Abstract
Background and Objectives:
Statins improve portal hypertension and may reduce mortality in cirrhosis, though their impact on chronic liver disease (CLD) progression before developing advanced fibrosis or cirrhosis remains unclear. We aimed to assess the association of statin use with severe liver disease (SLD) outcomes (cirrhosis or HCC) and fibrosis progression in CLD patients.
Methods:
We performed a retrospective cohort study of primary care electronic health record data from 2012-2021. Included were patients with CLD, lab inputs for a Fibrosis-4 (FIB-4) calculation and no prior SLD. Analysis 1 assessed the association of statin prescriptions with time to SLD in all patients. Analysis 2 evaluated the association of statin prescriptions with time to high-risk FIB-4 (≥3.25) in patients with baseline FIB-4 <3.25. Unadjusted and adjusted Cox and Weibull regression models were used.
Results:
The cohort included 2,872 CLD patients (median index FIB-4 1.2 IQR: 0.7-1.9) followed for a mean of 5.7 (±2.6) years. 884 (31%) patients received statins and 378 (13%) experienced an SLD outcome. Statin prescription was associated with a lower risk of SLD (adjusted HR 0.51; 95% CI 0.39-0.66). Analysis 2 included 2,547 patients with an index FIB-4 <3.25 followed for a mean of 5.4 (±2.6) years. 855 (34%) patients received statins and 600 (24%) progressed to a FIB-4 ≥3.25. Statin prescription was associated with a lower hazard of progressing to high-risk FIB-4 (adjusted HR 0.36; 95% CI 0.32-0.42).
Conclusion:
In a primary care population, statin prescriptions were associated with a lower risk of fibrosis progression and SLD.
Keywords: Cirrhosis, fibrosis, hepatocellular carcinoma, primary care
Introduction
Chronic liver disease (CLD) and its complications are associated with a great burden of morbidity, mortality (>44,000 deaths in the United States and 2 million deaths worldwide annually), disability, and health care utilization.1,2 Despite this, few pharmacologic options are approved to attenuate disease progression in patients with CLD. Statins (3-hydroxyl-3-methylglutaryl coenzyme A [HMG-CoA] reductase inhibitors) have been shown to have numerous pleiotropic effects beyond their cholesterol-lowering properties. In patients with cirrhosis, statins have been shown to reduce the severity of portal pressure by reducing hepatic vascular resistance.2 They are also associated with a reduced risk of cirrhosis, hepatocellular carcinoma (HCC), and liver-related deaths in patients with CLD.3 In patients with chronic Hepatitis C infection, statins were also associated with a reduction in fibrosis progression.4 However, the impact of statins on liver disease progression before the development of advanced fibrosis remains unclear. In this study, we aimed to determine the relationship between statin use and the risk of fibrosis progression as well as the development of a severe liver disease outcome (including cirrhosis, a complication of cirrhosis, HCC, or orthotopic liver transplantation [OLT]).
Methods
Study Design and Setting
We performed a retrospective cohort study utilizing electronic health record (EHR) data from July 2012 – December 2021 in an internal medicine primary care clinic located in Charleston, South Carolina.
Patients
We identified patients with a diagnosis of CLD, comprising viral hepatitis (B and C), alcohol-related liver disease (ALD), metabolic dysfunction-associated steatotic liver disease (MASLD, formerly nonalcoholic fatty liver disease), hemochromatosis, autoimmune hepatitis, primary biliary cholangitis, Wilson’s disease, and α-1 anti-trypsin deficiency, using a previously validated set of ICD-9/10 codes (Supplementary Table 1).5-7 For cohort inclusion, patients with a diagnosis of CLD were required to have outpatient lab inputs for calculating a baseline Fibrosis-4 Index (FIB-4) score. The FIB-4 calculation includes values for aspartate aminotransferase (AST), alanine aminotransferase (ALT), platelet count, and the patient's age (FIB-4=[Age×AST]/[Platelets×√ALT]).8 Scores were calculated so long as aminotransferase levels were <350 IU/L, to ensure that FIB-4 values were not calculated for acute or extreme elevations in AST and ALT values. Platelet counts within 6 months of the aminotransferase results were required for FIB-4 calculation. The date of the first set of qualifying aminotransferases for FIB-4 calculation after CLD diagnosis was used as baseline for included patients. After the index FIB-4, subsequent FIB-4 inputs were captured to calculate FIB-4 longitudinally throughout follow-up. Longitudinal FIB-4 scores were required to have unique inputs. FIB-4 scores were then categorized by advanced fibrosis risk: low (FIB-4 < 1.3), indeterminate (1.3 ≤ FIB-4 < 3.25), and high (FIB-4 ≥3.25).9,10 Patients with a diagnosis of cirrhosis, complications of cirrhosis (i.e. ascites, portal hypertension, hepatic encephalopathy, hepatorenal syndrome, or varices), HCC, or prior liver transplant (all identified by a composite of ICD-9/10 codes, Supplementary Table 1) preceding the index FIB-4 were excluded from the study.5-7
Primary Exposure
Prescription of a statin medication was the primary exposure for the study. All statin prescriptions transmitted to outpatient pharmacies were identified in the EHR during follow-up and in the 1 year preceding the index FIB-4. Patients were excluded if their first statin prescription was before the index FIB-4, to ensure a new-user design for the analysis.11 Statin prescriptions were primarily treated as a dichotomous covariate (yes/no). Statin exposure was also categorized according to maximum intensity (low/moderate/high) of the statin prescribed during the study period, as defined by the American College of Cardiology and the American Heart Association (Supplementary Table 2) to explore the association between statin prescription intensity and development of a severe liver disease (SLD) outcome (Analysis 1) and FIB-4 fibrosis risk progression during the study period (Analysis 2).12
Outcomes
Our primary outcome of interest in Analysis 1 was the time to a severe liver disease event, including cirrhosis, complications of cirrhosis, HCC, or OLT identified by a composite of ICD-9/10 previously validated for detecting cirrhosis in administrative health records (Supplementary Table 1).5-7 Complications of cirrhosis included ICD-9/10 codes for varices, hepatic encephalopathy, ascites, portal hypertension, and hepatorenal syndrome. The composite outcome was used to capture all points along the clinical continuum of severe liver disease to improve data quality and avoid making multiple comparisons that may contribute to increasing type I error.
