Objective Chronic liver disease (CLD) is a major cause of mortality and morbidity worldwide. The aim of this study was to assess the overall and liver-related mortality and their predictors in patients with CLD using population data.
Methods The Third National Health and Nutrition Examination Survey (NHANES III) and linked mortality data were utilised. Participants from NHANES III (1988–1994) and their mortality status were updated by the Center for Disease Control (CDC) as of 31 December 2006. In this study, the aetiology of CLD was based on available serological tests and clinical data. Each diagnostic cohort was compared with a cohort without liver disease using stratum-specific χ2. The Cox proportional hazard model was used for analysis, and HRs adjusted for all major confounders; overall mortality and cause-specific mortality were calculated for each type of liver disease. All analyses were run using SAS-callable SUDAAN 10.0 functions using remote access to the CDC Research Data Center server with restricted use linked mortality data.
Results The study cohort included 15 866 NHANES III participants with complete clinical, demographic, laboratory and mortality follow-up data. Of these, 235 subjects had alcohol-related liver disease (ALD), 66 had chronic hepatitis B (CH-B), 264 had chronic hepatitis C (CH-C), 991 were presumed to have non-alcoholic fatty liver disease (NAFLD) and 505 had other liver disease. Additionally, 13 004 subjects without evidence of liver disease served as controls. The analysis shows that type II diabetes (DM) and/or insulin resistance (IR) are independent predictors of overall mortality in CH-B, NAFLD and ALD (p <0.05). Additionally, DM, IR, obesity and metabolic syndrome could be independent predictors of liver-related mortality in CH-C, NAFLD and ALD.
Conclusions Components of metabolic syndrome are associated with overall and liver-related mortality in subjects with CLD.
- Non-alcoholic steatohepatitis
Statistics from Altmetric.com
Significance of this study
What is already known about this subject?
Data from tertiary care centres suggest that the presence of diabetes and obesity has a negative impact on the progression of liver disease related to hepatitis C and non-alcoholic fatty liver disease.
Obesity also impacts treatment outcomes for hepatitis C.
A population-based study to assess the impact of component of metabolic syndrome on the outcome of different types of liver disease is not available.
What are the new findings?
This population-based study suggests that components of metabolic syndrome (diabetes, insulin resistance and obesity) are associated with liver-related mortality for major causes of liver disease.
The study also suggests that components of metabolic syndrome may be associated with overall mortality for some types of liver disease.
How might it impact on clinical practice in the foreseeable future?
The evidence from this study suggests that aggressive treatment of components of metabolic syndrome for any type of liver disease may favourably impact liver-related outcomes such as liver-related mortality.
In recent years, National Vital Statistics reported chronic liver disease (CLD) as the 12th leading causes of death in the USA.1 2 Furthermore, growing evidence suggests that patients with chronic liver disease in the setting of metabolic syndrome (MS), specifically type 2 diabetes (diabetes mellitus (DM)), are at higher risk for disease progression.3–8 This elevated risk is of special interest given the currently observed worldwide epidemic of obesity and MS. Accumulating evidence suggests that the association of MS with the prognosis of patients with CLD could be bidirectional and largely independent of the aetiology of CLD.9 Indeed, while metabolic conditions affect the liver, damage to hepatic tissue from steatosis and fibrosis is known to affect the development of insulin resistance (IR) independently of changes in adipose tissue.10 Though confirmed by several well-designed clinical and experimental studies,11–18 this association has never been convincingly validated using population-based mortality data. Also, it is a matter of debate whether DM in the absence of obesity is a risk factor for CLD.19 Considering the rate of obesity and DM in the US population, it is important to confirm whether DM or IR or both represent independent risks factors for increased mortality among people with CLD, or whether the harmful effects observed in clinical studies can be solely attributed to concomitant obesity.
This study examines Third National Health and Nutrition Examination Survey (NHANES III) population data linked with nearly 18 years of mortality follow-up provided by the US National Death Index (NDI) to study the association of MS and its components with overall and liver-related mortality in subjects with CLD. Identification of independent risk factors for liver-related mortality, which obviously poses a significant burden to society, will help to improve public health management strategies and provide relevant information for physicians.
