Sense of Coherence as a Mediator Between Functional Status and Health-Related Quality of Life in Patients With Heart Failure

ABSTRACT Background Poor functional status relating to heart failure (HF) negatively affects health-related quality of life (HRQOL). Patients with HF, especially those with New York Heart Association (NYHA) Class III or IV HF, often exhibit poor HRQOL because of physical limitations and HF-related symptoms. Although sense of coherence (SOC) has been reported to be a determinant of HRQOL, its role as a mediator between functional status and HRQOL remains unclear, and few studies have explored the prevalence of HF in patients in NYHA Classes I and II. Purpose This study was designed to investigate SOC as a mediator between different functional status classes and HRQOL in patients with HF. Methods A cross-sectional study was conducted on patients with HF recruited from a hospital in northern Taiwan from April 2020 to September 2020. The Minnesota Living with Heart Failure Questionnaire and a questionnaire on sociodemographic characteristics; functional classification in terms of NYHA Classes I, II, and III; and SOC were administered. The PROCESS v3.5 (by Andrew F. Hayes) macro was applied to analyze the effects, and Model 4 was used to examine the mediating role of SOC on the relationship between NYHA functional class and HRQOL. Results Of the 295 participants, SOC was found to mediate the effects of functional status on HRQOL more significantly in patients in Class II than those in Class III but not more significantly in patients in Class I than those in Class III. A weaker mediating effect of SOC was noted on the relationship between functional status and HRQOL in patients with HF in NYHA Class II than those in Class III. Conclusions In patients with HF, poor functional status often reduces HRQOL significantly. SOC mediates the relationship between functional status and HRQOL more significantly in those in NYHA Class II than those in Class III. Nursing staff should work to increase patients' SOC by strengthening their coping capacity and improving their functional status to improve their HRQOL.


Introduction
Heart failure (HF), a severe health condition with high prevalence, leads to substantial morbidity and mortality as well as poor outcomes (Murphy et al., 2020).The prevalence of HF is increasing not only because of population aging but also because of improved cardiovascular disease treatment (Dunlay et al., 2021).HF is a complex clinical condition with various signs and symptoms, including poor cardiac function and physical capacity, that results in poor health-related quality of life (HRQOL; Costa et al., 2020).One study conducted across 15 countries reported patients with HF to be older and to have generally poorer physical and mental functioning, sense of coherence (SOC), and quality of life (Lu et al., 2022).A study reported that patients with HF and classified as New York Heart Association (NYHA) Functional Class III or IV, patients with depression, and older patients faced higher risks of poor HRQOL (Gott et al., 2006).Patients with HF and classified as NYHA Classes III and IV have poor functional status and are thus particularly vulnerable to poor HRQOL (Mehrotra et al., 2019).
Functional status, which is commonly determined using the NYHA functional classification, is a predictor of hospitalization, HRQOL, and mortality (Bredy et al., 2018;Yeh & Shao, 2021).NYHA functional status is crucial in chronic disease management for determining patient prognoses and hospital readmission need.NYHA functional status is a predictor of hospital readmission (Albuquerque, 2018).An international, multicenter, randomized trial including 613 patients with HF examined the prognostic utility of NYHA functional status, particularly in terms of HRQOL improvement and treatment outcomes (Giustino et al., 2020).A large international study on 4,028 patients with congenital heart disease from 15 countries revealed associations between functional status and mental health outcomes and quality of life (Moons et al., 2021).Although functional status determined using the NYHA functional classification is a significant predictor of HRQOL in patients with HF (Molla et al., 2021), more psychosocial strategies are required.
SOC is an adaptive, dispositional orientation that refers to the ability of an individual to manage negative events and identify internal and external resources to cope with life stress.According to Antonovsky (1996), SOC can serve as a stress-resisting resource in salutogenic approaches in health improvement.Systematic reviews have reported SOC, as determined using the scale developed by Antonovsky, to be strongly related to perceived health status and quality of life and to be a significant mediator in the relationship between health status (Eriksson & Lindström, 2007) and quality of life (Tang & Dixon, 2002).SOC has been shown to be closely related to HRQOL in patients with chronic diseases (Eriksson & Lindström, 2007;Qiu et al., 2020).In addition, a study reported higher SOC to be a predictor of better quality of life in patients with early breast cancer (Vähäaho et al., 2021).Furthermore, the results of another study indicate low SOC may partially explain the effect of hostility on mental HRQOL in patients with heart disease at 12-28 months of follow-up, although neither hostility nor SOC was found to be associated with physical HRQOL (Silarova et al., 2016).Thus, interventions aimed at strengthening SOC can improve HRQOL.
SOC is related to the psychosocial effects of health status and may mediate the relationship between disability and functional status (Schnyder et al., 1999).Strengthening SOC through activities related to empowerment and reflective learning can enable people to perceive meaningfulness, manageability, and comprehensibility (Super et al., 2016).Several studies have shown that SOC reduces the burden posed by functional limitations (Ekman et al., 2002;Potier et al., 2018).One study reported that, in the general Swedish population, weaker SOC was associated with poorer functional status (Langius & Björvell, 1993), indicating SOC may be a protective factor against poor functional status.In addition, SOC is a marker of HRQOL and may mediate the relationship between physical health status and HRQOL (Chen et al., 2022).However, the effect of SOC on the HRQOL of patients with heart disease remains unclear, causing difficulties in developing interventions for improving the HRQOL in these patients (Galletta et al., 2019).
Poor functional status adversely affects the HRQOL of patients with HF and in NYHA Classes III and IV (Mehrotra et al., 2019).Most studies have focused on the treatment of patients with HF and poor functional status, with few exploring the high prevalence of HF in patients in NYHA Classes I and II.Thus, whether SOC mediates the relationship between NYHA functional class and HRQOL in patients with mild HF remains unclear.Therefore, a theoretical framework was developed in this study to examine the mediating role of SOC on the relationship between functional status and HRQOL in patients with mild HF.In this study, functional status was hypothesized to be associated with poor HRQOL in patients with mild HF, and SOC was hypothesized to mediate the relationship between their functional status and HRQOL.