In Analysis 2, the outcome of interest was the time to a high-risk FIB-4 score (FIB-4 ≥3.25). In this analysis, only patients in the cohort with low- or indeterminate-risk FIB-4 scores (FIB-4 <3.25) at baseline were included. Patients were followed from the index FIB-4 until a FIB-4 indicating high risk of advanced fibrosis (FIB-4 > 3.25) was obtained, or until the end of the study period.
Demographics and Covariates
Demographic variables extracted from the EHR and used in the analyses included age (continuous), sex (female/male), race (Black/non-Black), marital status (married/unmarried), and smoking status (yes/no). The race variable was dichotomized due to the low proportion of patients who identified as non-black and non-white (<3%). The FIB-4 score at baseline was treated as a continuous variable and a categorical variable (low/indeterminate/high) based on advanced fibrosis risk, as noted previously. In the full cohort, FIB-4 scores were not normally distributed, thus, FIB-4 medians and interquartile ranges (IQR) were calculated.
Comorbidity variables included body mass index (BMI, kg/m2), hypertension, diabetes mellitus, CVD (composite of coronary artery disease, peripheral artery disease, and cerebral artery disease), and chronic kidney disease (CKD). Comorbidities were determined by the presence of diagnostic codes during the study period. Diagnostic codes were identified using validated Elixhauser coding algorithms.13,14
Data Sources
All data elements for the study came from the EHR, EPIC Clarity Database (EPIC, WI), and the Enterprise Data Warehouse at the Medical University of South Carolina.
Statistical Analysis
Univariate analyses were performed to describe cohort characteristics overall and by statin prescription status. Continuous variables were compared using two sample t tests for normally distributed covariates (age, BMI, and FIB-4 in Analysis 2) and Mann Whitney U tests were used to compare FIB-4 scores in Analysis 1 (non-normal distribution). Categorical variables were compared using X2 tests. Statin prescription, statin prescription intensity, demographic, and clinical variables were compared by the occurrence of severe liver disease.
We developed unadjusted and adjusted Cox regression models for the outcome of time to a severe liver disease diagnosis during the study period. Statin prescription during follow-up was treated as a dichotomous covariate (yes/no). We then performed stepwise adjustment: first by baseline FIB-4 fibrosis risk (continuous variable), then by demographics, and finally fully adjusted by all comorbidity variables. We then developed similar unadjusted and adjusted Cox regression models with statin intensity (none/low/moderate/high) as the primary exposure variable. We tested the proportional hazards assumption by testing the interaction term between the statin prescription covariate and a function of time.
In our second analysis, we created a subset of the cohort with low- or indeterminate-risk FIB-4 scores at baseline (FIB-4 <3.25), excluding patients with a high-risk index FIB-4 (FIB-4 ≥3.25). We performed univariate analyses to describe the sub-cohort by statin prescription status and by progression to a high-risk FIB-4. We developed unadjusted and adjusted regression models to evaluate the association between statin prescription and time to a high-risk FIB-4. We then performed stepwise adjustment: first by baseline FIB-4 fibrosis risk (continuous), then by demographic variables, and finally fully adjusted by comorbidity variables. We also tested the proportional hazards assumption for this FIB-4 outcome by testing the interaction term between the statin prescription covariate and a function of time. Because the proportional hazards assumption was not met, we used Weibull regression models for the analysis plan focused on the FIB-4 ≥3.25 outcome. Weibull regression models can accommodate a non-constant hazard that changes over time. We then developed unadjusted and adjusted regression models with statin intensity (none/low/moderate/high) for each outcome (Cox for severe liver disease and Weibull for FIB-4 ≥3.25).
We also performed a sensitivity analysis evaluating the association of a statin prescription with the time to a severe liver disease diagnosis outcome in this cohort of patients with low- and indeterminate-risk FIB-4 scores (FIB-4 < 3.25) at baseline using unadjusted and adjusted Cox regression models. All models were tested for multicollinearity using variance inflation factor. We tested the assumption of the proportionality of hazards in all models using Kaplan Meier curves. SAS 9.4 (Cary, NC) was used for all statistical analyses.