NHANES III is a nationwide survey conducted in the United States between 1988 and 1994 to gather information representing the health and nutritional status of the of the non-institutionalised civilian US population. The survey consists of interviews, standardised physical examinations and data from blood samples collected in special examination centres. After weighting on the basis of age, gender, level of education, and race or ethnic group, the distribution of participants was representative of that of the US population. Stratification and sampling units describing the design stages of NHANES III were used to account for the complex, multistage probability sample design of these data.20
Inclusion criteria were as follows: age of 17 years or older and the availability of complete demographic data (age, gender and ethnicity), social history data (history of smoking and alcohol consumption) and clinical data (history of hypertension, DM and family history of related problems). Furthermore, body mass index (BMI), waist circumference and blood pressure were measured at the time of examination for all eligible participants. The following laboratory tests were also available for all individuals included in the study: fasting serum glucose and insulin, triglycerides, high-density lipoprotein (HDL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), transferrin saturation, and viral hepatitis serologies for hepatitis B virus (HBV) and hepatitis C virus (HCV). Participants with incomplete data were excluded from the study.
The following definitions were used for the purpose of this study:
DM was defined as a fasting glucose ≥126 mg/dl or the history of oral hypoglycaemics or insulin use, or both.
Hypertension was defined as a systolic blood pressure (BP) of ≥140 mm Hg, diastolic BP of ≥90 mm Hg or history of oral antihypertensive medications.
Hypercholesterolaemia was defined as an elevated cholesterol of >200 mg/dl, low-density lipoprotein (LDL) ≥139 mg/dl, or HDL <40 md/dl for men and <50 mg/dl for women.
MS was defined by the National Cholesterol Education Program Adult Treatment Panel III (ATP III) criteria.21 According to ATP III, a subject is presumed to have metabolic syndrome if at least three of the following are present: waist circumference >102 cm in men or >88 cm in women, fasting plasma glucose >110 mg/dl, BP >130/85 mm Hg, elevated triglycerides >150 mg/dl and HDL <40 mg/dl in men or <50 mg/dl in women.
Elevated serum aminotransferases were defined as ALT >40 U/l or AST >37 U/l in men and ALT or AST >31 U/l in women.
Elevated transferrin saturation was diagnosed at the level of ≥50%.
IR was determined by the homeostasis of model assessment score (HOMA). A HOMA score of ≥3 was defined as IR.
Four major race or ethnic groups included non-Hispanic whites, non-Hispanic blacks, Hispanics and ‘other,’ which included Aleut, Eskimo, American Indian, Asian or Pacific Islander.
A positive smoking history was defined as ongoing smoking or >100 cigarettes in the past 5 years.
Excessive alcohol consumption was defined as >20 g per day for men and >10 g per day for women. The alcohol intake was calculated using self-reported data on amount and frequency of alcohol consumption collected as a part of NHANES III examination questionnaire.
Obesity was defined as BMI ≥30 or waist circumference ≥102 cm in men or ≥88 cm in women.
Aetiology of CLD
Four major categories of CLD were considered: alcoholic liver disease (ALD), chronic hepatitis B (CH-B), chronic hepatitis C (CH-C) and non-alcoholic fatty liver disease (NAFLD). ALD was presumed in subjects who reported excessive alcohol use for the past 5 years and had elevated serum aminotransferases. Participants were tested for hepatitis B core antibody and, if positive, hepatitis B surface antigen (HBsAg) was obtained. Those who were positive for HBsAg were considered to have CH-B. Similarly, hepatitis C antibody was tested and, if positive, HCV RNA was obtained. Participants with positive HCV RNA were considered to have CH-C. Finally, subjects were presumed to have NAFLD if they had elevated serum aminotransferases in the absence of any other evidence of CLD such as excessive alcohol use, elevated transferrin saturation or positive hepatitis B or hepatitis C tests. Another group of patients with transferrin saturation >50% may have had iron overload and were included in the initial analysis as ‘other liver disease’. Subjects without evidence of liver disease by our inclusion criteria were considered to have no CLD, and those individuals were used as a control cohort.
Mortality data including causes of death were obtained through the US NDI. The NDI is a computerised database of all certified deaths in the USA since 1979. All NHANES III participants ≥17 years had mortality follow-up until 31 December 2006. In addition to mortality status, the NHANES III-Linked Mortality File (accessible by request through the Research Data Center of the US Centers for Disease Control and Prevention (CDC)) contains dates of death as well as causes of death. According to the NDI database, people who died in the USA prior to 1998 were classified by the 9th revision of the International Statistical Classification of Diseases, Injuries, and Causes of Death (ICD-9) guidelines. After 1998, they were classified by the 10th revision (ICD-10) guidelines.22 The NHANES III-Linked Mortality File used the Underlying Cause of Death 113 (UCOD_113) code to recode all deaths according to ICD-9 and ICD-10 criteria.
