|Home | About | Journals | Submit | Contact Us | Français|
Nonalcoholic Fatty Liver Disease (NAFLD) is highly prevalent and associated with decreased quality of life (QOL). The currently available treatment is weight loss through lifestyle modification. However, longitudinal QOL data are lacking on whether weight loss improves QOL. We prospectively enrolled 151 patients with NAFLD from 2009-2014. All patients received a liver biopsy, lifestyle assessment, blood tests and QOL tools including the Chronic Liver Disease Questionnaire (CLDQ), a validated health-related quality of life measurement. All patients were followed with a repeat CLDQ at 6 months. The cohort included 91 (60%) men, aged 51.5 ± 12.6 years, 46 (30%) of whom were diabetic. Thirty (21%) had advanced fibrosis or cirrhosis and 67 (47%) had a NAFLD activity score (NAS) > 4. Overall, 47 (31%) patients achieved at least a 5% reduction in weight. The cohort's median baseline total CLDQ value was 5.6 IQR (4.8 – 6.2). Those who achieved at least a 5% reduction in weight had a 0.45 (95% CI:0.24 – 0.66, p<0.0001) point improvement in the total CLDQ, compared to 0.003 (95% CI:−0.12 – 0.12, p=0.95) in those who did not. Nondiabetic patients with Nonalcoholic Steatohepatitis and without advanced fibrosis are most likely to achieve QOL benefits from weight loss. For every decrement in body mass index (BMI) there was a corresponding increase of 0.09 95% CI (0.03 – 0.16) points in the CLDQ scale (p = 0.005), adjusting for histology, diabetes, gender, age, and change in alanine aminotransferase level and change in FIB-4 index. A decrease by 5 points in BMI leads to a 10% adjusted improvement in QOL.
Patients with NAFLD can experience significant improvements in QOL that appear specific to weight loss and not biochemical improvements.
Nonalcoholic fatty liver disease (NAFLD) is rapidly becoming the major cause of liver disease related morbidity and mortality in most developed nations.(1-3) A spectrum of disease states, NAFLD begins as simple steatosis, but may progress to steatohepatitis (NASH), cirrhosis and hepatocellular carcinoma.(4, 5) With the increasing prevalence and complications of progressive NAFLD, healthcare costs attributable to this condition have risen sharply.(2, 6) Along with the monetary costs, NAFLD exacts an important and significant price by way of reduced quality of life (QOL).(7, 8) Beyond their importance for counselling patients, understanding the baseline determinants of QOL in the setting of NAFLD is critical for the planning and execution of patient-centered outcomes and cost-effectiveness research.(9) However, these data are limited. Equally important are data regarding the predictors and magnitude of QOL changes following a therapeutic intervention for which the present literature also offers limited guidance.
In the first report on QOL in NAFLD, Dan and colleagues surveyed 106 patients with using the Chronic Liver Disease Questionnaire (CLDQ) and compared the results to patients with viral hepatitis.(7) This study established that QOL is significantly lower for patients with NAFLD, particularly those with cirrhosis, female sex and high body mass index (BMI). Neither diabetes, hypertension, metabolic syndrome nor liver enzyme levels were associated with QOL. David et al confirmed these findings using the Short Form 36 (SF-36) in a cohort of 713 subjects from the NASH clinical research network with the additional finding that histologic NASH contributed to reduced QOL.(8)
In the absence of longitudinal QOL data in NAFLD, hypotheses regarding predictors of QOL improvement during treatment may be inferred in part from obesity research. Weight loss, particularly through lifestyle changes, is the mainstay of NAFLD therapy.(10) However, the published effect of weight loss on QOL is variable.(11) For example, two observational studies demonstrated QOL improvements following 5% weight loss(12, 13) but improvements in QOL were not directly linked to weight loss in a randomized trial of diet, exercise or both.(14) For patients with NAFLD, specifically, weight loss can achieve remarkable improvements in liver tests and, often, histology.(15, 16) The varied benefits of weight loss may therefore prove more impactful for those with NAFLD. Herein, we describe a prospective, longitudinal study of QOL in a cohort of American patients with NAFLD to determine predictors of QOL improvement.
