Applying the fear-avoidance model to exercise rehabilitation after coronary artery bypass grafting a nurse-led intervention study Authors: Ling Xia, Jingyu Hua, Liping Xuan Published: 2026 ABSTRACT ObjectiveTo examine whether a nurse-led intervention grounded in the Fear-Avoidance Model (FAM) reduces postoperative kinesiophobia indirectly by improving patients’ perceived physical capability.MethodWe conducted a single-center, parallel-group randomized controlled trial. A total of 122 patients who underwent coronary artery bypass grafting (CABG) were randomized 1:1 to an intervention group or a usual-care control group (n = 61 each). The intervention included graded activity, functional feedback, goal setting, structured education, and post-discharge follow-up. Perceived physical capability and kinesiophobia were assessed 30 days after discharge. Mediation analysis was used to test whether perceived physical capability mediated the effect of the intervention on kinesiophobia.ResultsThe intervention was associated with higher perceived physical capability (a = 2.43, p = 0.031). Perceived physical capability was inversely associated with kinesiophobia (b = − 0.775, p < 0.001). The indirect effect of the intervention on kinesiophobia through perceived physical capability was significant (ab = − 1.879; bootstrapped 95% CI [− 3.702, − 0.288]). In contrast, the total effect (c = − 1.557, p = 0.433) and the direct effect (c′=0.322, p = 0.860) were not statistically significant.ConclusionsThese findings suggest a rehabilitation-relevant pathway in which strengthening perceived capability may help reduce avoidance-oriented responses after CABG. A nurse-led approach that combines graded mobilization, performance feedback, and goal-based progression may support engagement in early recovery activities and may offer a practical framework for transitional rehabilitation support. Introduction Coronary artery bypass grafting (CABG) is often followed by kinesiophobia (fear of movement), which refers to an emotional response, maladaptive beliefs, heightened bodily vigilance, and avoidance behaviors driven by concerns that activity may trigger cardiac discomfort or recurrence [1].Patients after CABG represent a particularly relevant population for this issue because early rehabilitation requires timely mobilization, whereas postoperative sternal pain, chest discomfort, fatigue, and concerns about cardiac safety may lead patients to interpret normal recovery-related sensations as warning signs. This tension between the need for activity and the fear of movement makes CABG patients a clinically meaningful population for examining FAM-informed rehabilitation strategies. Kinesiophobia is closely associated with rehabilitation adherence, physical activity, and quality of life; multiple studies suggest a relatively high prevalence in cardiac surgery populations and links to unfavorable rehabilitation outcomes [2, 3]. Measurement tools for assessing kinesiophobia in cardiovascular populations have been examined, and available evidence suggests that such instruments can support dynamic assessment during cardiac rehabilitation and postoperative recovery [4, 5].Conventional cardiac rehabilitation (CR) may help reduce kinesiophobia, but evidence for theory-driven interventions with clearly specified mechanisms remains limited; prior research suggests that behavioral interventions incorporating components such as graded activity and education may benefit patients by reducing avoidance behaviors [6]. The Fear-Avoidance Model (FAM) describes a maladaptive cycle in which “pain/threat appraisal→fear→avoidance→deconditioning and functional decline” reinforces itself over time. Within this framework, perceived physical capability—an individual’s internal judgment of their own physical state—shapes how symptoms are interpreted as threatening and, in turn, guides subsequent behavioral choices. In patients after CABG, direct evidence using the term “perceived capability” remains limited; however, related studies have shown that self-efficacy, fatigue, kinesiophobia, and physical activity are closely interconnected in this population. These findings suggest that patients’ subjective appraisal of their physical capacity may be an important proximal factor in rehabilitation behavior [7, 8]. Patients with lower perceived capability may be more likely to interpret postoperative sensations as threatening and to remain within the fear-avoidance cycle [9]. Therefore, intervening on perceived capability may be a feasible strategy to disrupt this pathway. Based on this rationale, we targeted patients after CABG and developed a multicomponent intervention grounded in the FAM. By strengthening perceived capability as the entry point, we tested whether the intervention could indirectly reduce kinesiophobia, thereby providing mechanism-oriented evidence to inform postoperative rehabilitation nursing. Materials and methods Research design This study was a single-center, parallel-group randomized controlled trial (RCT) with a 1:1 allocation ratio. Participants were patients who underwent CABG at Anhui Chest Hospital