In addition to overall mortality, four cause-specific mortalities were also studied. For this purpose, cardiovascular mortality (UCOD_113 58–63, 67, 70–74) covered causes of death such as ischaemic heart diseases (ICD-10 I20–I25), heart failure (I50), atherosclerosis (I70), cerebrovascular diseases (I60–I69), aortic aneurysm and other diseases of arteries, arterioles and capillaries (I71–I78). Solid organ malignancy-related mortality (UCOD_113 20–36, 43) included all malignant neoplasms except for those of lymphoid, haematopoietic and related tissue (C00–C80). Liver-related mortality (UCOD_113 15, 24, 93–95) included causes of death such as viral hepatitis (B15–B19), hepatocellular carcinoma (C22), alcoholic liver disease (K70) and other CLD and cirrhosis (K73–K74). Diabetes-related mortality (UCOD_113 46) accounted for the subjects for which DM (E10–E14) was listed as an underlying cause of death. Individuals without available mortality follow-up data were ineligible for the study.
As a first step, subjects with any type of CLD (ALD, CH-B, CH-C, NAFLD and other liver disease) were compared with the controls without CLD by using the stratum-specific χ2 test for independence. Then, to study mortality risk factors, the Cox proportional hazards model was used to identify independent predictors for the outcome. The studied outcomes included overall mortality and cause-specific mortalities in subjects with various aetiologies of CLD. In the hazards models, mortality predictors included the presence of CLD and MS components adjusted for potential confounders such as demographic parameters, history of smoking and alcohol use. Adjusted HRs (aHRs) were calculated. Follow-up time was from the date of the NHANES III examination to the date of death, which was followed until 31 December 2006. Proportional hazard assumption was controlled for all potential mortality predictors. Sample weights were used to account for non-response and unequal selection probabilities for certain categories of population, and stratum and sampling units accounted for the survey design effects using Taylor series linearisation. All analyses were run using SAS-callable SUDAAN 10.0 functions (SAS Institute, Cary, North Carolina, USA) using remote access to the CDC Research Data Center (Hyattsville, Maryland, USA) server with restricted use-linked mortality data.
The study was approved by the Inova Institutional Review Board.
NHANES III population
After applying inclusion and exclusion criteria, a total of 15 866 subjects were examined from NHANES III. These subjects had complete demographic, clinical, laboratory, social history and mortality data. The subjects were matched with a restricted access mortality database generated by the National Center for Health Statistics using NDI. Demographic data of the study cohort revealed that 47.9±0.5% were male, 76.9±1.2% Caucasian, 10.4±0.6% African American and 5.2±0.4% Hispanic. Furthermore, 11.78±0.51% of the cohort had CLD (ALD, 1.38±0.18%; CH-B, 0.36±0.07%; CH-C, 1.44±0.22%; NAFLD, 5.51±0.33%; and other liver disease, 3.59±0.25%). Of the rest of the population, 81.5% had no evidence of liver disease. The proportions with and without CLD do not add up to 100% because we could not rule in or rule out CLD in subjects who reported a history of excessive alcohol use but had normal liver enzymes, so we did not include them in any of those groups.
As returned by pairwise comparison, subjects with CLD were younger than controls (41.1±0.6% vs 44.4±0.4%, p<0.0001), predominantly male (56.5±1.9% vs 45.7±0.6%, p<0.0001), non-white (72.7±1.9% vs 77.1±1.2%, p=0.002) and a significantly higher portion was Hispanic (8.0±0.8% vs 4.9±0.4%, p<0.0001). In addition, CLD subjects had a higher prevalence of smoking history (34.5±1.8% vs 26.6±0.8%, p=0.0001) and MS (30.3±1.7% vs 25.8±0.8%, p=0.002) as well as its components such as DM (8.1±0.8% vs 5.5±0.3%, p=0.003) and IR (34.4±2.1% vs 22.5±0.9%, p<0.0001). CLD subjects also had a higher BMI (27.1±0.3 vs 26.4±0.1%, p=0.014), although the prevalence of obesity was not significantly different (40.30±2.04 vs 38.0±0.8, p=0.29). All other parameters of these two major cohorts were not significantly different. A descriptive summary of the subjects with different aetiologies of CLD is given in table 1.
