Introduction

Chronic obstructive pulmonary disease (COPD) is a prevalent chronic respiratory disease and is regarded as the third leading cause of death globally1,2. According to the China Pulmonary Health study, the overall prevalence of COPD in China was 8.6%, which translated to approximately 99.9 million people with COPD in China3. Progressive airflow restriction and severe airway obstruction diminish patients’ lung function and physical activity (PA). Due to activity-related breathlessness and reduced exercise tolerance, the majority of patients with COPD are usually compelled to decrease their PA and adopt a sedentary lifestyle4,5. Physical inactivity is notably prevalent among individuals with COPD6,7, posing a significant risk factor for adverse outcomes, including early mortality, particularly in patients with chronic diseases8,9. One study revealed that two-thirds of elderly patients with COPD reported low levels of PA, with more severe dyspnea and the presence of depression being independently associated with this reduced activity10. While it is crucial to enhance PA levels in individuals with COPD, achieving this remains a challenging task.

Background

Kinesiophobia is defined as an irrational and debilitating fear of movement, while dyspnea-related kinesiophobia refers to the excessive avoidance of physical activities that cause or may cause dyspnea11. The involvement of emotion-related areas such as the insula, anterior cingulate cortex, and amygdala during dyspnea anticipation most likely reflects anticipatory fear in patients, leading to activity avoidance, exacerbating breathlessness, and further affecting the quality of life12,13. A recent study reported that more than 90% of people with COPD had a high degree of kinesiophobia14, and an increase in dyspnea-related kinesiophobia was associated with lower PA levels in these individuals15. According to studies, dyspnea-related kinesiophobia in patients with COPD further limit their daily activities, exacerbate the degree of dyspnea, negatively affect lung function rehabilitation, and create a vicious cycle that ultimately impacts their quality of life16. Given the detrimental effects of kinesiophobia on PA and exercise behaviors, it is essential to address this issue in patients with COPD.

According to Social Cognitive and Self-Efficacy Theories, as sentient beings, we humans can interact with our environment and regulate our thoughts and behavior (Self-Regulation)17. The stronger our belief is in our abilities (Self-Efficacy), the more likely we are to regulate and adjust our thoughts and behavior in ways that help us persist through challenges, overcome obstacles, and achieve desirable results (Self-Regulatory Efficacy)18. Exercise self-regulatory efficacy is an individual’s confidence to exercise regularly despite facing difficulties or impediments17,19. This form of efficacy has been reported to positively influence adherence to regular exercise in patients with COPD and other chronic diseases20,21. Consequently, we hypothesized that fear of movement may lead to reduced exercise self-regulatory efficacy, further contributing to physical inactivity.

Cognitive evaluation theory highlights the importance of motivation regarding behavior, its perceived value as a reward, and its relevance22. In other words, whether individuals with COPD engage in or maintain regular PA may depend on the balance between their perception of exercise benefits and barriers. The perception of exercise benefits encompasses promoting physical and psychological health and boosting social interaction, while the perception of exercise barriers involves physical exertion and lack of family support23. One study showed that when patients perceived more benefits of exercise, they had more confidence in engaging in rehabilitation practice24. The existing literature review demonstrated that dyspnea- related kinesiophobia, exercise perception, and exercise self-regulatory efficacy were associated with exercise adherence. However, to our knowledge, none have examined the interrelationships among all four constructs in individuals with COPD. Clarifying and understanding the mechanisms underlying the strong correlations among these psychosocial variables and health-promoting behaviors could assist in tailoring effective measures and interventions aimed at promoting exercise.

In this study, a moderated mediation, also referred to as conditional indirect effects, occurs when the effect of an independent variable (A) on an outcome variable (C) via a mediator variable B differs depending on the levels of a moderator variable (D). Specifically, the effect of A on B, the effect of A on C, and/or the effect of B on C depends on the level of D25. The model aimed to examine the impact of exercise self-regulatory efficacy on the association between dyspnea-related kinesiophobia and PA, and then assess the moderating effects of exercise perception on the relationship between dyspnea-related kinesiophobia and exercise self-regulatory efficacy. The moderated mediation effect model was created to test the hypotheses put forward in line with the purpose of the research, as shown in Fig. 1.

