Relationship Between Symptoms and Stage of Change in Both Adopting a Healthy Life Style and Quality of Life in Patients with Liver Cirrhosis: A Cross-Sectional Study


 Background: Previous studies of patients with liver cirrhosis have not considered the broad range of symptoms or the association between healthy behavior and quality of life. The purposes of this study were to examine the association between symptoms and adopting exercise and consuming fruits and vegetables and to identify factors associated with quality of life in patients with liver cirrhosis.Methods: This cross-sectional study enrolled 91 consecutive patients with liver cirrhosis in one tertiary general hospital in South Korea between February 2016 and January 2017. Each study participant completed a self-administered questionnaire that measured symptom, stage of change in engaging in exercise and consumption of fruits and vegetables, and the Korean version of the 36-item Short-Form Health Survey. Multivariate ordinal logistic regression analysis and multiple regression models was used, respectively, to examine the association between each symptom with stage of change in engaging in exercise and consumption of fruits and vegetables and to evaluate factors affecting quality of life.Results: Experiencing nausea was associated with a higher stage of change in engaging in exercise, but experiencing shortness of breath was associated with a lower stage of change in engaging in exercise. Experiencing right upper quadrant pain was associated with a higher stage of change in engaging in consumption of fruits and vegetables. Muscle cramps, anorexia, right upper quadrant pain and body pain, itching, ascites or edema, bruising, and change in appearance negatively affected quality of life.Conclusions: The results suggest that the types of symptoms experienced by a patient with liver cirrhosis hinder or promote the patient’s adoption of exercise and dietary behavior. Experiencing symptoms may negatively affect quality of life. Caregivers should provide supportive care to patients with liver cirrhosis, which includes assessing and managing symptoms to improve quality of life.

fatigue can alter their aerobic capacity. Although we were unable to nd recent prior studies on symptoms and adherence to diet in patients with liver disease, we can predict that as compared with healthy individuals, nutritional de ciencies may occur in LC patients because of decreased dietary quantity, decreased digestive and absorption functions, changes in nutrient metabolism, increased nutritional requirements, ascites, or swelling. Thus, it is necessary to manage these patients' diets. The diverse symptoms of patients with LC can signi cantly impair QOL [24]. To date, few studies on liver disease patients have dealt with symptoms, exercise and dietary behavior, and QOL together.
Healthy behaviors, such as exercise and eating F&V, are tied to the diverse systemic symptoms experienced by LC patients, which conversely hinder the adoption of these behaviors. By determining that certain symptoms experienced by LC patients are correlated with adopting healthy behaviors, we can work to improve their QOL by controlling patients' subjective symptoms or encouraging the adoption of allied behaviors. The aims of the present study were to examine the association between symptoms and SOC of engaging in exercise and consuming F&V and to examine the factors in uencing QOL in a population of patients with LC.

Methods
This study used a cross-sectional design. The participants were recruited sequentially from one tertiary general hospital in South Korea between February 2016 and January 2017. Enrolled patients were older than 19 years, had a diagnosis of LC within the previous 2 years, and were undergoing regular treatment or follow-up. Patients were excluded if they were not physically or mentally healthy enough to complete the questionnaires or unable to understand the purpose of the study or provide informed consent.
We reviewed the hospital registries and initial assessments of patients returning to the hospital for follow-up or treatment. From this review, we identi ed 98 patients as eligible, and 91 of them (95%) accepted the informed consent document and completed the questionnaire. For the 7 individuals who refused, the most common reason for rejecting participation was "inconvenience." G*Power 3.1.9.7 calculated a sample size of 84, which was based on indices such as statistical analysis multiple regression, effect size f 2 0.21 [25], type 1 error 5%, power 80%, and nine predictors. Thus, the nal sample size of 91 was acceptable.
Research staff contacted eligible patients at the ambulatory clinic and inpatient ward to explain the details of the study personally. After providing informed consent, each participant was given a packet of questionnaires in a quiet and independent meeting room, and they lled out the self-reported questionnaire in the presence of the research staff. The questionnaires collected information on sociodemographic characteristics, symptom, SOC for adopting healthy behaviors based on the TTM [14], and QOL. The study protocol conformed with the ethical guidelines of the 1975 Declaration of Helsinki as re ected in a priori approval by the a liated Institutional Review Boards of the hospital (2015-09-020-003) and the university (2015-87).