Results
The overall cohort included 2,872 patients with CLD, no preceding severe liver disease outcome, and qualifying inputs for a baseline FIB-4 calculation (Figure 1). Included patients had a mean age of 49 years; 53% were female and 46% identified as Black. The most common CLD was MASLD (39%) followed by alcohol liver disease (27%), viral hepatitis (13%) and other etiologies. The cohort had a median index FIB-4 of 1.2 (IQR: 0.7-1.9), and 11% (325) of patients had a high-risk FIB-4 (≥3.25) at baseline (Table 1). Patients were followed for a mean 5.7 (±2.6) years (Figure 2). Of the overall cohort, 884 (31%) patients received a statin prescription and 378 (13%) experienced a severe liver disease outcome (Supplementary Table 3). A lower proportion of patients experiencing a severe liver outcome were prescribed a statin (22%) compared to patients with no severe liver disease (32%, p<0.001). Predictors of a severe liver disease event included: having a higher index FIB-4, older age, Male gender, smoking status, and the presence of hypertension, diabetes mellitus, cardiovascular disease, and chronic kidney disease (Supplementary Table 3).
Figure 1.
Development of the retrospective cohort of primary care patients with chronic liver disease (CLD) for Analysis 1, and an index FIB-4 < 3.25 for Analysis 2.
Table 1.
Characteristics of the cohort of patients with chronic liver disease and no statin prescriptions at baseline overall and by statin prescription during follow-up.
| Overall | Statin Rx | |||||
|---|---|---|---|---|---|---|
| Characteristics | Yes | No | p-value | |||
| n=2,872 | n=884 | n=1,988 | ||||
| Fibrosis risk at baseline | ||||||
| FIB-4 score, median (IQR) | 1.2 (0.7, 1.9) | 1.2 (0.8, 1.7) | 1.1 (0.7, 2.1) | 0.7438* | ||
| FIB-4 ≥ 3.25, % (n) | 11.3% (325) | 3.3% (29) | 14.9% (296) | <0.001† | ||
| FIB-4 < 3.25, % (n) | 88.7% (2547) | 96.7% (855) | 85.1% (1692) | <0.001† | ||
| Demographics, % (n) | ||||||
| Age, mean (SD) | 49.1 (14.9) | 54.6 (11.5) | 46.6 (15.5) | <0.001** | ||
| Gender | 0.254† | |||||
| Female | 53.1% (1524) | 51.5% (455) | 53.8% (1069) | |||
| Male | 46.9% (1348) | 48.5% (429) | 46.2% (919) | |||
| Race | 0.026† | |||||
| Black | 46.4% (1334) | 49.5% (438) | 45.1% (896) | |||
| Non-Black | 53.6% (1538) | 50.5% (446) | 54.9% (1092) | |||
| Unmarried | 64% (1838) | 58.7% (519) | 66.3% (1319) | <0.001† | ||
| Smoking hx | 25.3% (728) | 22.5% (199) | 26.6% (529) | 0.020† | ||
| BMI, mean (SD) | 29.4 (8.3) | 31.1 (7.6) | 28.6 (8.5) | <0.001** | ||
| Chronic Liver Disease Etiology, % (n) | <0.001† | |||||
| Viral hepatitis | 13.4% (385) | 13.3% (118) | 13.4% (267) | |||
| Alcohol | 27.3% (784) | 26% (230) | 27.9% (554) | |||
| MASLD / NAFLD | 38.5% (1105) | 48.1% (425) | 34.2% (680) | |||
| Other | 9.1% (262) | 3.3% (29) | 11.7% (233) | |||
| Mixed | 11.7% (336) | 9.3% (82) | 12.8% (254) | |||
| Comorbidities, % (n) | ||||||
| Hypertension | 72.2% (2073) | 91.4% (808) | 63.6% (1265) | <0.001† | ||
| Diabetes mellitus | 32.7% (939) | 52.4% (463) | 23.9% (476) | <0.001† | ||
| CVD | 35.4% (1018) | 47.7% (422) | 30% (596) | <0.001† | ||
| CKD | 20.2% (580) | 28.2% (249) | 16.6% (331) | <0.001† | ||
| Outcome | ||||||
| Severe liver disease | 13.2% (378) | 9.5% (84) | 14.8% (294) | <0.001† | ||
Mann Whitney U test.
Two sample t test.
Chi square test. BMI=body mass index, kg/m2. CKD=chronic kidney disease. CVD=cardiovascular disease. FIB-4=Fibrosis-4 Index. Hx=history. Rx=prescription. SD=standard deviation.
Figure 2.
Kaplan Meier survival curves for the time (in years) to the outcomes of: (A) time to a severe liver disease outcome; and (B) time to a Fibrosis-4 Index (FIB-4) score at high-risk for advanced fibrosis (FIB-4 ≥ 3.25) by statin prescription status.