The subjects for this study were enrolled in a NAFLD registry at Beth Israel Deaconess Medical Center. All patients were referred to a hepatology clinic with a specialization in NAFLD for evaluation of elevated liver enzymes or incidental hepatosteatosis on imaging. Criteria for enrollment into the registry included a histologic diagnosis of NAFLD within 3 months prior to the enrollment visit. Other forms of chronic liver diseases or daily consumption of greater than 20 grams of alcohol were excluded. Each patient was evaluated with serologic testing for viral hepatitis, hemochromatosis, autoimmune hepatitis, alpha-1 antitrypsin deficiency and Wilson Disease when appropriate.
Between December 2009 and September 2014, 206 patients received a formal intake visit, 173 underwent biopsies within 3 months, 162 completed baseline quality of life questionnaires and 151 completed 6 month follow-up questionnaires. Lifestyle modifications were recommended to all patients. Patients were instructed to perform a total 150 minutes of moderate intensity or 75 minutes of vigorous intensity cardiovascular exercise with an additional two days of resistance/weight training each week. Moderate intensity was defined for the patients as any activity that results in increased heart rate and lead to sweating; vigorous intensity as activities that achieve sweating and hard breathing. The same dietary recommendations were provided to all: portion control, elimination of sugary beverages and limited saturated fats. Compliance with these measures was patient-reported and assessed during clinical follow-up. No patient received any drug therapy (e.g. vitamin E) within the first 6 months of follow up as part of their trial of lifestyle modifications. This study has been approved by the BIDMC institutional review board.
Our primary outcome was an improvement in quality of life (QOL) as measured by the CLDQ. The CLDQ is a tool used to measure health related QOL that has been validated in multiple chronic liver diseases.(7, 17-20) It consists of 29 instrument items compiled from six domains which query patient reported experiences during the 2 weeks prior to questionnaire administration. The six domains include: Fatigue, Abdominal pain, Emotional function, Systemic symptoms, Activity and Worry. Each question requires a likert scaled response that ranges from ‘all of the time’ to ‘none of the time.’ Each dimension receives a score from 1-7 and from that an overall CLDQ score is determined between 1–7. Higher scores indicate better QOL. The CLDQ was administered at the time of a clinical visit by a trained research assistant.
The primary exposure variable was 5% weight loss, consistent with other studies of weight loss and QOL(11) as well as those examining weight loss and histologic improvements.(15, 16, 21) Additional exposure variables were recorded at study enrollment. These included demographics (age, ethnicity, sex), medical history (comorbidities including diabetes mellitus), physical exam findings (e.g. body mass index (BMI)), biochemical data (e.g. alanine and aspartate aminotransferase (ALT, AST)), behavioural data (enrollment in a weight loss program, minutes of exercise, cigarette smoking), non-invasive assessments for fibrosis (NAFLD Fibrosis Score (NFS) and FIB-4 index) and histological data (e.g. NAFLD activity score (NAS) and fibrosis stage). Physical exams, behavioural data and biochemical data were also collected at the 6 month visit.
Ultrasound guided liver biopsy was performed at study enrollment. Biopsies were interpreted by specialized hepatopathologists and reported in standardized fashion according to the Brunt scoring system with specific mention of METAVIR fibrosis stage and NAFLD activity score (NAS) as described previously.(22) NASH was defined by NAS score of 5 – 8. Advanced liver fibrosis was defined as fibrosis stages 3 – 4.
Two noninvasive indices of fibrosis were calculated at baseline and 6-month follow up for all patients: the NFS and FIB-4. The NFS is an algorithm based on the patient's age, body mass index, diabetes or impaired fasting glucose (yes/no), AST, ALT, platelet count and albumin. The result is a negative or positive number where lower values reflect lower risk. For reference, a cut-off of 0.676 optimizes NFS test characteristics for prediction of advanced fibrosis when the prevalence of advanced fibrosis in our sample exceeds 15%.(23) The FIB-4 is an algorithm based on age, ALT, AST and platelet count; scores of 1.45 and 3.25 possess excellent negative and positive predictive value, respectively.(24)
Data are summarized as mean ± standard deviation for normally distributed, median [25th and 75th percentiles] for non-normally distributed continuous outcomes, or counts and percentages for categorical outcomes. Normality was assessed using a Shapiro-Wilk test. A two-tailed p-value was considered significant when < 0.05.