Study Design
This cross-sectional study examined the association among functional status, SOC, and HRQOL in patients with HF.Convenience sampling was used to recruit patients with HF as participants from April 2020 to September 2020.

Participants
Patients aged older than 20 years (the legal threshold of adulthood in Taiwan) who had documented HF and were able to provide informed consent were recruited from a cardiovascular outpatient clinic in a large medical center in northern Taipei, Taiwan.Patients diagnosed by a clinical physician as being in NYHA Class I, II, or III were considered to have HF.Patients were excluded if they were not able to answer the study questionnaire, were in NYHA Class IV, or could not communicate in Mandarin Chinese or Taiwanese.Critically ill patients in NYHA Functional Class IV were excluded because of the low number of patients in this category in the outpatient clinic (Greene et al., 2021), their higher risk of mortality and hospitalization (Ahmed et al., 2006), and their higher likelihood of being unable to complete the study questionnaire.Potential participants were initially identified from daily cardiology outpatient clinic lists at the target hospital.Written informed consent was obtained from patients before participation.
Using a Monte Carlo power analysis, we determined that 269 participants would be required to detect a clinically meaningful mediation effect, obtain a medium effect size, and determine the relationship between quality of life and self-rated health (Patrick et al., 2000), with an intraclass correlation coefficient of .4 and a power of 0.8 (Pan et al., 2018).Considering a dropout rate of 10%, we recruited 300 participants.After excluding five participants because of incomplete questionnaire responses, data from 295 participants were available for analysis.This sample size satisfied the criteria for drawing clinically meaningful conclusions.

Instruments
The questionnaire included a demographic datasheet, the SOC scale, and the Minnesota Living with Heart Failure Questionnaire (MLHFQ).The demographic datasheet gathered information on gender, age, marital status, educational level, occupational status, and socioeconomic status.Details on left ventricular ejection fraction (LVEF) and NYHA functional classes were obtained from medical records.Functional status was determined using NYHA class.

The Journal of Nursing Research
Hsiao-Ping LEE et al.

Functional status
Functional status was estimated in accordance with the NYHA functional classification.The NYHA classification includes four classes (I-IV) and is used to evaluate HF symptoms and limitations in physical activity.A higher class indicates more severe symptoms and limitations on physical activity (Greene et al., 2021).The NYHA class of each participant was determined through direct observation during cardiologist clinic visits, indirect interpretation of their symptoms, the results of clinical tests for cardiac function, and their medical history.