In the unadjusted Cox regression model, statin prescription was associated with a lower risk of a severe liver disease outcome (HR 0.59; 95%CI 0.46 – 0.75) compared to patients without a statin prescription (Table 2). The proportional hazards assumption was verified by finding there was no significant time-dependent interaction of the statin covariate with follow-up time (p=0.107). After adjusting for baseline FIB-4, comorbidities, and demographics, statin prescription remained associated with a lower hazard of developing the composite severe liver disease outcome (adjusted HR 0.51; 95% CI 0.39-0.66). The unadjusted Cox regression model with statin intensity as the primary predictor variable demonstrated a lower risk of having a severe liver disease outcome for moderate (HR 0.41; 95%CI 0.26-0.62) and high intensity statins (HR 0.70; 95%CI 0.52-0.93; Supplementary Table 4). In the fully adjusted Cox regression model, moderate (HR 0.38; 95%CI 0.24-0.59) and high intensity statins (HR 0.58; 95%CI 0.43-0.79) were also associated with a lower hazard of experiencing a severe liver disease outcome. Low intensity statins were not significantly associated with the time to severe liver disease events in the models.
Table 2.
Estimated hazard ratios and 95% confidence intervals for the unadjusted and adjusted Cox regression models for the association of statin prescription with the time to a severe liver disease outcome.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| Unadjusted | Adjusted for index FIB-4 |
Adjusted for comorbidities |
Adjusted for demographics |
|
| Variables | HR 95% CI |
HR 95% CI |
HR 95% CI |
HR 95% CI |
| Statin Rx | 0.59 0.46-0.75 |
0.64 0.50-0.82 |
0.70 0.54-0.89 |
0.51 0.39-0.66 |
| Index FIB-4 | 1.07 1.06-1.08 |
1.07 1.06-1.08 |
1.08 1.07-1.09 |
|
| Male | 1.24 1.01-1.54 |
1.11 0.90-1.37 |
||
| Black | 0.70 0.56-0.87 |
0.50 0.40-0.62 |
||
| Unmarried | 1.53 1.20-1.95 |
1.52 1.19-1.94 |
||
| Smoking history | 1.20 0.95-1.51 |
1.39 1.10-1.75 |
||
| BMI | 0.99 0.97-1.00 |
0.98 0.96-0.99 |
||
| Hypertension | 1.33 1.01-1.75 |
|||
| Diabetes | 1.86 1.48-2.33 |
|||
| CVD | 1.04 0.84-1.30 |
|||
| CKD | 2.41 1.90-3.06 |
BMI=body mass index. CKD=chronic kidney disease. CVD=cardiovascular disease. FIB-4=Fibrosis-4 Index. Rx=prescription. Model 1 is the unadjusted Cox regression model. Model 2 is Model 1 with adjustment for index FIB-4 score as a continuous variable. Model 3 is Model 2 with adjustment for demographic and BMI variables. Model 4 is Model 3 with additional adjustment for comorbidity variables.
In the second analysis 2,547 patients with an index FIB-4 <3.25 were followed for a mean 5.4 (±2.6) years until the occurrence of a FIB-4 at high-risk for advanced fibrosis (FIB-4 ≥3.25), or the end of the study period (Figure 2). During follow-up, 855 (34%) patients received a statin prescription. Like the prior analysis, patients with a statin prescription were older, had a higher mean BMI, and a higher proportion had hypertension, diabetes mellitus, CVD, and CKD compared to patients without a statin prescribed (Table 3). During follow-up 600 (24%) patients progressed to a high-risk for advanced fibrosis. A significantly lower proportion of patients with a high-risk FIB-4 outcome had a statin prescribed (27%) compared to patients without progression to a high-risk FIB-4 (36%, p<0.001, Supplementary Table 5).
Table 3.
Characteristics of the cohort of patients with chronic liver disease, an index Fibrosis-4 Index (FIB-4) score < 3.25, and no statin prescriptions at baseline overall and by statin prescription during follow-up.
| Overall | Statin Rx | |||||
|---|---|---|---|---|---|---|
| Characteristics | Yes | No | p-value | |||
| n=2,547 | n=855 | n=1,692 | ||||
| Demographics, % (n) | ||||||
| Age, mean (SD) | 49.1 (14.9) | 54.3 (11.5) | 44.9 (15.4) | <0.001* | ||
| Gender | 0.046† | |||||
| Female | 55% (1402) | 52.3% (447) | 56.4% (955) | |||
| Male | 45% (1145) | 47.7% (408) | 43.6% (737) | |||
| Race | 0.047† | |||||
| Black | 46% (1172) | 48.8% (417) | 44.6% (755) | |||
| Non-Black | 54% (1375) | 51.2% (438) | 55.4% (937) | |||
| Unmarried | 63.3% (1611) | 58.2% (498) | 65.8% (1113) | <0.001† | ||
| Smoking history | 23.7% (603) | 21.9% (187) | 24.6% (416) | 0.128† | ||
| Chronic Liver Disease Etiology, % (n) | <0.001† | |||||
| Viral hepatitis | 12.8% (327) | 13.5% (115) | 12.5% (212) | |||
| Alcohol | 26.7% (679) | 25.4% (217) | 27.3% (462) | |||
| MASLD / NAFLD | 41.6% (1060) | 49% (419) | 37.9% (641) | |||
| Other | 9.8% (249) | 3.4% (29) | 13% (220) | |||
| Mixed | 9.1% (232) | 8.8% (75) | 9.3% (157) | |||
| Fibrosis risk at baseline | ||||||
| FIB-4 score, median (IQR) | 1.1 (0.7, 1.6) | 1.2 (0.8, 1.6) | 1 (0.6, 1.5) | <0.001* | ||
| Comorbidities, % (n) | ||||||
| BMI, mean (SD) | 29.4 (8.3) | 31.3 (7.6) | 29 (8.7) | <0.001* | ||
| Hypertension | 71.7% (1825) | 91.2% (780) | 61.8% (1045) | <0.001† | ||
| Diabetes mellitus | 32.8% (836) | 52.4% (448) | 22.9% (388) | <0.001† | ||
| CVD | 35.2% (897) | 46.9% (401) | 29.3% (496) | <0.001† | ||
| CKD | 19.7% (501) | 28% (239) | 15.5% (262) | <0.001† | ||
Two sample t test.