A power calculation was performed for this study of a continuous response variable from matched pairs of study subjects. Prior data indicated that the standard deviation of the difference in QOL scores was 1.0.(7) Given our baseline QOL score of 5.6, we sought to power our study in order to detect a 5% improvement. Therefore, if the true difference in the mean response of matched pairs is 0.28, we would need to study 102 subjects with paired values to be able to reject the null hypothesis that this response difference is zero with probability (power) 0.8. The Type I error probability associated with this test of this null hypothesis is 0.05.
Logistic regression was performed to assess associations with binary outcomes (weight loss) with an odds ratio as its output. Linear regression was performed to assess continuous outcomes (QOL). For longitudinal assessments of QOL, we used paired t-testing for matched-pairs to determine the significance of changes on follow up. We divided our cohort using clinical features we felt, a priori, to affect the capacity for QOL improvement (the presence of diabetes, NASH or advanced fibrosis). After establishing the subsets in which significant changes in QOL could be achieved, we performed a multilevel longitudinal regression to quantify corresponding differences in QOL after weight loss, adjusting for baseline clinical factors. Given that we lack follow up histology, we examined whether the inclusion of longitudinal noninvasive fibrosis assessments were associated with changes in QOL. As the indices’ components overlap with other model variables (i.e. ALT, BMI, diabetes), the NFS and FIB-4 were assessed for collinearity before inclusion into the model. The NFS was highly collinear while the FIB-4 was mildly collinear. A stepwise selection procedure was performed for the regression model to confirm the priority of model elements leading to the inclusion of FIB-4 and the exclusion of the NFS. JMP Pro statistical discovery software (version 11) was used for statistical analyses.
The baseline demographics and clinical characteristics of our 151-patient cohort are detailed in Table 1, divided by whether they achieved significant weight loss. The significant differences between those who lost weight and those who did not included sex and Hispanic ethnicity. When both variables were examined in a multivariable logistic regression, only male sex was associated with weight loss (odds ratio 2.24 (95% CI (1.11 – 4.55)).
An association between weight loss and behavioural changes on follow-up was also queried. The only significant change observed was that, after 6 months of follow-up, patients who lost weight exercised an additional 9.0 95% CI (0.71 – 17.3) minutes daily on average compared to 4.1 95% CI (−11.8 – 3.6) fewer minutes for those who did not (p = 0.04).
We assessed all patients for a baseline measure of QOL using the CLDQ. The median score was 5.6 Interquartile range (IQR) (4.8 – 6.2). We represent our data as medians as they are nongaussian variables. However, for reference, our mean score was 5.4 ± 1.02 while the mean score Dan et al's landmark paper was 5.4 ± 1.0.(7) There were no baseline differences between patients with or without diabetes mellitus, active histologic NASH and advanced fibrosis or below versus above the median BMI. However, there were significant differences between men and women – 5.9 95% CI (5.2 – 6.2) versus 5.2 95% CI (4.4 – 6.0), p = 0.0002. Male sex accounted for an additional 0.32 95% CI (0.15 – 0.48) points in the overall CLDQ score. Sensitivity analyses for the absence of NASH, advanced fibrosis and adjustment for BMI or ALT did not change the magnitude or significance of this association. The remaining baseline values for the CLDQ dimensions are described in Table 2.
Table 2 also shows the degree of change in QOL and serologic evidence of disease activity for the subjects who did and did not lose weight. Weight loss was associated with significant incremental improvements in QOL overall and with respect to abdominal symptoms (p = 0.003), fatigue (p = 0.0003), activity (p = 0.02) and worry (p = 0.01). Notably, patients who did not lose weight did not report an improvement in any QOL dimension save for worry where the average improvement was 0.27 95% CI (0.08 – 0.46). With respect to serologic assessments, weight loss was associated with larger improvements in ALT and AST but non-significant improvements in the NFS and FIB-4.