Minnesota living with heart failure questionnaire
The Chinese version of the MLHFQ was used to measure HRQOL in this study.The MLHFQ consists of three subscales: physical, mental, and social.The Chinese version of the MLHFQ has been shown to be a valid and reliable instrument for measuring HRQOL in patients with NYHA Class II or III HF (Ho et al., 2007).The MLHFQ consists of 20 items that are scored on a 6-point Likert scale ranging from 0 (no effect on HRQOL) to 5 (strongest effect on HRQOL), with lower scores indicating better HRQOL.The Cronbach's α values and correlation coefficients of the subscales ranged from .93 to .95 and from .55 to .80,respectively (Ho et al., 2007).The content validity index was .98,and on the basis of the construct validity results, three factors had a 20-item loading above .50,accounting for 71% of the variance (Ho et al., 2007).The Cronbach's alpha was .83 for the entire scale and .71,.57,and .70 for the physical, emotional, and social subscales, respectively.We obtained permission to use the MLHFQ from the University of Minnesota as well as permission from Dr. Ho, who was responsible for the translation and validation of the Chinese version.We naturally log-transformed the MLHFQ scores for more favorable comparison.

Sense of coherence scale
A short version of the Mandarin SOC scale was employed in this study to measure SOC.According to Tang and Dixon (2002), the 13-item short version of the Mandarin SOC scale, which measures comprehensibility, manageability, and meaningfulness, is valid and reliable (α = .89).This scale, translated from the original English, is a general conceptual equivalent to the English SOC-13 scale and was determined by experts to be an accurate translation and to be valid and reliable (Tang & Dixon, 2002).Participants were asked to respond to each item on a 7-point scale ranging from 1 (very seldom) to 7 (very often), with higher scores indicating better coping capability.In this study, Cronbach's alpha was .88 for the entire scale.To ensure a more favorable comparison among the domains, we standardized the SOC scale scores to obtain a t score with a mean of 50 and a standard deviation of 10 (Lin et al., 2013).

Ethical Considerations
This study was approved by the hospital's ethics committee (Approval No. A202005040).All of the participants provided written informed consent.Permission was obtained to use the questionnaires from their respective authors.Patients were assured of the confidentiality of their data and their right to withdraw from the study at any time for any reason.All participants voluntarily participated in this study.

Statistical Analysis
Descriptive statistics were calculated for the characteristics of the study sample, and bivariate correlations were used to examine the association among functional status, SOC, and HRQOL.Raw data were evaluated for normality and outliers before the analysis.One-way analysis of variance and the chi-square (or Fisher's exact) test were performed to examine the homogeneity in terms of characteristics among the three functional status groups.The PROCESS plug-in computational macro v3.5 was used for mediation and moderation analyses.Model 4 (PROCESS macro) was used to examine the mediating role of SOC between NYHA functional class and HRQOL in the population with HF.The alpha criterion for significance was set at p < .05(two tailed), and 95% confidence interval (CI) was calculated using the bootstrapping method.

Participant Characteristics
The characteristics and descriptive statistics of the participants are shown in Table 1.The mean age was 70.12 (SD = 11.80)years, approximately half were men (n = 148, 50.2%), and most were married (n = 225, 76.3%).Approximately half of the participants (n = 134, 45.4%) had a junior high school degree or lower, and a relatively small percentage (n = 23, 7.8%) were of a low socioeconomic status.In terms of employment status, approximately two of three participants (n = 200, 67.8%) were retired, and most could care for themselves independently (n = 209, 70.8%), perceived their health as average (n = 170, 57.6%), and had good family relations (n = 252, 85.4%).Most had NYHA Functional Class II (n = 213, 72.2%), and their mean LVEF was 52.66 (SD = 14.42).In terms of demographic and medical variables, only age, marital status, retirement status, socioeconomic status, primary caregiver status, and LVEF differed significantly among the three groups ( p < .05;Table 1).