between June 2024 and June 2025 and met the eligibility criteria. All patients were enrolled after their postoperative condition was stable and written informed consent had been obtained. Routine postoperative mobilization followed the standard care pathway in both groups. For patients allocated to the intervention group, the FAM-oriented enhanced components were introduced after enrollment and then continued through discharge and the 30-day post-discharge period.The study protocol was approved by the Ethics Committee of Anhui Chest Hospital (KJ2024-009). After consent and confirmation of eligibility, participants were assigned a study number by the research coordinator in the order of enrollment. Group allocation was determined by the parity of the assigned number: odd numbers were allocated to the intervention group and even numbers to the control group (1:1). To ensure allocation concealment, the allocation result was disclosed only after enrollment was completed; the nurses delivering the intervention were unaware of group assignment before allocation. The generation and custody of the allocation list were managed by an independent staff member, and intervention providers were not involved in sequence generation or assignment. Eligibility criteria Inclusion criteria Age ≥ 18 years;First-time CABG with stable vital signs after surgery;Able to communicate and understand instructions, and able to complete questionnaire assessments;Willing to participate and complete the 30-day post-discharge follow-up. Age ≥ 18 years; Age ≥ 18 years; First-time CABG with stable vital signs after surgery; First-time CABG with stable vital signs after surgery; Able to communicate and understand instructions, and able to complete questionnaire assessments; Able to communicate and understand instructions, and able to complete questionnaire assessments; Willing to participate and complete the 30-day post-discharge follow-up. Willing to participate and complete the 30-day post-discharge follow-up. Exclusion criteria Severe perioperative complications or reoperation, defined as conditions that significantly impair participation in rehabilitation or follow-up, such as hemodynamic instability requiring prolonged intensive care, reoperation, or severe functional deterioration;Severe cognitive or psychiatric disorders that would impair participation;Severe functional limitation that precludes participation in the training program;Refusal of follow-up or inability to complete follow-up assessments.Common postoperative events that were clinically managed and did not interfere with rehabilitation or follow-up were not considered withdrawal criteria. Severe perioperative complications or reoperation, defined as conditions that significantly impair participation in rehabilitation or follow-up, such as hemodynamic instability requiring prolonged intensive care, reoperation, or severe functional deterioration; Severe perioperative complications or reoperation, defined as conditions that significantly impair participation in rehabilitation or follow-up, such as hemodynamic instability requiring prolonged intensive care, reoperation, or severe functional deterioration; Severe cognitive or psychiatric disorders that would impair participation; Severe cognitive or psychiatric disorders that would impair participation; Severe functional limitation that precludes participation in the training program; Severe functional limitation that precludes participation in the training program; Refusal of follow-up or inability to complete follow-up assessments. Refusal of follow-up or inability to complete follow-up assessments. Common postoperative events that were clinically managed and did not interfere with rehabilitation or follow-up were not considered withdrawal criteria. Common postoperative events that were clinically managed and did not interfere with rehabilitation or follow-up were not considered withdrawal criteria. Sample size The sample size was determined based on consecutive enrollment of eligible patients during the study period. For reference, in studies comparing categorical outcomes between two groups, a sample size of approximately 100–120 participants is generally considered sufficient to detect medium effect sizes (Cohen’s w ≈ 0.30) at a significance level of 0.05 with 80% power. The final sample size of 122 participants in this study is consistent with this range. Intervention and control Intervention group (FAM-oriented multicomponent program) The intervention targeted perceived physical capability as the entry point. Some mobilization-related activities were also part of routine postoperative care; however, in the intervention group these activities were embedded within a structured FAM-informed framework emphasizing symptom reinterpretation, capability-focused feedback, goal-based progression, and continuity of support after discharge.Drawing on the core components of cardiac rehabilitation defined by the British Association for Cardiovascular Prevention and Rehabilitation (BACPR) [10], the program integrated graded activity exposure, capability-focused feedback, collaborative