Chi square test. BMI=body mass index (kg/m2). CKD=chronic kidney disease. CVD=cardiovascular disease. FIB-4=Fibrosis-4 Index. Rx=prescription. SD=standard deviation.
In this second analysis, there was a significant time-dependent interaction of the statin covariate with follow-up time (p <0.001), thus the proportional hazards assumption was not met. In the unadjusted Weibull regression model evaluating the association of a statin prescription with time to a high-risk FIB-4, a statin prescription was associated with a lower risk of progressing to a FIB-4 ≥3.25 (HR 0.63; 95%CI 0.55-0.74, Table 4). Statin prescription remained associated with a lower hazard of progressing to a high-risk FIB-4 after adjusting for the index FIB-4 fibrosis risk (HR 0.53; 95%CI 0.47-0.61) and patient demographics (HR 0.54; 95%CI 0.47-0.63). In the fully adjusted model, that also adjusted for comorbidities, a prescription for a statin was still associated with a lower hazard of progressing to a high-risk for advanced fibrosis (HR 0.36; 95%CI 0.32-0.42). For the unadjusted Weibull regression model with statin intensity as the primary predictor, moderate (HR 0.54; 95%CI 0.44-0.71) and high intensity statins (HR 0.64; 95%CI 0.54-0.79) were associated with a lower risk of progressing to a high-risk FIB-4 score, while low intensity statin prescriptions were not associated with the outcome (Supplementary Table 6). In the fully adjusted Weibull regression model, moderate (HR 0.36; 95%CI 0.29-0.46) and high intensity statins (HR 0.33; 95%CI 0.29-0.40) are all associated with a lower risk of progressing to a FIB-4 at high-risk for advanced fibrosis.
Table 4.
Estimated hazard ratios and 95% confidence intervals for the unadjusted and adjusted Weibull regression models for the association of statin prescription with the outcome of time to a Fibrosis-4 Index score at high-risk for advanced fibrosis (FIB-4≥ 3.25).
| Model 1 | Model 2 | Model 3 | Model 4 | ||
|---|---|---|---|---|---|
| Unadjusted | Adjusted for index FIB-4 |
Adjusted for demographics |
Adjusted for comorbidities |
||
| Variables | HR 95% CI |
HR 95% CI |
HR 95% CI |
HR 95% CI |
|
| Statin Rx | 0.63 0.55 – 0.74 |
0.53 0.47 – 0.61 |
0.54 0.47 – 0.63 |
0.36 0.32 – 0.42 |
|
| Index FIB-4 score | 3.00 2.48 – 3.71 |
2.84 2.36 – 3.50 |
2.73 2.29 – 3.34 |
||
| Male | 1.06 0.90 – 1.28 |
1.05 0.89 – 1.26 |
|||
| Black | 1.59 1.31 – 1.99 |
1.12 0.94 – 1.37 |
|||
| Unmarried | 1.11 0.93 – 1.36 |
1.07 0.90 – 1.32 |
|||
| Smoking history | 0.90 0.76 – 1.09 |
1.05 0.87 – 1.29 |
|||
| BMI | 0.98 0.97 – 0.99 |
0.98 0.97 – 0.99 |
|||
| Hypertension | 1.55 1.20 – 2.08 |
||||
| Diabetes | 1.37 1.14 – 1.70 |
||||
| CVD | 2.02 1.63 – 2.58 |
||||
| CKD | 1.91 1.53 – 2.46 |
||||
BMI=body mass index. CKD=chronic kidney disease. CVD=cardiovascular disease. FIB-4=Fibrosis-4 Index. Rx=prescription. Model 1 is the unadjusted Weibull regression model. Model 2 is Model 1 with adjustment for index FIB-4 score as a continuous variable. Model 3 is Model 2 with adjustment for demographic variables and BMI. Model 4 is Model 3 with additional adjustment for comorbidities.
In the sensitivity analysis evaluating the association of statin prescription with the time to a severe liver disease diagnosis (n=216, 8.4%) in the sample of patients with low- and indeterminate-risk FIB-4 scores (FIB-4 < 3.25) at baseline (n=2,547), statin prescription had an estimated hazard ratio of 0.74 (95%CI 0.55-1.00) for the time to event outcome.