Fig. 1
figure 1

Hypothesized theoretical model.

Full size image

Methods

Study design

A cross-sectional study was conducted. Data were gathered from three hospitals in Zhejiang Province, China between December 2023 and May 2024. This study was conducted in accordance with guidelines of the Declaration of Helsinki and approved by the Institutional Research Committees of the First Affiliated Hospital of Zhejiang University, the Affiliated Hospital of Hangzhou Normal University, and Haining People’s Hospital. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement was followed when producing the manuscript (File S1)26.

Research samples

Patients with COPD were selected as the study participants using a convenience sampling method. Study participants were informed of the purpose and process of the study prior to participation and of their right to withdraw at any time. The inclusion criteria for participants were: (i) the diagnosis of stable patients with COPD according to the Global Initiative for Chronic Obstructive Lung Disease (2022) guidelines, with a postbronchodilator ratio of the forced expiratory volume in one second (FEV1) to the forced vital capacity (FVC) < 0.70. The exclusion criteria were: (i) long-term bedridden status or wheelchair use; (ii) inability to perform PA due to non-respiratory disease; (iii) inability to complete the questionnaire or answer questions independently; and (iv) participation in other intervention studies. The sample size estimation was based on the opinion of Kline that the sample size that is acceptable for research using structural equation model is approximately 200 cases27.

Tools

Dyspnea-related kinesiophobia

The independent variable was dyspnea-related kinesiophobia, assessed using the Breathlessness Beliefs Questionnaire (BBQ), translated into Chinese by Wu Q et al.28 The questionnaire comprises two dimensions: somatic focus (BBQ-SF, 5 items) and activity avoidance (BBQ-AA, 6 items). High scores on the BBQ-SF scale indicate the belief that breathlessness signifies a dangerous disease process, while high scores on the BBQ-AA scale reflect the belief that PA causing breathlessness should be avoided as it might exacerbate the underlying condition. We specifically focused on the BBQ-AA subscale, to estimate dyspnea-related kinesiophobia in people with COPD. Each item in the subscale is rated on a 5-point scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Based on a previous study, we considered a total score exceeding 6 on the BBQ-AA subscale indicative of the presence of dyspnea-related kinesiophobia11. The Cronbach’s alpha of the subscale was 0.81.

Exercise self-regulatory efficacy

The Exercise Self-Regulatory Efficacy Scale (EX-SRES), developed by Davis et al.18, serves as an effective tool for assessing exercise efficacy beliefs in individuals with COPD. Translated into Chinese by Tsai et al.29, the scale has 16 items, which reflect the confidence of patients with COPD to continue exercising under conditions of bad weather, pain, exercise alone, busyness, no support from others, hypoxia, and fatigue. Each item is rated on a scale from 0 to 10, where 0 means “no confidence” and 10 means “very confident”. The total score is the sum of the scores of 16 items, and a total score < 53 is classified as low-level exercise self-efficacy, a total score ranging from 53 to 106 indicates moderate exercise self-efficacy, while a total score > 106 is classified as high exercise self-efficacy. The higher the score is, the higher the confidence that individuals with COPD can adhere to exercise. The Chinese version of the instrument had an overall Cronbach’s alpha of 0.925.

Exercise perception

The exercise perception was assessed using the Exercise Benefits/Barriers scale (EBBS) developed by Sechrist et al.23. A Chinese version of this scale has been previously used in patients with COPD, demonstrating good internal consistency15. This scale consists of 43 items across 2 dimensions: perceived exercise benefits (29 terms) and perceived exercise barriers (14 terms). Each item is based on a 4-point scale ranging from 1 (strongly disagree) to 4 (strongly agree). Notably, the subscale measuring exercise barriers used a reverse scoring method. A higher score indicates greater perceived exercise benefits and lower perceived exercise barriers. The Cronbach’s alpha coefficient for EBBS in our sample was 0.925.