goal setting, and education with follow-up. These components were intended to intervene in the fear-avoidance loop described by the Fear-Avoidance Model (FAM), strengthen perceived physical capability, and thereby indirectly reduce kinesiophobia. The intervention was delivered by trained specialist nurses after patients had become clinically stable, completed informed consent, and were enrolled in the study, and continued through discharge and the 30-day post-discharge period. Details are presented in Table 1. A schematic overview of intervention delivery and outcome assessment timing is shown in Fig. 1. Table 1FAM-oriented multicomponent intervention (Intervention group)PhaseKey contentDeliveryFrequency/durationNotesEarly graded exercise training [11]Bed-/bedside joint range-of-motion and low-intensity resistance→sitting up/transfer→ indoor and ward ambulationProgressed stepwise along the routine postoperative recovery pathway and was reinforced within the FAM-oriented intervention after enrollmentGradually increased based on individual statusKeep intensity within Borg Rating of Perceived Exertion (Borg RPE; 6–20 scale) at 11–13 (“somewhat light to somewhat hard”) [12]; monitor heart rate, oxygen saturation, and symptom tolerance dynamically; stop training and reassess when needed.Functional feedback and enhancement of perceived capability [13]Record walking time/distance, peak heart rate, and Borg score after each session; provide visual feedback on “today’s improvement compared with yesterday.”Immediate feedback by the primary nurseAfter each training sessionReduce catastrophic expectations and strengthen perceived capability [14].Goal-oriented motivation [15]Set short-term goals jointly (e.g., number of walking laps, duration); provide immediate reinforcementGoals set jointly by nurse and patientAssessed once dailyReinforce positive experiences; adjust the exercise prescription according to tolerance. Pre-discharge rehabilitation education Education content: exercise safety, risks and benefits, graded exposure, self-monitoring; correct misconceptions such as “activity will harm the heart.”Group teaching + individual coachingOnce before discharge, (30 min)Provide structured education and a home exercise prescription. Post-discharge follow-up Telephone/WeChat follow-up once weekly; verify adherence and symptom feedback; provide individualized guidanceDelivered by specialist nurses4 weeks after dischargeAdjust dynamically according to goal attainment.Some early mobilization activities were consistent with routine postoperative rehabilitation; the distinctive features of the intervention were the structured FAM-informed delivery, capability-focused feedback, collaborative goal setting, interpretive communication regarding symptoms and safety, and planned post-discharge follow-up Table 1FAM-oriented multicomponent intervention (Intervention group)PhaseKey contentDeliveryFrequency/durationNotesEarly graded exercise training [11]Bed-/bedside joint range-of-motion and low-intensity resistance→sitting up/transfer→ indoor and ward ambulationProgressed stepwise along the routine postoperative recovery pathway and was reinforced within the FAM-oriented intervention after enrollmentGradually increased based on individual statusKeep intensity within Borg Rating of Perceived Exertion (Borg RPE; 6–20 scale) at 11–13 (“somewhat light to somewhat hard”) [12]; monitor heart rate, oxygen saturation, and symptom tolerance dynamically; stop training and reassess when needed.Functional feedback and enhancement of perceived capability [13]Record walking time/distance, peak heart rate, and Borg score after each session; provide visual feedback on “today’s improvement compared with yesterday.”Immediate feedback by the primary nurseAfter each training sessionReduce catastrophic expectations and strengthen perceived capability [14].Goal-oriented motivation [15]Set short-term goals jointly (e.g., number of walking laps, duration); provide immediate reinforcementGoals set jointly by nurse and patientAssessed once dailyReinforce positive experiences; adjust the exercise prescription according to tolerance. Pre-discharge rehabilitation education Education content: exercise safety, risks and benefits, graded exposure, self-monitoring; correct misconceptions such as “activity will harm the heart.”Group teaching + individual coachingOnce before discharge, (30 min)Provide structured education and a home exercise prescription. Post-discharge follow-up Telephone/WeChat follow-up once weekly; verify adherence and symptom feedback; provide individualized guidanceDelivered by specialist nurses4 weeks after dischargeAdjust dynamically according to goal attainment.Some early mobilization activities were consistent with routine postoperative rehabilitation; the distinctive features of the intervention were the structured FAM-informed delivery, capability-focused feedback, collaborative goal setting, interpretive communication regarding symptoms and safety, and planned post-discharge follow-up FAM-oriented multicomponent intervention (Intervention group) Some early mobilization activities were