Bivariate Correlations Among Outcome Variables
In this section, we refer to variables using the names of measured constructs rather than the names of the scales.In terms of NYHA functional class, Class III indicates severe functional status problems, whereas Class I indicates no or insignificant functional status problems.In terms of SOC, the highest scores indicate excellent SOC, whereas the lowest scores indicate poor SOC.In terms of HRQOL, the highest scores indicate severe HRQOL problems, whereas the lowest scores indicate no HRQOL problems.
Bivariate correlations are shown in Table 2.However, for the NYHA functional classes (ordinal data), Spearman correlation was used rather than Pearson correlation.NYHA functional class was found to be negatively correlated with SOC (r = −.12,p = .036)and positively correlated with HRQOL scores (r = .36,p < .001) as well as its physical (r = .28,p < .001),emotional (r = .29,p < .001),and social (r = .35,p < .001)dimensions.Furthermore, SOC was found to be negatively correlated with HRQOL scores (r = −.36,p < .001) as well as its physical (r = −.31,p < .001),emotional (r = −.36,p < .001),and social (r = −.21,p < .001)dimensions.HRQOL was found to relate positively to its physical (r = .94,p < .001),emotional (r = .75,p < .001),and social (r = .73,p < .001)domains.The participants in NYHA Class II had significantly higher SOC scores (B path a 1 = 0.34, p = .03)than those in NYHA Class III, and the participants in NYHA Class I had significantly higher SOC scores (B path a 2 = 0.54, p = .03)than those in NYHA Class III (Figure 1).The participants with higher SOC scores had lower HRQOL scores when the holding condition was constant (B path b = −0.32,p < .001).After adjusting for group differences in SOC, those in NYHA Class II had lower HRQOL scores (B path c' 1 = −0.68,p < .001)than those in NYHA Class III, and those in NYHA Class I had lower HRQOL scores (B path c' 2 = −1.01,p < .001)than those in NYHA Class III.

Discussion
The mediating effect of SOC on the relationship between functional status and HRQOL was investigated in this study.
Elucidating the mediating effect of variables can help provide better explanations of outcomes.To the best of the authors' knowledge, no previous study has examined the role of SOC as a mediator in the relationship between NYHA functional class and HRQOL in patients with mild HF.A small mediating effect of SOC on the relationship between functional status and HRQOL was observed in the participants in NYHA Class II versus Class III but not in those in NYHA Class I versus Class III.Because of the low number of patients with HF in NYHA Class I in the sample, we found a nonsignificant mediating role of SOC on the relationship between functional status and HRQOL in the patients in NYHA Class I versus Class III.We observed a mediating effect of SOC on HRQOL in those in NYHA Class II versus Class III, which is consistent with prior studies on patients with chronic diseases (Galletta et al., 2019;Qiu et al., 2020).Although studies have suggested a positive effect of SOC on HRQOL, few studies have examined whether SOC mediates the relationship between NYHA functional status and HRQOL in patients with HF.Most prior studies have investigated the mediation effect of SOC on the relationship between psychosocial variables and HRQOL and observed a significant mediation effect of SOC on the mental component of HRQOL and a considerably weaker effect on the physical component of HRQOL (Hori et al., 2022).The results of our mediation model indicate a similar mediation effect of SOC on the relationship between physical functional status and HRQOL in the participants with NYHA Class II HF versus Class III HF.
The results of this study indicate a weak mediation effect of SOC on the relationship between functional status and HRQOL.A cross-lagged prospective study on 382 adolescents with congenital heart disease reported the relationship between SOC and health status to be reciprocal.SOC is a crucial determinant of general and disease-specific perceived health, and different domains of perceived health are less consistent predictors of SOC over time (Løvlien et al., 2017).NYHA functional class represents the physical health status of patients and is a less reliable predictor of the relationship between SOC and functional status.A cross-sectional study of 227 patients with HF from Hong Kong identified other variables significantly associated with HRQOL, namely, psychological distress, patients' self-perceived health status, NYHA grading, and educational level (Rohani et al., 2015).Future studies should consider other variables related to HRQOL in patients with HF.The finding in this study explains the weak relationship between functional status and SOC as well as the weak mediation effect of SOC.
Another reason for the weak mediation effect of SOC on the relationship between functional status and HRQOL was the functional status class composition of the participants.High disease severity predicts a significant association between SOC and HRQOL (Vähäaho et al., 2021), whereas low disease severity predicts a nonsignificant correlation between SOC and HRQOL (Anguas-Gracia et al., 2021).Because most of the participants in this study were in NYHA Class II, the effect of SOC on the relationship between functional status and HRQOL was weaker than in samples with participants in higher disease severity classes (i.e., NYHA Classes III and IV).Moreover, the mediating pathway of SOC in the context of HRQOL appears to be an inner resource for the psychological adaptation process (Hori et al., 2022).Patients with lower disease severity require less psychological adaptation.This explains the weak mediation effect of SOC on the relationship between functional status and HRQOL found in the participants with low disease severity.In addition, among the participants in NYHA Class I and with no symptoms or limitations on daily physical activities, those with lower disease severity had higher HRQOL and required less psychological adaptation.This may explain the nonsignificant mediating effect found for the SOC on the relationship between functional status and HRQOL in the participants with HF in NYHA Class I versus those in NYHA Class III.
In this study, the participants with better functional status and SOC were shown to be more likely to have better HRQOL.Similar to another study, this study found poor SOC leads to poor HRQOL (Yu et al., 2008).SOC can serve as a stress-resisting resource because it can improve energy and attenuate depression, thus improving both physical and psychological health (Eriksson & Lindström, 2007).Strengthening SOC can improve HRQOL by helping individuals understand that addressing feelings of loss and negative mood is meaningful and feasible (Apers et al., 2013).In addition, nurse-patient interaction is crucial to patients' SOC and its subdimensions of comprehensibility and manageability (Drageset et al., 2021).Empowering patients with HF to cope with symptoms of distress, feelings of powerlessness and hopelessness, and social role dysfunction can help them adjust to their illness (Nordfonn et al., 2019).Nurses play a vital role in strengthening SOC and improving HRQOL in patients with HF using positive nurse-patient interaction and in strengthening the coping capacity of patients by empowering health-related activities.