Physical activity

The Chinese version of the International Physical Activity Questionnaire - Long Form (IPAQ-L) demonstrated adequate reliability and showed sufficient evidence of validity, making it an appropriate tool for this study30. The IPAQ-L consists of 27 items that assess PA across four domains: leisure activities, household tasks, transportation, and work. Participants are asked to self-report the number of days and duration of their PA over the past 7 days, based on their subjective perception. This questionnaire uses standardized metabolic equivalent (MET) values for various activities. Each level of PA is assigned a specific MET value, and the weekly energy expenditure from PA (MET-minutes/week) is calculated using the following formula: Weekly energy expenditure (MET-min/week) = MET value × Activity duration (minutes/day) × Activity days (days/week).

Control variables

The control variables encompassed demographic and clinical characteristics. The demographic characteristics comprised age, gender, body mass index (BMI), marital status, employment state, education levels, and family income monthly per person RMB. The clinical variables included disease duration, complications, and the degree of dyspnea measured by the modified British Medical Research Council (m-MRC) scale. This scale is a self-rating tool to measure the degree of disability that breathlessness impacts day-to-day activities, ranging from 0 (no breathlessness except on strenuous exercise) to 4 (too breathless to leave the house or breathless when dressing or undressing)31.

Data collection

Participants were recruited continuously until the required sample size was achieved. Sampling criteria were evaluated, and informed consent was obtained from all participants. Trained research staff collected data on-site, either at the bedside or in the waiting room. Data collectors were specifically trained to ensure a high response rate, minimize systematic differences between respondents and non-respondents, and reduce the amount of missing data. Participants self-completed the questionnaires within 15–30 min, with data collectors present to provide assistance and address any questions.

Data analysis

Data analysis was conducted in SPSS 25.0. Descriptive statistics were employed to present the demographic and clinical characteristics of the participants, including scores (Mean ± Standard Deviation, Mean ± SD), medians (first quartile-third quartile, Q1-Q3), frequency (N), and percentage (%). Univariate analysis was performed to assess the effects of demographic and clinical variables on PA, using the Kruskal-Wallis and Mann-Whitney tests. Spearman’s correlation analysis was used to examine the relationships between dyspnea-related kinesiophobia, exercise perception, exercise self-regulatory efficacy, and PA. The mediation and moderation analyses were analyzed using the PROCESS 3.5 macro program compiled by Hayes, with control variables accounted for. A total of 5,000 bootstrapping samples were used to calculate the bias-corrected 95% confidence interval (CI). First, we tested the association between dyspnea-related kinesiophobia and PA by assessing exercise self-regulatory efficacy using Model 4 of the PROCESS. If the 95% CI of a*b did not contain 0, the indirect effect was considered significant. Model 59 of PROCESS was subsequently used to examine the moderated mediation effect of exercise perception, namely, whether exercise perception moderates the direct and indirect effects of dyspnea-related kinesiophobia and PA. A 95% CI of the interaction that did not contain 0 indicated that the moderated mediation effect was established32. Simple slope analysis was used to further analyze the moderating effect of exercise perception. The test level was α = 0.05.

Results

Characteristics of study participants

A total of 250 individuals with COPD were invited to participate in this study, of which 239 questionnaires were collected (Response rate: 95.6%). The demographic characteristics are presented in Table 1. The median age of participants was 72.00 (66.00, 77.00) years and the average BMI was 22.57 ± 3.73 kg/m2. The median disease duration was 6.00 (2.00, 11.00) years. The range of PA energy consumption was 0.00-494.20 MET-h/week, and the median was 64.50 (28.00, 126.55) MET-h/week. The median PA energy expenditure in different domains was as follows: work, 0.00 (0.00, 0.00) MET-h/week; transportation, 3.50 (0.00, 13.55) MET-h/week; household tasks, 21.00 (0.00, 45.00) MET-h/week; and leisure activities, 11.55 (0.00, 23.10) MET-h/week. Participants were categorized into three PA levels: low (n = 30, 12.6%), moderate (n = 73, 30.5%), and high (n = 136, 56.9%). The median dyspnea-related kinesiophobia score was 16.00 (12.00, 20.00), the median exercise self-regulatory efficacy score was 94.00 (79.00, 108.00), and the median exercise perception score was 101.00 (96.00, 110.00).