Symptoms
Multidimensional symptom was measured using a self-reported scale developed in a previous study [25]. This scale assesses 18 symptoms: fatigue, muscle cramps, decrease in memory, itching, dyspepsia, anorexia, dark urine, drowsiness, dry mouth, bruising, right upper quadrant (RUQ) pain, shortness of breath (SOB) and/or dyspnea, nausea and/or vomiting, body pain, urinary di culty, tarry stools, ascites/edema, and change in appearance. The patient rated each item in frequency, intensity, and distress on a four-point Likert-type scale (0 = never experienced to 3 = extremely experienced in frequency, intensity, and distress). The symptom score was calculated by summing the scores of frequency, intensity, and distress for each symptom. Thus, the score ranged from 0 to 162, with a higher score indicating having experienced more severe symptoms. The Cronbach's alpha for the symptom scale in this study was 0.96.

Adopting Healthy Behaviors
Adopting healthy behaviors was evaluated from the SOC of TTM. The SOC assessments for engaging in exercise and consumption of F&V were adapted from a previous study [26], in which the measure was developed from the TTM of behavior change [14]. These measures characterized differences in the SOC of goal behaviors related to exercise and diet, from preadoption to adoption. The ve items of each measure characterized participants into the following ve stages: precontemplation, contemplation, preparation, action, and maintenance. Precontemplation describes an individual who is not engaged in a new behavior and has no intention of adopting the behavior in the near future. Contemplation indicates an individual who is not engaged in the behavior but is considering adopting the behavior in the next 6 months. Preparation describes an individual who has initiated some behavior changes, but the changes are not regular. Action describes an individual who is regularly engaged in the behavior but only initiated this new behavior in the past 6 months. Maintenance indicates an individual who has been regularly engaging in the behavior for more than 6 months.
Before questioning, exercise was de ned as "brisk walking, cycling, swimming, mountain climbing, or another form of exercise that makes your heart pound, or makes you break out in a sweat, and is not part of your normal job activity"; F&V servings were de ned as "one serving of vegetables is equal to 1/2 cup of cooked (or parboiled) vegetable, 1 cup raw vegetable or vegetable juice, and 2 cups leafy salad greens" and "one serving of fruits is 1 cup fruit or 1/2 cup of fruit juice (orange juice, etc.) or 1/3 cup of a fruit juice blend." Participants were then asked, "On average, do you exercise regularly at least 30 minutes per day, 5 days a week and do you regularly eat at least ve servings of F&V per day?" [27]. If the participant responded "yes," the next questions were posed: "How long have you been exercising regularly at least 30 minutes per day, 5 days a week and regularly eating at least ve servings of F&V per day?" The responses were "less than 6 months" (action stage) and "6 months or longer" (maintenance stage). If a patient replied "no" or "I am not sure" to the initial question, they were asked the following question: "Have you initiated exercising at least 30 minutes per day, 5 days a week and eating at least ve servings of F&V per day?" The responses were "yes" (preparation stage) and "no." If the response was "no," the patient was asked the following question: "Are you earnestly considering exercising at least 30 minutes per day, 5 days a week and eating at least ve servings of F&V per day within the next 6 months?" The possible responses were "yes" (contemplation stage), "no," and "I am not sure" (precontemplation stage). Precontemplation, contemplation, and preparation were classi ed into "not engaging in exercise or consumption of F&V," and action and maintenance were classi ed into "engaging in exercise or consumption of F&V" ( Figure   1). QOL QOL was assessed using the validated Korean version of the Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) [28]. This questionnaire consists of 36 items among eight subscales: physical functioning, role limitations due to physical health problems, role limitations due to emotional problems, social functioning, body pain, general health, vitality, and mental health. Among the subscales of the SF-36, we considered only the functioning scales in this study; thus, subscales measuring general health, mental health, and vitality were excluded. In addition, because symptoms were considered independent variables, the bodily pain subscale was excluded from the dependent variable because of the problem of multicollinearity. The present study used 19 items from the following four subscales: physical functioning, role limitations due to physical health problems, role limitations due to emotional problems, and social functioning. The score on each subscale ranged from 0 to 100, with a higher score indicating better functioning; however, a higher score on role limitation indicated greater limitation. The internal consistency for the SF-36 was good (Cronbach's alpha = 0.85) [28]. Use of the questionnaire was approved by Optuminsight Life Sciences, Inc (Lincoln, RI, USA).