Limitations
This study is affected by several limitations.This study was designed as a cross-sectional trial and recruited patients by convenience sampling.Thus, the causal effect of variables could not be assessed, and findings should be generalized with appropriate caution to patients with different functional statuses and in different contexts.In addition, several sociodemographic and psychological variables such as psychological distress, family-related factors, and socioenvironmental factors were not measured and thus not controlled in this study.The proportion of patients with NYHA Class I HF was low in this study, indicating that the mediating effect was not significant.More patients with NYHA Class I HF should be included in future studies.Finally, the recruitment of participants from only one medical center reduces the generalizability of the findings to other populations and settings.

Conclusions
Increasing attention has been focused on HRQOL in healthcare.Functional status is known to play an essential role in improving HRQOL in patients with HF.The findings of this study indicate SOC is a significant mediator that protects the HRQOL of those with NYHA Class II versus Class III HF.The findings contribute to the literature by expanding scholarly knowledge on the direct effects of functional status on HRQOL in patients with mild HF.Understanding how functional status affects HRQOL in patients with mild HF can assist nurses to make more informed decisions and tailor treatments to individual patients.Nursing staff should help patients with HF improve their functional status through health promotion activities and improve their SOC through positive interaction and coping-ability strengthening through empowerment to prevent emotional distress and physical, psychological, and social burdens and thus improve HRQOL.

Table 3
Effects of Functional Status on HRQOL Outcomes Mediated by Sense of Coherence AfterControlling for Demographic Characteristics (Model 4) Note.Reference group: NYHA Class III."Class" refers to New York Heart Association Functional Classes I-III of heart failure.Mediation model of HRQOL: R 2 = 27.25%, F = 13.39,p < .001.SOC = sense of coherence; HRQOL = health-related quality of life; B = coefficient; SE = standard error.

Table 2
Bivariate Correlations Among the Main Study Variables (N = 295) Note.SOC was measured using the Sense of Coherence Scale; HRQOL was measured using the Minnesota Living with Heart Failure Questionnaire (MLHFQ).X = independent variable; M = mediator; Y = outcome variable; SOC = sense of coherence; HRQOL = health-related quality of life; HRQOL-physical = domain of physical health in the MLHFQ scale; HRQOL-emotional = domain of emotional health in the MLHFQ scale; HRQOL-social = domain of the social scale in the MLHFQ scale. a

Table 4
Relative Total, Direct, and Indirect Effects of Functional Status on Health-Related Quality of Life Mediated by Sense of Coherence Note.BSE, BLCI, and BUCI refer to the bootstrap standard error, bootstrap lower limit, and bootstrap upper limit of 95% confidence interval of each effect estimated by deviation-corrected percentile bootstrap method, respectively.Reference group: NYHA Class III.