Table 1 Univariate analysis of difference of PA scores regarding demographic and clinical variables (N = 239).
Full size table

Correlations among study variables

As shown in Table 2, The dyspnea-related kinesiophobia was negatively correlated with exercise perception, exercise self-regulatory efficacy, and PA (r = -0.503, P < 0.01; r = -0.739, P < 0.01; r = -0.657, P < 0.01). The exercise perception was positively correlated with exercise self-regulatory efficacy scores and PA (r = 0.375, P < 0.01; r = 0.289, P < 0.01). There was also a positive correlation between exercise self-regulatory efficacy and PA (r = 0.650, P < 0.01).

Table 2 Correlations among dyspnea-related kinesiophobia, exercise perception, exercise self-regulatory efficacy and PA (N = 239).
Full size table

Testing for the mediating effect

This study standardized the data and applied Model 4 of the PROCESS macro in SPSS to identify the mediation effect. Figure 2 presents the results of the mediation model, along with the standardized path coefficients for each path. As shown in Table 3, the results indicated that dyspnea-related kinesiophobia had a negative direct effect on both exercise self-regulatory efficacy (β = -0.706, P < 0.001) and PA (β = -0.170, P = 0.033). Additionally, exercise self-regulatory efficacy had a positive direct effect on PA (β = 0.361, P < 0.001). The total effect of dyspnea-related kinesiophobia on PA was statistically significant (β = -0.425, P < 0.001, 95% CI = [-0.552, -0.299]), and the indirect effect of dyspnea-related kinesiophobia on PA through exercise self-regulatory efficacy was also significant (β = -0.255, 95% CI = [-0.353, -0.164]). Consequently, we could confirm that exercise self-regulatory efficacy played a negative partial mediating role in the relationship between dyspnea-related kinesiophobia and PA in people with COPD, but such an indirect effect was lower than the direct effect.

Table 3 Mediation of exercise perception between dyspnea-related kinesiophobia and PA (N = 239).
Full size table
Fig. 2
figure 2

The mediation model depicted the effects of dyspnea-related kinesiophobia on PA in people with COPD. * P < 0.050; ** P < 0.010.

Full size image

Testing for the moderated effect

Model 59 of the PROCESS macro was used to examine the moderated effect of exercise perception on each path of the mediation model. When exercise perception was used as a moderating variable, the model indicated that both dyspnea-related kinesiophobia scores and the interaction between dyspnea-related kinesiophobia and exercise perception scores were significantly related to exercise self-regulatory efficacy. This suggests that exercise perception moderates the relationship between dyspnea-related kinesiophobia and exercise self-regulatory efficacy (Table 4). In addition, both exercise self-regulatory efficacy scores and the interaction between exercise self-regulatory efficacy and exercise perception scores were related to PA, which meant that exercise perception moderated the relationship between exercise self-regulatory efficacy and PA.

Model 58 of the PROCESS macro was adopted to further examine the moderated mediation effect. The regression analysis showed that dyspnea-related kinesiophobia negatively was associated with exercise self-regulatory efficacy (β = -0.654, P < 0.001), and the interaction of dyspnea-related kinesiophobia*exercise perception was also significantly associated with exercise self-regulatory efficacy (β = 0.100, P = 0.030). The results indicated that exercise perception moderated the first half of the mediation model “dyspnea-related kinesiophobia → exercise self-regulatory efficacy → PA”. The regression analysis showed that exercise self-regulatory efficacy was negatively associated with PA (β = 0.412, P < 0.001), and the interaction of exercise self-regulatory efficacy*exercise perception was also significantly associated with PA (β = 0.110, P = 0.018). The results indicated that exercise perception moderated the second half of the mediation model “dyspnea-related kinesiophobia → exercise self-regulatory efficacy → PA”. Figure 3 presents the moderated mediation model, along with standardized coefficients for each path.