consistent with routine postoperative rehabilitation; the distinctive features of the intervention were the structured FAM-informed delivery, capability-focused feedback, collaborative goal setting, interpretive communication regarding symptoms and safety, and planned post-discharge follow-up Fig. 1Timeline of intervention delivery and outcome assessment Timeline of intervention delivery and outcome assessment Control group (usual care) Participants in the control group received routine postoperative rehabilitation, including mobilization as prescribed by the medical team, general activity guidance from ward nurses, routine monitoring for postoperative complications, and standard discharge education. These elements reflected standard ward-based postoperative care in our setting. However, usual care did not include structured capability-focused feedback, collaborative goal setting, explicit reinterpretation of postoperative symptoms and safety boundaries, or a planned post-discharge follow-up program. These latter elements constituted the main theory-informed additions of the FAM-oriented intervention. Outcomes and measurement Primary outcomes ① Kinesiophobia (Y) Kinesiophobia was the primary outcome (Y) in this study and was assessed using an 11-item questionnaire based on the content of the Tampa Scale for Kinesiophobia (TSK-11) items. In the implementation of this study, a 5-point response format (1–5) was used to capture more nuanced gradations of postoperative kinesiophobia, yielding a total score range of 11–55, with higher scores indicating greater kinesiophobia and a stronger avoidance tendency. For descriptive group comparisons, kinesiophobia levels were categorized into low, moderate, and high based on tertiles of the total score distribution.For mediation analysis, the total score was used as a continuous outcome variable. This study-administered version showed good internal consistency in the present sample (Cronbach’s α = 0.879). Given that its scoring format differs from more commonly used applications, the findings are primarily interpreted for within-study group comparisons and exploratory mechanism analysis [16]. ② Perceived physical capability (M) Perceived physical capability was modeled as the mediator (M) and operationalized as self-reported physical function measured by the PROMIS Physical Function short form (PF-10a). The PF-10a includes 10 items rated on a 5-point scale. Raw scores were converted to standardized T-scores according to the PROMIS scoring manual; higher T-scores indicate better perceived physical function (i.e., stronger perceived capability). This measure has demonstrated good reliability (Cronbach’s α ≈ 0.92) and construct validity in cardiovascular disease and postoperative rehabilitation populations and supports both cross-sectional comparisons and longitudinal monitoring [17]. Secondary outcomes Secondary outcomes included length of hospital stay, ejection fraction (EF) at discharge, pain score within 24 h after transfer from the ICU to the general ward (Numeric Rating Scale, NRS 0–10), and activities of daily living (ADL) at discharge, assessed using the Activities of Daily Living (ADL) scale [18]. Pain within 24 h after ICU transfer was included as an auxiliary early in-hospital recovery indicator rather than a core intervention-sensitive outcome. Data collection and quality control Due to the nature of the intervention, blinding of participants and intervention nurses was not feasible. To minimize measurement bias, the 30-day post-discharge outcomes were assessed via telephone interview by trained research nurses who were not involved in intervention delivery and were blinded to group allocation. A standardized follow-up script and response prompts were used. Patient identity was verified before each call. Data recording and entry were performed by two staff members independently, followed by cross-checking. Statistical analyses were conducted by another researcher using de-identified and coded datasets. No participants were lost to follow-up. All enrolled participants (n = 122) completed the 30-day post-discharge assessment (follow-up rate: 100%). This high retention rate may be attributable to the relatively short follow-up period, the single-center design, and the use of structured telephone follow-up conducted by trained nursing staff. Statistical analysis All analyses were performed using SPSS version 26.0. Continuous variables were assessed for normality prior to analysis. Normally distributed continuous variables are presented as mean (standard deviation) and were compared between groups using independent-samples t tests. Non-normally distributed continuous variables are presented as median (interquartile range, IQR) and were compared using the Mann–Whitney U test. Categorical variables are presented as frequency (percentage) and were compared using the chi-square test. Fisher’s exact test was used when expected cell counts were < 5 or when any expected count was < 1. Effect size for contingency analyses was reported as Cramér’s V (range 0–1), with larger values indicating greater between-group differences. To