Discussion
Our study aimed to determine the association between statin prescription and CLD outcomes in a primary care setting before the development of cirrhosis. In this large retrospective cohort study, statin prescription was associated with a significantly reduced risk of developing advanced fibrosis assessed by FIB-4 or a severe liver disease outcome, including development of cirrhosis (or one of its complications), HCC, or need for OLT. It is noteworthy that patients in the statin cohort had a lower likelihood of the study outcomes, even though they were on average older and had a higher prevalence of metabolic comorbidities, including diabetes. Within the statin cohort, higher intensity statin doses were more protective than the low intensity dosing. However, a true dose effect was not evident in this study since there was not an incremental reduction in the hazard of developing the study outcomes between medium and high intensity cohorts. It is not clear if this was due to a lack of power with fewer subjects in the high intensity group or that there is a dosing plateau in the protective effect of statins in liver disease outcomes.
Our results add to the growing body of evidence that in patients with CLD, statins may be associated with a reduced risk of liver fibrosis progression and developing a SLD outcome. Initial studies focused on patients with specific liver disease etiologies or on those with established cirrhosis.15-17 For example, Yang et al. performed a large cohort study in Taiwan that examined 226,856 patients with HCV and demonstrated a dose-response relationship between statin use and cirrhosis risk.18 Similarly, Oliver et al. showed a decreased hazard (0.68, 95% CI of 0.47-0.98) of developing cirrhosis in patients prescribed a statin in a select patient population with HIV/HCV coinfection.19 A potential benefit with statin use in cirrhosis patients was also demonstrated in a large retrospective cohort study of Veterans Affairs patients.20
More recently, two studies demonstrated that the beneficial effects of statins in CLD patients may exist, regardless of liver disease etiology or fibrosis severity. Similar to our study, Sharma et al. published a large nationwide cohort study in Sweden that also showed a decreased risk of developing a SLD outcome (HR 0.60, 95% CI 0.8-0.74).[21] Our analysis of SLD outcomes was remarkably similar (adjusted HR 0.51; 95% CI 0.39-0.66) despite the obvious differences in demographic composition as well as in statin prescription prevalence. More recently, Choi et al. also investigated the association between statin use and CLD outcomes in the Mass General Brigham health care system in Boston, MA.3 Our study did have several noteworthy differences. First, our study population had greater racial diversity, with a high prevalence of Black patients (46.4%). In our study, Black patients had a greater risk of progressing to high-risk FIB-4 after adjusting for multiple confounding variables in the fully adjusted model. Conversely, Black patients were less likely to experience a SLD (HR 0.50; 95% CI 0.40-0.62). The reason for this discrepancy between the risk of fibrosis progression and subsequent development of a SLD outcome is not clear and would require further investigation. Also, compared to the Choi study, ours had more robust BMI reporting that allowed for us to better correct for its impact on liver disease progression. We also did not exclude patients with a low FIB-4 score (< 1.3), that positioned us to capture progression from lower-risk CLD patients. Finally, by using the cutoff of ≥3.25 for defining high-risk for advanced fibrosis, our study may be more specific for capturing advanced fibrosis, especially in older patients as age is a contributing factor to FIB-4 scores.
Our study has some limitations. First, our study is a single center one. However, as we already noted, the population was not only large, but it is also uniquely diverse compared to prior studies. We also used a surrogate measure of fibrosis (FIB-4) to assess disease progression. Although liver biopsy is the gold standard for determining liver fibrosis, FIB-4 is a well validated and recommended non-invasive liver fibrosis scoring tool.23 Despite its high specificity, using FIB-4 exclusively may lead to misclassification, especially when compared to more direct fibrosis measurements like elastography.26 Additionally, since patient age contributes to the FIB-4 score, longitudinal increases may partially reflect patient aging rather than progression of hepatic fibrosis. Our models adjusted for index FIB-4 as a continuous covariate, of which age is a contributor. Given that our exposed group was on average older than the unexposed group, we would expect a faster progression to high-risk FIB-4 in the exposed group if FIB-4 increases were primarily being driven by age. Instead, our exposed group was found to have a lower hazard of progression to high-risk FIB-4, providing some reassurance that our findings are not primarily driven by age-related confounding.
Another limitation was our inability to account for possible confounding factors regarding statin prescription, including healthy user bias and medication adherence. Although statin recipients were generally older and had higher burden of comorbidities, these factors likely also lead to increased healthcare contact and therefore may be more likely to exhibit healthier behaviors or have well managed comorbidities compared to the unexposed group. This in turn may bias our results toward a stronger protective association. The data structure was also not able to account for dynamic changes in the specific statin prescribed and the dosing during follow-up. The analysis of the maximum potency of statin prescribed provides an evaluation of the most intense statin prescription, but future work that includes pharmacy prescription fill data may address this gap. We were also unable to account for the factors that may influence the natural history of chronic liver disease and fibrosis progression, such as antiviral therapy, and alcohol abstinence.27-30
In conclusion, our study adds to the body of evidence that statins are associated with a reduced risk of fibrosis progression and developing a SLD outcome in patients with CLD, regardless of fibrosis severity. Additional prospective randomized controlled trials and studies with direct measurements of fibrosis are needed to help confirm our findings.