Statistical Analysis
Sociodemographic characteristics and degree of dimensional symptoms were presented using descriptive statistics. First, in the univariate analyses, an independent samples t-test was used to examine the associations between sociodemographic parameters and total score of symptoms, and Fisher's exact test was used to examine the associations between sociodemographic parameters and engaging in exercise and consumption of F&V. For the Fisher's exact test, we categorized the stages of precontemplation, contemplation, and preparation into "not engaging in exercise or consuming F&V" and the stages of action and maintenance into "engaging in exercise and/or consuming F&V." Then, because the ve categories of SOCs constituted a dependent variable, we used multivariate ordinal logistic regression analysis with stepwise selection to examine the independent effect of each symptom on engaging in exercise or consumption of F&V, controlling for age, sex, marital status, having a religion, job status, monthly income, residence area, and number of family members living together. The results were presented as adjusted odds ratio (aOR) with 95% con dence interval (CI). Descriptive statistics were shown for the symptoms demonstrating signi cant changes according to the ve SOCs in adopting exercise and consumption of F&V, and we analyzed the correlations between the SOC and symptoms using Pearson correlation coe cient. Multiple regression analysis with stepwise selection, in which the main independent variables were engaging in exercise and consumption of F&V, sociodemographic characteristics, and all symptom subscales, was used to identify the independent effect on QOL. These results are presented as beta coe cients and standard errors. For the multivariate ordinal logistic regression and multiple regression analyses, we selected independent variables based on a previous study on factors associated with poor QOL in patients with LC [17]. All statistical analyses were conducted using SAS 9.4 package (SAS Institute, Cary, NC, USA).

Participant Characteristics
Of the 91 patients, 56% were male; the mean patient age was 59 years, and 70% of patients were younger than 65 years (Table 1). Most participants were married and covered by the national health insurance system (71.4%). More than half of the patients had no job at the time of the questionnaire (59.3%), two-thirds (65.9%) lived in a metropolitan area, and about 60% were living with two or fewer family members. The mean of the total symptom score of the participants was 33.6 (standard deviation [SD], 24.76) ( Table 1) Table 2). An analysis of symptom experiences signi cantly associated with QOL (please refer to Table 5 indicated that anorexia was the 6th-most common symptom (M = 2.1, SD = 2.5), RUQ pain was 8th (M = 1.9, SD = 2.1), bruising was 11th (M = 1.6, SD = 2.1), body pain was 14th (M = 1.3, SD = 2.0), ascites/edema was 17th (M = 1.0, SD = 1.9), and change of appearance was the least experienced symptom (M = 0.9, SD = 1.9) ( Table 2).

Correlations of Sociodemographic Characteristics with Symptom and Engaging in Healthy Behaviors
Unemployment (p = 0.001), low monthly household income (p = 0.014), and joining Medical Aid, the national insurance for low-income individuals in South Korea (p = 0.037), were correlated with more symptoms. Sociodemographic characteristics were not signi cantly associated with engaging in exercise and dietary behavior (Table 1).