examine whether perceived physical capability mediated the association between the intervention and kinesiophobia, a simple mediation model was tested using the Hayes PROCESS macro (Model 4). Indirect effects were estimated using bias-corrected bootstrap resampling (5,000 samples) with 95% confidence intervals. Total, direct, and indirect effects were reported, along with the proportion mediated (indirect effect / total effect). All tests were two-sided, with a significance level of 0.05. Result Comparison of baseline characteristics This randomized controlled trial enrolled 122 patients after CABG and randomized them 1:1 to the intervention or control group (n = 61 per group). All participants completed the 30-day follow-up and were included in the analysis. Participant flow is shown in Fig. 2. Fig. 2CONSORT flow diagram of participant recruitment, randomization, follow-up, and analysis CONSORT flow diagram of participant recruitment, randomization, follow-up, and analysis There were no statistically significant between-group differences in baseline and admission-related characteristics, including age, sex and education level, and ejection fraction at admission (all p > 0.05), indicating general comparability between the two groups at study entry (Table 2). Because the “bachelor’s degree or above” category yielded sparse cells in the control group, Fisher’s exact test was used for education level, no between-group difference was detected (p = 0.435). In-hospital recovery indicators, including length of hospital stay, ejection fraction at discharge, pain score within 24 h after transfer from the ICU to the general ward, and ADL at discharge, are presented separately in Table 3. Table 2Baseline and admission-related characteristics (n = 122)VariableControl (n = 61)Intervention (n = 61)satisticP valueAge (years)65.28 ± 8.7864.10 ± 9.06t = 0.7310.466Sex, n(%)Male44(72.1%)47(77.0%)χ²= 0.3890.533Female17(27.9%)14(23.0%)Education level, n (%)Fisher0.435Illiterate11(18.0%)6(9.8%)Primary school14(23.0%)18(29.5%)Junior/Senior high school31(50.8%)29(47.5%)Technical secondary school / junior college5(8.2%)6(9.8%)Bachelor’s degree or above02(3.3%)Ejection fraction at admission (%)53.93 ± 10.2250.74 ± 7.85t = 1.9370.055 Table 2Baseline and admission-related characteristics (n = 122)VariableControl (n = 61)Intervention (n = 61)satisticP valueAge (years)65.28 ± 8.7864.10 ± 9.06t = 0.7310.466Sex, n(%)Male44(72.1%)47(77.0%)χ²= 0.3890.533Female17(27.9%)14(23.0%)Education level, n (%)Fisher0.435Illiterate11(18.0%)6(9.8%)Primary school14(23.0%)18(29.5%)Junior/Senior high school31(50.8%)29(47.5%)Technical secondary school / junior college5(8.2%)6(9.8%)Bachelor’s degree or above02(3.3%)Ejection fraction at admission (%)53.93 ± 10.2250.74 ± 7.85t = 1.9370.055 Baseline and admission-related characteristics (n = 122) Table 3In-hospital recovery indicators (n = 122)VariableControl (n = 61)Intervention(n = 61)satisticP valueLength of hospital stay (days)18.56 ± 10.2318.03 ± 6.18t = 0.3430.732Ejection fraction at discharge (%)53.03 ± 6.4651.79 ± 5.12t = 1.1810.240Pain score within 24 h after transfer from ICU to general ward (NRS 0–10)2.23 ± 0.762.52 ± 1.18t = -1.6430.103ADL score at discharge79.75 ± 15.1585.25 ± 6.42t = -2.607<0.01Continuous variables are presented as mean±standard deviation (SD) and were compared between groups using independent-samples t tests. Categorical variables are presented as number (%) and were compared using the chi-square test. For education level, Fisher’s exact test was used due to expected cell counts < 5. All tests were two-sided with a significance level of p < 0.05 Table 3In-hospital recovery indicators (n = 122)VariableControl (n = 61)Intervention(n = 61)satisticP valueLength of hospital stay (days)18.56 ± 10.2318.03 ± 6.18t = 0.3430.732Ejection fraction at discharge (%)53.03 ± 6.4651.79 ± 5.12t = 1.1810.240Pain score within 24 h after transfer from ICU to general ward (NRS 0–10)2.23 ± 0.762.52 ± 1.18t = -1.6430.103ADL score at discharge79.75 ± 15.1585.25 ± 6.42t = -2.607<0.01Continuous variables are presented as mean±standard deviation (SD) and were compared between groups using independent-samples t tests. Categorical variables are presented as number (%) and were compared using the chi-square test. For education level, Fisher’s exact test was used due to expected cell counts < 5. All tests were two-sided with a significance level of p < 0.05 In-hospital recovery indicators (n = 122) Continuous variables are presented as mean±standard deviation (SD) and were compared between groups using independent-samples t tests. Categorical variables are presented as number (%) and were compared using the chi-square test. For education level, Fisher’s exact test was used due to expected cell counts < 5. All tests were two-sided with a significance level of p < 0.05 Primary outcomes at 30 days after discharge: perceived physical capability and kinesiophobia Perceived physical capability was originally categorized into five levels ranging from severe limitation to above-average function. Due to small counts in some categories, levels were combined for analysis (Table 4) to improve statistical power and interpretability. The distribution of perceived physical capability differed significantly between groups, with a higher proportion of “good function” in the intervention group compared with the control group(88.5% vs. 75.4%), suggestinga more favorable perceived capability profile at 30 days after discharge. Kinesiophobia was categorized into three tertile-based levels (low/moderate/high). The distribution of kinesiophobia levels differed between groups (χ²=7.218, p = 0.027; Cramér’s V = 0.243), with a lower proportion of high kinesiophobia in the intervention group (24.6% vs. 44.3%) and a higher proportion of moderate kinesiophobia (41.0% vs. 21.3%) compared with the control group (Table 4). Table 4Outcomes at 30 days after dischargeVariableControl (n = 61)Intervention (n = 61)StatisticP valueEffect sizePerceived physical capability level, n (%)χ²=9.230<0.01Cramér’sV = 0.275Level 1–38(13.1%)0Level 47(11.5%)7(11.5%)Level 546(75.4%)54(88.5%)Kinesiophobia level (tertiles), n (%)χ²=7.218<0.05Cramér’sV = 0.243Level 121(34.4%)21(34.4%)Level 213(21.3%)25(41.0%)Level 327(44.3%)15(24.6%)Ordinal variables are presented as number (%). Between-group comparisons were performed using the chi-square test. Fisher’s exact test was used when expected cell counts were < 5. Cramér’s V was reported as the effect size, with larger values indicating greater between-group differences Table 4Outcomes at 30 days after dischargeVariableControl (n = 61)Intervention (n = 61)StatisticP valueEffect sizePerceived physical capability level, n (%)χ²=9.230<0.01Cramér’sV = 0.275Level 1–38(13.1%)0Level 47(11.5%)7(11.5%)Level 546(75.4%)54(88.5%)Kinesiophobia level (tertiles), n (%)χ²=7.218<0.05Cramér’sV = 0.243Level 121(34.4%)21(34.4%)Level 213(21.3%)25(41.0%)Level 327(44.3%)15(24.6%)Ordinal variables are presented as number (%). Between-group comparisons were performed using the chi-square test. Fisher’s exact test was used when expected cell counts were < 5. Cramér’s V was reported as the effect size, with larger values indicating greater between-group differences Outcomes at 30 days after discharge Ordinal variables are presented as number (%). Between-group comparisons were performed using the chi-square test. Fisher’s exact test was used when expected cell counts were < 5. Cramér’s V was reported as the effect size, with larger values indicating greater between-group differences Mediation analysis: indirect role of perceived physical capability A simple mediation model (Hayes PROCESS Model 4) was used to test whether the intervention (X) influenced kinesiophobia (Y) indirectly through perceived physical capability (M). The intervention was positively associated with perceived physical capability (a = 2.4262, p = 0.0305), and perceived physical capability was inversely associated with kinesiophobia after adjusting for group assignment (b = − 0.7746, p < 0.001). The bootstrap-estimated indirect effect was significant (ab = − 1.8793; bootstrapped 95% CI [− 3.7017, − 0.2879]). In contrast, neither the total effect (c = − 1.5574, p = 0.4328) nor the direct effect (c’= 0.3220, p = 0.8603) was statistically significant (Table 5; Fig. 3). Together, these findings suggest that the association between the intervention and kinesiophobia operated primarily through the capability pathway within the 30-day follow-up window. Table 5Mediation analysis resultsSectionPath/effectEffect (b)SEt p 95% CIInferencePath coefficientsa: X→M2.42621.10792.19000.0305[0.2327, 4.6198]SignificantPath coefficientsb: M→Y-0.77460.1475-5.2502<0.001[-1.0667, -0.4825]SignificantEffect decompositionc: Total effect-1.55741.9788-0.78700.4328[-5.4753, 2.3605]Not significantEffect decompositionc’: Direct effect0.32201.82600.17630.8603[-3.2936, 3.9375]Not significantEffect decompositionab: Indirect effect-1.8793----Significant(1) X=Group (1 = control,2 = intervention); M=Perceived capability score; Y =Kinesiophobia total score(2) The indirect effect was evaluated using bias-corrected bootstrap confidence intervals; significance is inferred when the 95% CI does not include zero(3) Because c’ and ab were in opposite directions (inconsistent mediation), we did not report the proportion mediated Table 5Mediation analysis resultsSectionPath/effectEffect (b)SEt p 95% CIInferencePath coefficientsa: X→M2.42621.10792.19000.0305[0.2327, 4.6198]SignificantPath coefficientsb: M→Y-0.77460.1475-5.2502<0.001[-1.0667, -0.4825]SignificantEffect decompositionc: Total effect-1.55741.9788-0.78700.4328[-5.4753, 2.3605]Not significantEffect decompositionc’: Direct effect0.32201.82600.17630.8603[-3.2936, 3.9375]Not significantEffect decompositionab: Indirect effect-1.8793----Significant(1) X=Group (1 = control,2 = intervention); M=Perceived capability score; Y =Kinesiophobia total score(2) The indirect effect was evaluated using bias-corrected bootstrap confidence intervals; significance is inferred when the 95% CI does not include zero(3) Because c’ and ab were in opposite directions (inconsistent mediation), we did not report the proportion mediated Mediation analysis results (1) X=Group (1 = control,2 = intervention); M=Perceived capability score; Y =Kinesiophobia total score (2) The indirect effect was evaluated using bias-corrected bootstrap confidence intervals; significance is inferred when the 95% CI does not include zero (3) Because c’ and ab were in opposite directions (inconsistent mediation), we did not report the proportion mediated Fig. 3Mediation model of perceived physical capability in the association between intervention group and kinesiophobia. Unstandardized coefficients are shown.Indirect effect was tested using bootstrap (5,000). * p < 0.05;** p < 0.01༛*** p < 0.001 Mediation model of perceived physical capability in the association between intervention group and kinesiophobia. Unstandardized coefficients are shown.Indirect effect was tested using bootstrap (5,000). * p < 0.05;** p < 0.01༛*** p < 0.001 Discussion Main findings in relation to the FAM In this study of patients after CABG, a nurse-led, FAM-oriented intervention was associated with higher perceived physical capability and a more favorable kinesiophobia profile at 30 days after discharge. Mediation analysis further suggested that the association with kinesiophobia operated primarily through perceived physical capability (i.e., a significant indirect effect), whereas the total and direct effects were not statistically significant. The FAM proposes that catastrophic interpretation of threat-related cues triggers fear and avoidance, which in turn may contribute to deconditioning and further reinforces fear, forming a vicious cycle; cognitive–behavioral strategies and graded exposure can disrupt this cycle [19].Our findings align with this core mechanism and suggest that the FAM is applicable in the postoperative CABG context. Importantly, the results provide empirical support for an intervention pathway that targets perceived capability as a practical entry point [2]. It should be emphasized that the practical value of this study is not to repeatedly “re-validate” the FAM, but to translate patients’ common concerns into assessable, monitorable, and adjustable behavioral rules. Specifically, when patients can reinterpret sensations such as chest tightness, incision pulling, and fatigue as expected and manageable recovery experiences—rather than equating them with danger or deterioration—their avoidance tendency may loosen accordingly. Accordingly, this study addresses a pragmatic nursing question: how can nurses help patients “dare to start and be willing to continue” rehabilitation activities, rather than simply telling them that they “should exercise.” These findings do not suggest a need to revise the core structure of the FAM. Rather, they indicate that, in the CABG context, the model may be best operationalized by emphasizing perceived capability as a proximal nursing target, helping patients reinterpret symptoms, clarify safety boundaries, and set achievable activity goals. Perceived capability as a key lever In the mediation model, the indirect pathway via perceived physical capability (perceived capability) was significant, whereas neither the total effect nor the direct effect on kinesiophobia reached statistical significance within the 30-day follow-up window. This pattern suggests that the intervention may first influence avoidance-oriented responses by changing patients’ judgments about their capability and safety boundaries, while a measurable overall shift in fear scores may not fully emerge in the short term. Self-efficacy theory suggests that strengthening the belief of “I can do it” helps reduce fear and sustain behavioral engagement [20].In postoperative rehabilitation, this belief is often built through a sense of control and mastery experiences, and can be reinforced through visible positive feedback [21].Prior reviews also indicate that systematic cardiac rehabilitation can reduce kinesiophobia scores, and that improvements in self-efficacy frequently co-occur with reductions in fear [22, 23]. Rogers et al. reported that enhancing self-efficacy can interrupt the negative loop of “fear-avoidance-functional decline” [23].Taken together with our findings, the nurse-led combination of graded exposure, capability-focused feedback, and goal management may operate through a capability-driven pathway-strengthening perceived capability, supporting continued participation, and thereby weakening avoidance tendency—an effect that is more consistently reflected as an indirect effect. Clinically, perceived capability may be a more proximal and actionable target: it directly shapes daily rehabilitation decisions and is more readily influenced by nursing actions (e.g., demonstration, supervised practice, feedback, and correction). Graded exercise, goal setting, functional feedback, self-monitoring, and follow-up are commonly used strategies that may strengthen perceived capability or related constructs such as exercise self-efficacy [9]. In CABG patients, their value may lie in making functional progress visible and helping patients reinterpret postoperative symptoms within safe activity boundaries. In contrast, fear ratings are influenced by multiple information sources (e.g., pain, concerns about complications, and prior experiences) and may not decline in a linear or stable manner during early recovery. Therefore, rather than interpreting a non-significant direct effect as “the intervention was insufficient,” it may be more appropriate to view it as a process in which the intervention first promotes a reappraisal of capability and safety boundaries; fear relief may gradually emerge as mastery experiences accumulate. This interpretation is consistent with the realities of early postoperative rehabilitation. Accordingly, nursing assessment may be shifted slightly upstream. Beyond fear screening, nurses may focus on whether patients have developed stable self-monitoring and whether they can adopt appropriate coping responses when mild discomfort occurs rather than stopping activity immediately and avoiding exercise. These process indicators may better reflect whether the intervention is truly taking effect. Practical implications of a nurse-led pathway with early mobilization and follow-up Both international and Chinese guidelines emphasize multidisciplinary cardiac rehabilitation, in which nurses play key roles in assessment, education, follow-up, and quality improvement, and are central to implementing rehabilitation pathways [24, 25].The present intervention covered inpatient functional training and post-discharge follow-up support, mapping onto core components such as assessment, education, exercise, psychological/behavioral support, outcome evaluation, and longer-term planning. This alignment suggests the potential feasibility of a nurse-driven rehabilitation approach. To prevent “nurse-led” from remaining a conceptual label, the intervention could be embedded into a standardized transitional rehabilitation workflow. The core is not to increase nursing workload, but to standardize critical clinical actions: clarifying safety boundaries, breaking rehabilitation tasks into manageable steps, making functional feedback explicit, and structuring follow-up prompts. Multiple RCTs suggest that home-based or remote cardiac rehabilitation is not inferior to center-based programs in outcomes such as mortality, exercise capacity, and quality of life, and may improve accessibility and transitional care [26].Among older CABG populations, transitional remote rehabilitation can also improve self-care and quality of life; wearable devices and mobile approaches can provide real-time functional feedback and strengthen adherence [27].This is consistent with our emphasis on “making progress visible” through capability feedback and goal management, supporting a scalable strategy of nurse leadership, early mobilization, and continuous follow-up. More broadly, the key advantage of home-based or remote rehabilitation may lie less in changing the setting than in sustaining rehabilitation support and accessibility after discharge. By using standardized communication and follow-up, nurses may help support the transition from passive compliance toward more active self-management [28, 29]. For phase II cardiac rehabilitation, the findings suggest value in assessing not only exercise capacity but also perceived capability and activity-related concerns. A brief FAM-informed module involving symptom reinterpretation, safety-boundary clarification, functional feedback, and short-term goal setting could be integrated into routine rehabilitation to support confidence and reduce avoidance-oriented responses. A further limitation is the lack of baseline kinesiophobia assessment, which limits evaluation of pre-intervention equivalence and change over time. Future multicenter studies with longer follow-up and the inclusion of objective functional assessments and adherence indicators are warranted to further validate the intervention effects and underlying mechanisms.In addition, the study-administered kinesiophobia measure used a scoring format that differs from more commonly used applications, which may limit direct comparability with studies using standard scoring approaches. Conclusion Based on the Fear-Avoidance framework, this study developed a nurse-led rehabilitation pathway for patients after CABG and suggests that perceived capability—operationalized in this study as perceived physical capability—has important relevance for modulating avoidance-oriented responses related to kinesiophobia. The findings are consistent with the applicability of the FAM in early postoperative cardiac rehabilitation and highlight the practical value of designing interventions around perceived capability through components, such as graded exposure, functional feedback, and goal management. This study provides preliminary evidence for a nurse-led, FAM-informed rehabilitation approach in patients undergoing CABG. The findings suggest that improvements in perceived physical capability may be associated with reduced kinesiophobia in the early postoperative recovery period. These results highlight a potential behavioral pathway that may inform the design of rehabilitation strategies, while further studies with larger samples and longitudinal designs are needed to confirm these findings.