) A hospital-integrated “Cloud Doctor” platform (enabling direct communication with multidisciplinary healthcare providers) was utilized to deliver disease-specific education, nursing guidance, and psychological support, ensuring seamless post-discharge care. (3) A fixed team of experienced follow-up personnel was established to guarantee communication consistency. All staff underwent standardized training to enhance clinical expertise and communication skills, thereby fostering patient trust. (4) Patient rights were rigorously protected through multi-layered privacy measures, including anonymized data collection, encrypted storage, and restricted access, in compliance with ethical guidelines.

Results

A total of 104 participants were enrolled in the study. One participant was excluded because he moved to another neighborhood, and two participants were excluded due to cognitive impairment (see Fig. 1). Survivor demographic data are presented in Table 2.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Flow chart of participant enrollment.

Table 2 Socio-demographics data comparison between two groups of patients (n, %).
Full size table

Primary outcomes

Comparison of Self-Management behavior scores before and after intervention

The repeated measures ANOVA revealed significant improvements in self-management behaviors across both groups (F = 77.048, P < 0.001), with notable interaction effects between intervention method and time (F = 19.714, P < 0.001). The intervention group demonstrated progressive improvement across all dimensions (P < 0.05), while the control group showed limited advancement in interpersonal management (P = 0.28). See Table 3 for details.

Table 3 Comparison of Self-Management behavior scores at different time points within each group before and after intervention (Score, x ± s).
Full size table

Comparison of Self-Management behavior stages before and after intervention

Stage progression analysis (Table 4) revealed divergent trajectories: by 6 months, 72.5% of intervention participants reached the maintenance stage versus 10.0% in controls (P < 0.001). This contrasted sharply with comparable baseline distributions (P = 0.69 at T0).

Table 4 Comparison of Self-Management behavior stages at different time points before and after intervention in two groups (n, %).
Full size table

Secondary outcomes

Comparison of Self-Efficacy scores at different time points before and after intervention

Significant between-group differences in self-efficacy were observed between pre- and post-intervention periods (F = 3.993, P = 0.05). Additionally, within-group differences in self-efficacy scores across intervention time points demonstrated statistically significant variations (F = 5.900, P < 0.001) (Fig. 2).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Comparison of Self-Efficacy Scores at Different Time Points before and after Intervention in Two Groups.

Comparison of Barthel index scores at different time points before and after intervention

The BI scores of both groups improved significantly over time. Notably, the intervention group demonstrated sustained gains compared to the control group (P < 0.001). A significant interaction effect (F = 14.977, P < 0.001) indicated that the intervention accelerated functional recovery, particularly evident at T2 and T3 (Table 5).

Table 5 Comparison of BI scores at different time points before and after intervention in two groups.
Full size table

Discussion

This study demonstrated that the TTM-based Empowerment intervention significantly improved self-management behaviors, self-efficacy, and activities of daily living in young and middle-aged stroke survivors. These findings are consistent with previous research emphasizing the value of theory-driven interventions in chronic disease management8,9. However, the innovation of this study lies in the integration of TTM with HET to construct a dynamic framework: TTM focuses on the progressive stages of behavioral change, providing a road-map for precise matching of intervention strategies, while HET enhances patients’ autonomous decision-making capabilities through collaborative goal-setting, family involvement, and positive reinforcement. This dual-theory integration addresses the limitations of single-theory applications, with their synergistic effects providing more comprehensive support for sustained behavioral change.

Although both groups demonstrated significant time-effect differences in self-management behaviors, the intervention group exhibited greater magnitude of improvement with sustained benefits over time. This contrasts with findings by Sit JW et al., where Empowerment intervention effects diminished at 6 months27. This discrepancy may stem from fundamental differences in intervention design: unlike single-component Empowerment interventions that wane with reduced supervision, our TTM model-based Empowerment intervention synergistically incorporated behavioral change stages, processes, self-efficacy enhancement, and decisional balance28,29. By systematically addressing behavioral barriers while reinforcing perceived benefits of change, the TTM model-based Empowerment approach may better sustain self-management improvements.

The limited between-group differences observed at early time-points (1 month) may reflect a “delayed-onset effect” inherent to stage-matched interventions. The TTM model-based Empowerment intervention emphasizes gradual progression from cognitive preparation to behavioral implementation, with initial phases prioritizing awareness-building and motivation rather than immediate behavioral change30,31. Future studies could explore the feasibility of incorporating brief, intensive booster sessions during early post-intervention phases to accelerate initial improvements.