Aetiologies of CLD as a risk factor for overall and cause-specific mortality
A total of 3662 subjects died during the follow-up, with median length of 160 months. The overall mortality rate for the study cohort was 16.3±0.7%. The overall mortality rate of patients with CLD was 14.6±1.0% (16.3±2.9% in ALD, 22.4±7.1 in CH-B, 22.5±4.61% in CH-C and 11.8±1.6% in NAFLD).
Adjusted HRs for different aetiologies of CLD returned by the Cox proportional hazards model suggested that only CH-C was an independent predictor of overall mortality (HR (95% CI) 2.73 (1.77 to 4.20), p<0.0001).
Most of the aetiologies of CLD considered in this study were independent predictors of liver-related mortality (CH-B, 20.31 (4.74 to 87.04), p=0.0001; CH-C, 22.41 (4.99 to 100.74), p=0.0001; NAFLD, 7.53 (2.26 to 25.14), p=0.001).
For diabetes-related mortality, viral hepatitis B and C were independent predictors (CH-B, 26.72 (4.13 to 172.62), p=0.0009; CH-C, 7.55 (1.92 to 29.74), p=0.004). Finally, none of the CLDs was independently associated with cardiovascular- and solid organ malignancy-related mortalities.
Independent predictors of overall and cause-specific mortality in each type of CLD
Mortality in the CH-B cohort
Seventeen of the 66 CH-B subjects died during the follow-up. After weighting, the overall mortality rate in the CH-B cohort was slightly higher than in the control cohort (22.3±7.0% vs 16.5±0.7%, p=0.40). A demographic summary of the CH-B cohort in comparison with controls is included in table 1. Specifically, subjects with CH-B were predominantly non-white (47.6±10.0% vs 77.1±1.2%, p=0.0006) and a majority of them in fact belonged to the ethnic group designated as ‘other’ (23.2±7.1% vs 7.5±0.8, p=0.028). This group included Asians, Eskimos or Pacific Islanders in whom a higher prevalence of CH-B has been confirmed in numerous clinical studies.23–25 CH-B subjects were also more frequently male (65.2±8.8% vs 45.7±0.6%, p=0.042) and had a significantly lower prevalence of MS and its components (table 1). However, after adjusting for all studied confounders, DM was significantly associated with increased mortality in the CH-B cohort (aHR (95% CI)=30.79 (8.36 to 113.42), p<0.0001). This means that a subject with CH-B and DM has ∼30 times the risk of death as the same subject without DM. No other factors were independent mortality predictors. The low prevalence of CH-B in this cohort did not allow us to study other cause-specific mortalities.
Mortality in the CH-C cohort
Out of 264 subjects with CH-C, 81 died during the follow-up, with an overall mortality rate of 22.49±4.61%. A demographic summary of the CH-C cohort compared with controls is given in table 2. Specifically, subjects with CH-C were younger (40.8±1.1 years vs 4.4±0.4 years, p=0.011), predominantly non-Hispanic black (21.8%±3.2 vs 10.6±0.6%, p=0.0005) and male (66.0±5.0% vs 45.7±0.6%, p=0.004). They also reported a history of smoking significantly more often (64.0±5.3% vs 26.6±0.8%, p<0.0001), and had a higher prevalence of IR (37.4±3.2% vs 22.5±0.9%, p=0.001) and hypercholesterolaemia (48.8±6.1% vs 68.6±0.9%, p=0.020).
Independent predictors for overall mortality in CH-C subjects were older age (aHR 1.09 (1.06 to 1.12), p<0.0001) and male gender (aHR 2.39 (1.32 to 4.33), p=0.005). However, in patients with CH-C, in addition to older age (aHR 1.06 (1.01 to 1.11), p=0.020), both DM and IR were independently associated with liver-related mortality (aHR 2.20 (0.97 to 3.84), p=0.076; and aHR 3.48 (1.10 to 11.01), p=0.034, respectively). Furthermore, liver-related mortality was independently associated with other components of MS such as hypertension (aHR 5.17 (1.18 to 22.73), p=0.030) and obesity (aHR 11.16 (1.80 to 69.17), p=0.011) as well as with MS itself (aHR 3.38 (1.52 to 6.66), p=0.005).
In patients with CH-C, cardiovascular mortality was associated with age (aHR 1.10 (1.05 to 1.15), p=0.0001), male gender (aHR 8.45 (2.46 to 29.00), p=0.001) and hypertension (aHR 15.20 (1.40 to 165.55), p=0.026).