Independent Association of Symptoms with SOC in Engaging in Exercise and Consuming F&V
The results of the multivariable analysis indicated that experiencing nausea and/or vomiting was associated with higher SOC in engaging in exercise (aOR = 2.25, 95% CI = 1.70-2.98), whereas experiencing SOB was associated with lower SOC in engaging in exercise (aOR = 0.61, 95% CI = 0.48-0.78). In addition, experiencing RUQ pain was associated with lower SOC in engaging in consumption of F&V (aOR = 3.16, 95% CI = 2.23-4.49) ( Table 3).
Correlation Between Symptoms and SOC in Both Engaging in Exercise and Consuming F&V A higher SOC in engaging in exercise was correlated with lower SOB (r = −0.31, p = 0.003) and more severe nausea and vomiting (r = 0.54, p < 0.0001). A higher SOC in consuming F&V was correlated with more severe RUQ pain (r = 0.78, p < 0.0001) ( Table 4).
Role limitations due to emotional problems was positively associated with experiencing muscle cramps (β = 7.18, partial R 2 = 26%, p < 0.0001) and change in appearance (β = 5.25, partial R 2 = 26%, p = 0.005). The linear model with these independent variables explained 52% of total variance in role limitations resulting from emotional problems (Table 5).

Discussion
The main ndings of this study indicated that experiencing SOB was associated with a reduction in exercise, experiencing nausea and vomiting was associated with adopting exercise, and experiencing RUQ pain was associated with consuming F&V. In addition, the QOL of LC patients was negatively affected by experiencing muscle cramps, anorexia, RUQ pain and body pain, itching, ascites or edema, bruising, and change in appearance. Our results suggest that the types of symptoms experienced by LC patients affect their adoption of exercise, dietary behavior, and QOL.
The current study indicated that experiencing SOB was associated with lower SOC in engaging in exercise. In patients with cirrhosis, SOB can be complicated by ascites or pleural effusion. Dyspnea is well known to induce exercise intolerance, and individuals with daily-life dyspnea have more limited exercise capacity [29]. SOB may have inhibited exercise in the cirrhosis patient population because this symptom becomes worse during exercise. Patients who perceive a temporary worsening of symptoms during physical activity may be less physically active [30]. Previous studies of the relationship between dyspnea and exercise intolerance were conducted in patients with pulmonary disease and multiple sclerosis. Therefore, further study is required to determine the impact of SOB on exercise levels in LC patients.
On the other hand, we found that experiencing nausea correlated with engaging in exercise, and experiencing RUQ pain correlated with greater F&V consumption. According to the Health Belief Model, patients will perform healthy behaviors if they believe there is a high risk that the symptoms and complications could adversely affect their lives and that the performance of healthy behaviors is an effective method for dealing with these symptoms. The presence of symptoms and complications may prompt patients to adopt healthy behaviors. On the other hand, it can be interpreted that abdominal discomfort or pain can occur due to certain food intolerance [20]. Patients who believe that their abdominal discomfort or pain is related to food intolerance may restrict their diet, possibly causing nutritional de ciencies and sarcopenia, which are predictive factors for poorer survival in patients with LC [31]. Thus, it should be clinically recognized whether the presence of abdominal discomfort or pain is associated with a healthy diet or different clinical scenarios (i.e., food allergy and intolerance, functional gastrointestinal disorder). In addition, a prospective study design (cohort or experimental study) is required to examine the causal relationship between the presence of symptoms and healthy diet behaviors.
The positive association between the development of nausea and higher SOC in engaging in exercise may be because exercise can induce gastrointestinal symptoms, such as nausea, heartburn, and abdominal pain [32]. This is because exercise can increase the level of catecholamine, which activates adrenergic receptors and induces nausea and vomiting [33]. In contrast, most longitudinal studies of patients with cancer have reported that those who exercise experienced signi cantly less intense nausea and greater alleviation of other symptoms [34]. Because the current study had a cross-sectional design, we cannot determine whether nausea led to increased exercise or if increased exercise led to nausea. However, it is unlikely that patients with LC would exercise so intensely as to cause gastrointestinal symptoms. Thus, we believe that the experience of gastrointestinal symptoms was a cue that increased SOC in engaging in exercise, according to the Health Belief Model. Further longitudinal studies are required to make a de nitive conclusion on the causal relationship between symptoms and the adoption of healthy behaviors.