Table 3 examines the cohort subgroups in order to describe which clinical factors were associated with significant improvements in QOL. Among patients with weight loss, those who experienced significant improvements in QOL were those with active NASH, without advanced fibrosis, and without diabetes mellitus.
In a multilevel longitudinal regression model adjusting for baseline histologic status (fibrosis, NAS score), diabetes, gender, age, change in alanine aminotransferase level and change in FIB-4 for every decrement in BMI there was a corresponding increase of 0.09 95% CI (0.03 – 0.16) points in the CLDQ scale (p = 0.005). For reference, a decrease by 5 points in BMI leads to a roughly 10% improvement in QOL. Changes in ALT and FIB-4 were not significantly associated with change in QOL, though the change in FIB-4 approached significance. For each unit increase in FIB-4 there was a corresponding decrease in CLDQ score of −0.28 95% CI (−0.58 - −0.01), p = 0.06.
Clinicians who care for patients with NAFLD uniformly seek to improve their clinical outcomes. Beyond the prevention of liver disease complications, an important outcome is the patient's quality of life (QOL). Understanding the effect of weight loss on measures of QOL for patients with NAFLD is essential to counsel patients, power clinical studies with patient-centered outcomes and plan robust cost-effectiveness models. Data regarding baseline QOL for American patients with NAFLD has been established in two robust cohorts.(7, 17, 25) However, data regarding the effect of therapeutic interventions on QOL for patients with NAFLD is lacking. In this study of patients with biopsy proven NAFLD, we show that weight loss is associated with significant improvements in QOL. Furthermore, we show that the subgroups most likely to benefit include non-diabetics with active NASH and F0-F2 fibrosis.
This study confirms and extends the literature on NAFLD and QOL in a few important ways. First, and most importantly we show that weight loss affects QOL. Our results contrast significantly with a recent meta-analysis showing that weight loss is generally not associated with changes in QOL.(11) Notably, their analysis could not include CLDQ as no prior longitudinal studies had been conducted with this metric. Additionally, among the subset of studies in which a 5% weight loss was reported, patients attaining that goal did experience significant improvements in QOL.(12-14) In this study of patients evaluated at a NAFLD referral center, weight loss was significantly associated with improved QOL, particularly those with diabetes or NASH and without advanced fibrosis.
Second, we confirm that the mean baseline CLDQ score of patients presenting with NAFLD; our cohort's mean score was 5.4, the same as that seen 8 years prior in a different American region by Dan and colleagues.(7) Third, like Dan(7) and David et al(8), we confirm that among all clinical characteristics, sex is a significant determinant of baseline QOL scores in a cohort. Fourth, we also show that patients who lost weight increased their daily exercise commitments while patients who did not lose weight did not.
These findings must be understood in the context of our study design. First, our cohort has a significant burden of advanced NAFLD with many patients exhibiting active NASH and advanced fibrosis on biopsy. These findings reflect that ours is a referral cohort with inclusion requiring a liver biopsy. The requirement for histology is likely an intrinsic source of a bias leading to the inclusion of patients who are more likely to have advanced histology. Further, in view of the limited improvements in QOL for patients with advanced fibrosis who achieved weight loss, it is likely that these features rendered the findings with respect to the QOL benefits of weight loss more conservative. Indeed, the subgroups most likely to benefit were non-diabetics without advanced fibrosis and with active NASH. These subgroups may also be better able to lose weight. Second, we cannot compare changes in histology with QOL changes as we did not perform a biopsy at 6 months. However, both the NFS and FIB-4 – validated noninvasive indices of fibrosis – do not vary significantly with weight loss and changes in FIB-4 are not associated with changes in QOL. Further, while inflammation and steatosis may improve with weight-loss, it is unlikely that fibrosis burden could change significantly within 6 months of follow up or after 5% weight-loss.(21) Third, future studies of long-term effects from weight loss are needed to confirm the demonstrated 6 month results.
In conclusion, weight loss is associated with significant short-term improvements in QOL for patients with NAFLD.