Interestingly, in this CH-C cohort, diabetes mortality, even when adjusted for type 2 diabetes, was still significantly associated with excessive alcohol use (aHR 10.46 (3.28 to 33.35), p=0.0002) and hypercholesterolaemia (aHR 5.21 (1.04 to 26.12), p=0.045).
Finally, solid organ malignancy-related mortality was, in addition to demographics, only associated with smoking history (not shown). A summary of all independent mortality predictors for the CH-C cohort is given in table 3.
Mortality in the NAFLD cohort
Of 991 subjects presumed to have NAFLD, 177 died during follow-up. Overall mortality rate in this cohort was 14.31±1.58%. A demographic summary of the NAFLD cohort in comparison with controls is given in table 1. As shown, subjects with NAFLD were slightly younger (42.7±0.7 years vs 4.4±0.4 years, p=0.025); a larger proportion was Hispanic (10.2%±1.2 vs 4.9±0.4%, p<0.0001) and male (53.2±2.8% vs 45.7±0.6%, p=0.015). NAFLD subjects reported a history of smoking less frequently (20.8±2.2% vs 26.6±0.8%, p<0.0001) and had a significantly higher prevalence of all MS components including obesity, DM, IR and hypertension together with hypercholesterolaemia. These observations are consistent with the previously reported characteristics of patients with NAFLD and indirectly confirm that our definition of NAFLD for this study was reasonable despite the absence of confirmatory liver biopsy.
Independent predictors for overall mortality in NAFLD subjects were older age (aHR 1.10 (1.08–1.12), p<0.0001) and the presence of type 2 diabetes (aHR 2.31 (1.17 to 4.56), p=0.017).
For liver-related mortality in patients with NAFLD, both DM and IR carried significant hazard (aHR 1.05 (1.00 to 1.65), p=0.052 for DM, and aHR 53.55 (9.22 to 344.29), p<0.0001 for IR, respectively) in addition to older age (aHR=1.10 (1.08 to 1.12), p=0.020) and male gender (aHR 9.53 (1.36–66.55), p=0.024). Furthermore, liver-related mortality in the NAFLD cohort was also independently associated with obesity (aHR 11.19 (2.43 to 51.56), p=0.003) as well as for MS in general (aHR 12.08 (1.10 to 132.22), p=0.042).
Cardiovascular mortality in NAFLD subjects was associated with type 2 diabetes (HR 2.76 (1.26 to 6.05), p=0.012). Diabetes mortality in the NAFLD cohort, besides the presence of type 2 diabetes and IR, was significantly associated with hypercholesterolaemia (aHR 39.52 (13.30 to 235.47), p<0.0001). A summary of independent mortality predictors in the NAFLD cohort is given in table 4.
Mortality in the ALD cohort
Of a total of 235 participants with ALD, 56 died during the follow-up, with an overall mortality rate of 16.30±2.90%. A demographic summary of the ALD cohort compared with controls is given in table 1. As shown in the summary, subjects with ALD were slightly younger than controls (41.6±1.5 years vs 4.4±0.4 years, p=0.117), predominantly male (61.5±6.1% vs 45.7±0.6%, p=0.018), and represented a significantly higher proportion of the Hispanic population (8.7±1.6% vs 4.9±0.4%, p=0.004). They also more often reported a history of smoking (49.0±5.3% vs 26.6±0.8%, p=0.0003), and had higher prevalence of IR (33.7±3.2% vs 22.5±0.9%, p=0.043) and hypertension (36.5±5.7% vs 22.2±0.8%, p=0.014).
Independent predictors for overall mortality in ALD subjects were smoking (aHR 3.38 (1.04 to 11.02), p=0.044), DM (aHR 3.00 (1.06 to 8.54), p=0.039) and IR (aHR 3.21 (1.56 to 6.58), p=0.002).
The same risk factors were independently associated with cardiovascular mortality among participants with ALD (smoking, aHR 6.60 (1.06 to 41.12), p=0.044; DM, aHR 19.91 (1.67 to 237.51), p=0.019; IR, aHR 6.43 (1.77 to 23.35), p=0.006) together with hypertension (aHR 7.55 (1.54 to 36.89), p=0.014), but not other components of MS.