Our results indicated that symptoms that were more frequent, intense, and distressing differed from the symptoms that affected QOL. In particular, patients whose symptoms were more frequent, intense, and distressing experienced fatigue, muscle cramps, and decreased memory, respectively.
These ndings are in line with those of a previous integrative review, which reported that fatigue was common among populations with chronic liver disease [35]. The general ndings from the current study, which found an association between symptoms and poor QOL in patients with LC, agree with the ndings of previous studies [36].
Experiencing muscle cramps can have a negative in uence on physical functioning and increase role limitations. Patients with LC frequently report painful muscle cramps [37] and typically describe them as abrupt, uncomfortable squeezing or contraction of a muscle that lasts seconds to minutes. Our nding supports the results of a previous study, which reported that although muscle cramps are not life threatening, they are a major concern and closely associated with the perception of poor health status among patients with LC [38]. Another previous study showed that muscle cramps are related to signi cantly decreased QOL in patients with LC [39]. Because muscle cramps are often intermittent and subjective and the predisposition to cramping is di cult to measure using a diagnostic tool, clinicians and researchers often overlook this symptom [39]. Although muscle cramps are easy to ignore, they can signi cantly affect the QOL of patients with LC; thus, professional clinical nurses should help in the assessment of this symptom.
Anorexia can negatively affect physical and social functioning and increase role limitations. This nding was consistent with previous studies, which reported that patients with LC and gastrointestinal symptoms had profound reductions in physical functioning, based on SF-36 scores [40], and that malnutrition due to anorexia was signi cantly associated with self-perception of lower QOL in patients with LC [41].
RUQ pain and body pain can also negatively affect physical and social functioning. Our rating of symptoms based on their frequency, intensity, and distress indicated that RUQ pain ranked 11th and body pain ranked 14th among all symptoms. Although pain is not among the most serious symptoms, a small change in patients' pain experience may signi cantly affect QOL; thus, it is important that professional clinical nurses carefully assess each patient's individual experience of pain. This nding supports a previous study that compared age-and gender-matched controls and found that patients with LC were impaired in all QOL domain scores [38].
Itching and ascites/edema had a negative impact on physical functioning. Pruritus is an extrahepatic symptom that is the greatest burden for patients with LC, and our study rated it as the fourth-most serious symptom overall. In patients with LC, persistent itching can lead to severe sleep loss, depression, and suicidal thoughts [42]. The assessment of itching severity allows objective assessment of its in uence on a patient's health and QOL, so that different therapies can be implemented. Based on the reports of previous studies, ascites, serum sodium levels, and lower extremity edema were independently associated with impaired physical functioning [43] and decreased QOL in patients with LC [44]. Thus, professional clinical nurses should perform systematic assessments of patients with LC who present with itching, ascites, and edema.
Role limitations due to emotional problems were greater in those who experienced changes in appearance. We speculate that a patient with LC who has readily visible symptoms (itching, edema, and ascites) may experience greater problems with body self-image and consequently greater emotional distress. Previous clinical studies of patients with cancer reported that changes in physical appearance that resulted from disease progression or treatment produce psychological distress [45,46]. The development of changes in appearance can have a profound impact on multiple functions such as working with others [47]. The present study of patients with LC indicated that emotional problems due to a poor body self-image seemed to be associated with greater role limitations. When referring a patient with LC for psychosocial care, health care professionals should be aware of a patient's body self-image as a potential indicator of poor role functioning.