Table 4 Moderating effect of exercise perception in the model mediated by exercise self-regulatory efficacy (N = 239).
Full size table
Fig. 3
figure 3

The moderated mediation model depicted the effects of dyspnea-related kinesiophobia on PA in people with COPD. * P < 0.050; ** P < 0.010.

Full size image

Further simple slope analyses were conducted to illustrate the significant interaction at 1 standard deviation (SD) below the mean and 1 SD above the mean of exercise perception scoRes. As shown in Fig. 4, for individuals with weaker exercise perception, higher dyspnea-related kinesiophobia was associated with lower exercise self-regulatory efficacy (β simple = -0.754, P < 0.001). In contrast, for individuals with stronger exercise perception, the effect of dyspnea-related kinesiophobia on exercise self-regulatory efficacy was still significant but considerably weaker (β simple = -0.554, P < 0.001). This suggests that as exercise perception increases, the influence of dyspnea-related kinesiophobia on self-regulatory efficacy diminishes. In addition, for individuals with stronger exercise perception, exercise self-regulatory efficacy was significantly associated with PA (β simple = 0.522, P < 0.001). For individuals with weaker exercise perception, the effect of exercise self-regulatory efficacy on PA was still significant (β simple = 0.303, P = 0.001), but the regression coefficient decreased, indicating that as exercise perception increases, the influence of self-regulatory efficacy on PA becomes stronger.

Fig. 4
figure 4

(a) exercise perception as a moderator of the relationship between dyspnea-related kinesiophobia and exercise self-regulatory efficacy; (b) exercise perception as a moderator of the relationship between exercise self-regulatory efficacy and PA.

Full size image

Discussion

Essential finding

In this study, we investigated the impact of dyspnea-related kinesiophobia on PA. Our findings revealed that dyspnea-related kinesiophobia not only had a direct adverse effect on PA, but also indirectly influenced PA through the mediating effect of exercise self-regulatory efficacy. Moreover, increased exercise perception not only reduced the negative impact of dyspnea-related kinesiophobia on exercise self-regulatory efficacy, but also increased the positive influence of exercise self-regulatory efficacy on PA in people with COPD.

The direct effect of dyspnea-related kinesiophobia on PA

In our study, we observed that 237 (99.2%) COPD patients had dyspnea-related kinesiophobia, and this was associated with lower PA, consistent with a previous research15. The fear-avoidance model proposed by Leeuw W et al.33 served as a psychological framework used to explain the behaviors and psychological reactions of patients with chronic pain and other chronic health conditions. In patients with COPD, shortness of breath is an uncomfortable and potentially life-threatening sensation that easily triggers strong emotional reactions. This can lead to catastrophic thinking about PA, causing patients to exaggerate its severity and potential consequences. Tai Chi has been shown to enhance both physical and mental well-being by diminishing fear and embarrassment associated with breathlessness, potentially facilitating continued PA and improving the quality of life in individuals with COPD34. In addition, non-pharmacologic interventions, such as inspiratory muscle training and pulmonary rehabilitation demonstrated positive effects on the affective ___domain of dyspnea. Inspiratory muscle training significantly reduces dyspnea-related kinesiophobia in patients with COPD, leading to improved physical function and reduced levels of anxiety and depression11. Furthermore, pulmonary rehabilitation has been effective in reducing exercise-related dyspnea and fear, consequently enhancing their quality of life35. It is therefore essential for medical professionals to provide tailored recommendations based on the degree of dyspnea through the mMRC scale.