Finally, liver-related mortality in ALD subjects was independently associated with older age (aHR 1.15 (1.08 to 1.23), p=0.0001), male gender (aHR 47.14 (5.46 to 406.93), p=0.0008), DM (aHR 3.60 (0.96 to 13.52), p=0.057), obesity (aHR 16.22 (1.91 to 137.68), p=0.012) and MS in general (aHR 2.06 (1.21 to 3.31), p=0.001). The summary on mortality risk factors for people with ALD is given in table 5.
Diabetes-related mortality in the ALD cohort was not studied because of low sample size, and solid organ malignancy-related mortality was not independently associated with any risk factor other than older age (data not shown).
The prevalence of obesity in the USA has grown over the past 30 years, with almost a third of the US population now considered to be obese. Growing evidence also suggests that MS and its components may be associated with the development and progression of CLD. This association is especially important for NAFLD, which is considered to be the hepatic manifestation of MS.26 Studies originating from tertiary care centres and retrospective data analyses of patients with NAFLD have already demonstrated that MS and its components are risk factors for the development of liver fibrosis and its progression to more severe stages of liver disease, such as cirrhosis and hepatocellular carcinoma.3–8 15–18 Similar studies have suggested that the prevalence of type 2 diabetes and MS is increased in patients with CH-C and that patients with CH-C are significantly more likely to have evidence of liver damage such as steatosis and fibrosis.5 6 Less is known about the association of MS and CH-B, though some studies reporting preliminary results suggest that MS components, such as visceral obesity, diabetes, hypertension and hyperlipidaemia, may be associated with the presence of advanced fibrosis in patients with CH-B.27 28
This study is the first to report independent factors associated with overall mortality and cause-specific mortality in a large population-based cohort with an average of 160 months of follow-up. In part, this population-based study was designed to assess the independent association of MS and its components with increased overall and cause-specific mortality in patients with the most common causes of CLD. This is the largest study demonstrating the association of components of MS, specifically, type 2 diabetes, IR and obesity, with liver-related mortality and/or overall mortality in patients with viral hepatitis B and C as well as patients with NAFLD and ALD. Our results largely confirm previous findings from smaller studies originating from tertiary care medical centres and provide the strongest evidence to date supporting the association of components of MS with adverse outcomes in patients with HCV, NAFLD and ALD. Although in a smaller cohort, our data are also the first to show independent association of DM with mortality in patients with CH-B.
Although the association between NAFLD and components of MS has been previously reported, this study provides the strongest evidence to support the notion that DM and IR increase the risk for liver-related mortality in NAFLD. The data reporting a higher rate of IR in patients with CH-C indirectly support the hypothesis that hepatitis C infection alone may directly influence glucose and lipid metabolism,29 30 and increases the risk for development of type 2 diabetes, which begins with IR. Finally, this is the first study to report that both overall and liver-related mortality are independently associated with DM and IR in patients with ALD. This confirms previous reports that the other component of MS, obesity, has a detrimental effect on ALD.31
An important limitation of this study is that some participants with liver disease may have been classified as controls without liver disease if their liver enzymes were normal at the time of examination. In fact, some patients with normal liver enzymes may have had underlying NAFLD or ALD that would only have been determined by an ultrasound or liver biopsy. Additionally, a relatively low threshold for definition of excessive alcohol consumption might lead to classifying some subjects with NAFLD as having ALD. Nevertheless, the prevalence of metabolic conditions such as obesity and diabetes in the respective cohorts indirectly supports the representativeness of our NAFLD cohort and the lack of definition-caused bias in the ALD cohort. We were also unable to stage liver disease in the absence of histological data that might be useful for determination of the potential liver-associated risk factors. Finally, although the length of follow-up (12–18 years) is comparable with the time period typically necessary for CLD to progress to life-threatening stages, longer follow-up may still be necessary.
In conclusion, this study provides the strongest evidence to date showing that individuals with CH-C, NAFLD and ALD superimposed with components of MS (type 2 diabetes, IR, hypertension or elevated BMI) are at higher risk of mortality and/or liver-related mortality. We postulate that treating components of MS in patients with CLD may improve the outcomes of these patients. Prospective long-term studies of patients with CLD treated with these modalities are necessary to demonstrate the efficacy of such interventions.
Linked articles 219907.
Funding This study was partly supported by the Liver Outcomes Research Fund of The Center for Liver Diseases, Inova Fairfax Hospital, Falls Church, Virginia.
Competing interests None.
Ethics approval The study was approved by the Inova Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.