Supplementary Material
Funding Support:
National Institute of Diabetes and Digestive and Kidney Diseases (NIH/NIDDK K23DK118200 PI: Schreiner; R03DK129558 PI: Schreiner; R01DK139087). This project was also supported by the South Carolina Clinical & Translational Research Institute with an academic home at the Medical University of South Carolina CTSA National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH) under UL1 TR001450.
Abbreviations:
- ALT
Alanine aminotransferase
- AST
Aspartate aminotransferase
- CLD
Chronic liver disease
- CI
Confidence interval
- EHR
Electronic health record
- FIB-4
Fibrosis-4 Index
- GGPP
Geranylgeranyl pyrophosphate
- HCC
Hepatocellular carcinoma
- HCV
Hepatitis C Virus
- HIV
Human Immunodeficiency Virus
- HSC
Hepatic stellate cells
- HVR
Hepatic vascular resistance
- ICD
International Classification of Diseases
- KLF2
Kruppel-Like Factor 2
- MASLD
Metabolic Dysfunction-Associated Steatotic Liver Disease
- NO
Nitric oxide
- NOS
Nitric oxide synthase
- OLT
Orthotopic Liver Transplant
- PLT
Platelet count
- Rho-A
Rho-A
- SLD
Severe Liver Disease
- 3-hydroxyl-3-methylglutaryl coenzyme A reductase inhibitors
Statins
Footnotes
Conflicts of Interest: Dr. Schreiner reports financial support was provided by National Institute of Diabetes and Digestive and Kidney Diseases. Dr. Schreiner also reports a consulting relationship with Novo Nordisk. All other authors report no conflicts of interest with this work.
Declaration of Generative AI Use: No generative AI was used in the writing of this manuscript.
Ethical Review: This study was approved by the Institutional Review Board (IRB) at the Medical University of South Carolina (MUSC), and we certify that the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards (IRB #Pro00056541). The MUSC IRB approved this study with a waiver of consent, due to the retrospective nature of the data.
Data Availability Statement:
The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request. All data requests require review and approval by the Medical University of South Carolina (MUSC) Institutional Review Board (IRB). Data are located in controlled access data storage at MUSC.
References
- 1.Moon AM, Singal AG, Tapper EB. Contemporary Epidemiology of Chronic Liver Disease and Cirrhosis. Clin Gastroenterol Hepatol. 2020;18(12):2650–2666. doi: 10.1016/j.cgh.2019.07.060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Younossi Z, Anstee QM, Marietti M, et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol. 2018;15(1):11–20. doi: 10.1038/nrgastro.2017.109 [DOI] [PubMed] [Google Scholar]
- 3.Kreidieh M, Hamadi R, Alsheikh M, Al Moussawi H, Deeb L. Statin Use in Patients With Chronic Liver Disease and Cirrhosis: Current Evidence and Future Directions. Gastroenterol Res. 2022;15(1):1–12. doi: 10.14740/gr1498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Statin Use and Risk of Hepatocellular Carcinoma and Liver Fibrosis in Chronic Liver Disease ∣ Hepatobiliary Disease ∣ JAMA Internal Medicine ∣ JAMA Network. Accessed May 9, 2025. https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2831392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Simon TG, King LY, Zheng H, Chung RT. Statin use is associated with a reduced risk of fibrosis progression in chronic hepatitis C. J Hepatol. 2015;62(1):18–23. doi: 10.1016/j.jhep.2014.08.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hagström H, Adams LA, Allen AM, et al. Administrative Coding in Electronic Health Care Record-Based Research of NAFLD: An Expert Panel Consensus Statement. Hepatol Baltim Md. 2021;74(1):474–482. doi: 10.1002/hep.31726 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mapakshi S, Kramer JR, Richardson P, El-Serag HB, Kanwal F. Positive Predictive Value of International Classification of Diseases, 10th Revision, Codes for Cirrhosis and Its Related Complications. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc. 2018;16(10):1677–1678. doi: 10.1016/j.cgh.2018.01.042 [DOI] [PubMed] [Google Scholar]
- 8.Nehra MS, Ma Y, Clark C, Amarasingham R, Rockey DC, Singal AG. Use of administrative claims data for identifying patients with cirrhosis. J Clin Gastroenterol. 2013;47(5):e50–54. doi: 10.1097/MCG.0b013e3182688d2f [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vallet-Pichard A, Mallet V, Nalpas B, et al. FIB-4: an inexpensive and accurate marker of fibrosis in HCV infection. comparison with liver biopsy and fibrotest. Hepatol Baltim Md. 2007;46(1):32–36. doi: 10.1002/hep.21669 [DOI] [PubMed] [Google Scholar]