The mediating role of exercise self-regulatory efficacy

In addition to the direct impact of dyspnea-related kinesiophobia on PA, another indirect pathway included the mediation of exercise self-regulatory efficacy. In our study, the median exercise self-regulatory efficacy score of 94.00 suggested that most participants fall into the moderate-to-high self-efficacy category, which was related to a higher level of PA. Patients may face challenges in maintaining a regular exercise regime, so exercise programs can target self-regulatory barriers such as exacerbation, dyspnea, and hypoxia. Self-regulatory efficacy is a psychological resource that is necessary for the ongoing self-management of long-term PA. Targeting key self-regulatory mechanisms of behaviour change increases the potential for long-term adherence. Theories can guide which key psychological mechanisms of change are necessary to target and then behaviour change techniques (BCTs) can be implemented to change those targeted constructs36. The Small Steps for Big Changes intervention purposefully delivered BCTs (e.g.,1.2 problem solving, 1.1 goal setting [behaviour], 1.3 goal setting [outcome]) to specifically bolster participants’ self-regulatory skills and the self-efficacy to enact those skills37. Understanding the mechanisms of change through which BCTs influence a targeted behaviour is needed to advance behaviour change science and design more effective interventions.

The moderating role of exercise perception

Our results provided an interesting insight into the relationship between dyspnea-related kinesiophobia and PA, whereby exercise perception moderates the mediating process by which dyspnea-related kinesiophobia affects PA levels through exercise self efficacy in patients. The perception of behavioral benefits and obstacles was first proposed in the Health Belief Model (HBM), which holds that an individual’s cognitive attitude toward health behavior affects his or her behavioral intention, while an individual’s cognitive attitude toward exercise behavior is mainly embodied in his or her perception of the benefits and obstacles of exercise38. The perception of exercise benefits and obstacles is an important predictor of patients’ compliance with functional exercise. Perception of exercise benefits can motivate patients to complete rehabilitation exercise tasks more actively, while perception of many exercise obstacles may reduce their willingness to exercise. Encouraging patients to share their exercise experiences through communication platforms can improve their exercise perception. For example, health education using theoretically-based multimedia messaging service targeting exercise benefits and barriers has been effective and affordable in promoting and changing patients’ beliefs and PA behaviors39,40. In addition, The motivational communication method and the “5A” model were used to implement intervention management, which corrected the catastrophic understanding of dyspnea of patients, reduced the fear of PA, and strengthened the understanding of the importance of exercise41.

Strengths and limitations

The strength of this study lay in the thorough analysis of the potential mediator and moderator between dyspnea-related kinesiophobia and PA. However, this study inevitably had several limitations. First, the cross-sectional data could not prove the causal relationship between dyspnea-related kinesiophobia, exercise self-regulatory efficacy, and PA in people with COPD. Therefore, further investigation in longitudinal studies is needed to ascertain the relationship and strength of these variables. Second, despite accounting for some potential confounding variables, the association between dyspnea-related kinesiophobia and PA may be influenced by other factors, such as social support and environmental exposure. Comprehensive studies should explore the impact of these potential elements on this relationship. Third, the small sample of hospitalized patients may not adequately represent the broader population of individuals with COPD. Consequently, the findings of this study need to be considered with caution when generalizing to all individuals with COPD.

Implications

This study carries significant implications. First, this study reaffirms the impact of dyspnea-related kinesiophobia on patients’ physical activities, and managers need to consider strategies to assess the degree of dyspnea and dyspnea-related kinesiophobia when developing intervention programs to improve physical activities. Second, the mediating role of patients’ exercise self-regulatory efficacy provides a new perspective on the negative effects of dyspnea-related kinesiophobia on PA. Future studies could examine mediating effects of an intervention on longer-term behaviour. Third, the moderating results of exercise perception provide an idea for the design of interventions, such as providing communication platforms for patients. These platforms could facilitate the sharing of exercise experiences and benefits, thus potentially improving patients’ perceptions of exercise and promoting adherence to PA regimens.

Conclusion

Almost all individuals with COPD in this study experienced dyspnea-related kinesiophobia, which directly and adversely affected PA. Furthermore, the moderated mediation model indicated that exercise self-regulatory efficacy could mediate the association between dyspnea-related kinesiophobia and PA, and exercise perception played a moderating role in this relationship. Consequently, future interventions seeking to improve the level of PA in patients with COPD should focus on improving exercise self-regulation efficacy and increasing exercise perception.