Notably, the control group showed no sustained improvement in interpersonal management, a finding potentially attributable to the predominant focus of conventional health education on individual behavior change rather than social interaction support. While the control group received standardized guidance on medication adherence and lifestyle modifications, the intervention lacked structured modules addressing communication skills, social support networks, or community reintegration. Previous studies have emphasized that social interactions and peer support are critical for sustaining self-management behaviors in chronic conditions8,33. In contrast, the intervention group benefited from tailored strategies such as family-mediated goal-setting and role-playing exercises to practice interpersonal scenarios, which may explain their superior outcomes. Future, leveraging digital platforms (e.g., WeChat groups) to facilitate peer interaction and real-time feedback may further reinforce social engagement, thereby bridging the limitations of conventional approaches.

According to Bandura’s social cognitive theory, self-efficacy reflects an individual’s belief in their own capabilities and serves as a core driver of behavioral change32. This study found that the TTM model-based Empowerment intervention significantly enhanced self-efficacy in young and middle-aged stroke survivors, with the intervention group demonstrating early advantages (at 1 month) and maintaining higher levels at 6 months. These results align with the longitudinal study by Chen et al., indicating that empowerment strategies can strengthen patients’ confidence in their abilities through staged goal achievement and positive feedback, thereby producing sustained effects33. In contrast, the lack of significant early improvement in self-efficacy in the control group may stem from the unidirectional knowledge transmission model of conventional health education, which fails to adequately identify stage-specific barriers to behavioral change, resulting in patients receiving information but struggling to translate it into actionable beliefs.

Furthermore, additional evidence demonstrating the superiority of TTM-based Empowerment intervention over conventional health education lies in the significant improvement in BI scores. This finding differs from the study by Studenski et al. on subacute stroke survivor, where BI scores gradually declined post-intervention34. This discrepancy may arise from two factors: first, our intervention integrated psycho-social support modules, whereas Studenski’s program focused solely on exercise rehabilitation, neglecting the impact of psycho-social factors on functional maintenance; second, the staged support mechanism of TTM may have delayed functional deterioration. Therefore, future intervention designs should address multidimensional “biological-psychological-social” targets, particularly by involving patients and their families in rehabilitation planning to ensure goals align with individual lifestyles and family resources, thereby enhancing adherence.

However, the feasibility and scalability of implementing such personalized interventions in diverse clinical settings are crucial for practical application. The current study utilized the WeChat platform for health education delivery and remote follow-up, which reduced the frequency of face-to-face interventions and shows potential for implementation in regions with uneven healthcare resources. However, this approach requires either ensuring patients’ digital literacy or providing basic technical training to bridge potential digital divides. Furthermore, the successful implementation of this intervention relied on multidisciplinary collaboration among physicians, nurses, rehabilitation therapists, and other healthcare professionals. We therefore recommend integrating stage assessment tools into electronic medical record systems within healthcare institutions. This integration would enable real-time tracking of patients’ behavioral stages by the care team and facilitate timely strategy adjustments.

Future research should focus on developing culturally adapted intervention modules tailored to different national contexts while maintaining the core components of the TTM framework. Additionally, comprehensive cost-effectiveness analyses of TTM model-based empowerment interventions across various healthcare environments are needed to inform policy decisions and resource allocation. Such studies should particularly examine implementation outcomes in low-resource settings and evaluate strategies for sustaining intervention effects over a long-term.

Conclusion

This study confirms that the integration of the TTM and HET not only provides a structured “roadmap” for behavioral change but also enhances intervention sustainability through patient agency activation. The balanced combination of “theoretical rigor” and “practical flexibility” offers a replicable paradigm for personalized interventions in chronic disease management. Future research should further explore its applicability across broader populations and cultural contexts, along with cost-effectiveness analyses.

Limitations

First, the intervention in this study employed a multifaceted approach, making it difficult to isolate the specific components responsible for the effectiveness of the TTM-based Empowerment intervention. Second, to ensure smooth implementation, the study excluded patients with language barriers, hearing impairments, or severe motor dysfunction individuals who may actually require greater support in self-management. Third, the WeChat-based intervention may introduce potential selection bias, as participation requires a certain level of digital literacy. This could exclude individuals with limited digital skills, thereby limiting the generalizability of the findings to populations with higher proficiency.

Implications for service planning and health policy

Stroke is an acute event often accompanied by long-term chronic disability consequences. However, many survivors lack of sufficient self-management knowledge, making the transition from hospital to home rehabilitation remarkably challenging. Our findings suggest that TTM-based Empowerment interventions are more effective than single Empowerment interventions in promoting the establishment and maintenance of health behaviors among young and middle-aged stroke survivors, ultimately improving self-efficacy and self-management levels. This suggests that such interventions should be incorporated into the early stages of survivor rehabilitation. Continuous professional support can help stroke survivors learn and adapt to maintaining good self-management behaviors at home, reducing caregiver burden, lowering potential healthcare costs, and conserving social resources. These findings provide a basis for policy development, emphasizing the importance of introducing comprehensive and continuous interventions early in rehabilitation to ensure a smooth transition and sustained recovery outcomes after discharge.