Being employed was positively associated with social functioning [48], and being married was negatively associated with role limitations [49]. These results are in line with the common idea that a patient who is married or in a common-law relationship has better health-related QOL. Previous studies also found that socioeconomic factors affected QOL: in patients with hepatitis C virus who were married or who had a cohabitant had signi cantly better QOL than their single counterparts [50]. This is presumably because a spouse can provide social and emotional support that increases QOL.
This study indicated that symptoms had an impact on QOL, yet higher SOC in engaging exercise and consumption of F&V did not. Although some studies have reported an exercise-induced improvement in the metabolic pro le [51], a randomized controlled trial on home-based physical activity and diet intervention showed no QOL improvement [52]. Thus, more evidence based on prospective or experimental studies that supports the bene ts of exercise and diet in relation to cirrhosis is still needed, with emphasis on individuals with cardiovascular risk, musculoskeletal disorders, and complications related to cirrhosis. Cirrhosis patients require certi ed exercise and diet professionals who can perform a detailed functional assessment and design an individualized exercise and diet regimen to improve their QOL [6].
The symptom scores of our patients were lower than those reported in previous studies [25], possibly because the symptoms of LC patients may consist of a wide range of symptom development in those with advanced disease, whereas patients with LC may be asymptomatic for years.
Importantly, professional clinical nurses should keep in mind that although a symptom score may be low, the symptoms may still have a negative affect QOL and health-related behaviors.
Clinical nurses play pivotal roles in the systematic assessment of symptoms and the implementation of treatments in the management of patients with LC. Identifying symptoms that occur during the different stages of LC may inform the development of interventions that target multiple symptoms at once, thereby improving the effectiveness of symptom management. Knowledge of the impact of symptoms on QOL during the different stages of LC can guide best practices for symptom assessment. Our ndings suggest that professional clinical nurses require a deep understanding of the symptoms of LC patients.
The major limitations of this study are the use of a cross-sectional design and the small sample size. However, because many previous studies on this topic also examined about 100 individuals, we believe it is reasonable to compare our results with those of previous studies. In addition, our results suggest that the adoption of healthy behaviors had no signi cant effect on QOL. Therefore, further longitudinal or experimental studies on patients with LC are needed to identify the causal relationship between the adoption of healthy behaviors and QOL. This study should be a starting point for a prospective study design (cohort or experimental ones) to con rm whether healthy living habits have a long-term impact on the QOL of patients with cirrhosis. Despite these limitations, very few previous studies have identi ed the various types of symptoms and QOL in patients with LC.
In conclusion, this study of LC patients indicated that the type of symptoms experienced by a patient can lead to the adoption or rejection of healthy behaviors and that the patient symptoms can also negatively affect QOL. Patients with LC have symptoms that differ in frequency, intensity, and distress and that have different effects on QOL. Professional clinical nurses should carefully assess patients with symptoms of anorexia, RUQ pain and body pain, ascites/edema, and change of appearance, as these can have a negative effect on QOL, even among those who appear to be asymptomatic. Patients with LC require specialized support for the assessment and management of symptoms. Identi cation of the association between symptoms with the intention of adopting healthy behaviors and QOL provides a basis for the development of symptom management strategies and other interventions that may improve the QOL of patients with LC.

Consent for Publication
Not applicable.

Availability of Data and Materials
The data sets generated during and/or analyzed during the current study are available from the corresponding authors on reasonable request.

Competing Interests
The authors declare that they have no competing interests.      a The model included stage of change in engaging in exercise and consumption of vegetables, sociodemographic characteristics, and all subscales of symptoms as the main independent variables. A higher score on functioning indicates better functioning, and a higher score on role limitation indicates more severe role limitations.