- 10.Shah AG, Lydecker A, Murray K, et al. Comparison of noninvasive markers of fibrosis in patients with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc. 2009;7(10):1104–1112. doi: 10.1016/j.cgh.2009.05.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kanwal F, Shubrook JH, Adams LA, et al. Clinical Care Pathway for the Risk Stratification and Management of Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology. 2021;161(5):1657–1669. doi: 10.1053/j.gastro.2021.07.049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Yoshida K, Solomon DH, Kim SC. Active-comparator design and new-user design in observational studies. Nat Rev Rheumatol. 2015;11(7):437–441. doi: 10.1038/nrrheum.2015.30 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/ APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: Executive Summary. Circulation. 2019;139(25):e1082–e1143. doi: 10.1161/CIR.0000000000000625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Quan H, Li B, Saunders LD, et al. Assessing validity of ICD-9-CM and ICD-10 administrative data in recording clinical conditions in a unique dually coded database. Health Serv Res. 2008;43(4):1424–1441. doi: 10.1111/j.1475-6773.2007.00822.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–1139. doi: 10.1097/01.mlr.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
- 16.Kumar S, Grace ND, Qamar AA. Statin use in patients with cirrhosis: a retrospective cohort study. Dig Dis Sci. 2014;59(8):1958–1965. doi: 10.1007/s10620-014-3179-2 [DOI] [PubMed] [Google Scholar]
- 17.Pollo-Flores P, Soldan M, Santos UC, et al. Three months of simvastatin therapy vs. placebo for severe portal hypertension in cirrhosis: A randomized controlled trial. Dig Liver Dis. 2015;47(11):957–963. doi: 10.1016/j.dld.2015.07.156 [DOI] [PubMed] [Google Scholar]
- 18.Abraldes JG, Albillos A, Bañares R, et al. Simvastatin Lowers Portal Pressure in Patients With Cirrhosis and Portal Hypertension: A Randomized Controlled Trial. Gastroenterology. 2009;136(5):1651–1658. doi: 10.1053/j.gastro.2009.01.043 [DOI] [PubMed] [Google Scholar]
- 19.Yang YH, Chen WC, Tsan YT, et al. Statin use and the risk of cirrhosis development in patients with hepatitis C virus infection. J Hepatol. 2015;63(5):1111–1117. doi: 10.1016/j.jhep.2015.07.006 [DOI] [PubMed] [Google Scholar]
- 20.Oliver NT, Hartman CM, Kramer JR, Chiao EY. Statin drugs decrease progression to cirrhosis in HIV/hepatitis C virus coinfected individuals. AIDS. 2016;30(16):2469. doi: 10.1097/QAD.0000000000001219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mahmud N, Chapin S, Goldberg DS, Reddy KR, Taddei TH, Kaplan DE. Statin exposure is associated with reduced development of acute-on-chronic liver failure in a Veterans Affairs cohort. J Hepatol. 2022;76(5):1100–1108. doi: 10.1016/j.jhep.2021.12.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sharma R, Simon TG, Hagström H, et al. Statins Are Associated With a Decreased Risk of Severe Liver Disease in Individuals With Noncirrhotic Chronic Liver Disease. Clin Gastroenterol Hepatol. 2024;22(4):749–759.e19. doi: 10.1016/j.cgh.2023.04.017 [DOI] [PubMed] [Google Scholar]
- 23.Comparison of laboratory tests, ultrasound, or magnetic reso… : Hepatology. Accessed May 4, 2024. https://journals.lww.com/hep/abstract/2017/11000/comparison_of_laboratory_tests,_ultrasound,_or.18.aspx
- 24.Sterling RK, Patel K, Duarte-Rojo A, et al. AASLD Practice Guideline on blood-based noninvasive liver disease assessment of hepatic fibrosis and steatosis. Hepatol Baltim Md. Published online March 15, 2024. doi: 10.1097/HEP.0000000000000845 [DOI] [PubMed] [Google Scholar]
- 25.Siddiqui MS, Yamada G, Vuppalanchi R, et al. Diagnostic Accuracy of Noninvasive Fibrosis Models to Detect Change in Fibrosis Stage. Clin Gastroenterol Hepatol. 2019;17(9):1877–1885.e5. doi: 10.1016/j.cgh.2018.12.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Chang M, Chang D, Kodali S, et al. Degree of Discordance Between FIB-4 and Transient Elastography: An Application of Current Guidelines on General Population Cohort. Clin Gastroenterol Hepatol. 2024;22(7):1453–1461.e2. doi: 10.1016/j.cgh.2024.02.008 [DOI] [PubMed] [Google Scholar]
- 27.Reversal of liver fibrosis: From fiction to reality - ScienceDirect. Accessed October 5, 2024. https://www.sciencedirect.com/science/article/pii/S1521691817300288?via%3Dihub#bib9
- 28.Weight Loss Through Lifestyle Modification Significantly Reduces Features of Nonalcoholic Steatohepatitis - ScienceDirect. Accessed October 6, 2024. https://www.sciencedirect.com/science/article/pii/S0016508515004965?casa_token=VcDZZe_yzygAAAAA:uJGJOiHrf4KQvw83CF5_CrCC-ldpnub7h5Wya-yHiOIOHqbrBJK3od4lHMtGsQ7p4seFGY0bzg [DOI] [PubMed]
- 29.Kweon YO, Goodman ZD, Dienstag JL, et al. Decreasing fibrogenesis: an immunohistochemical study of paired liver biopsies following lamivudine therapy for chronic hepatitis B. J Hepatol. 2001;35(6):749–755. doi: 10.1016/S0168-8278(01)00218-5 [DOI] [PubMed] [Google Scholar]
- 30.Poynard T, McHutchison J, Manns M, et al. Impact of pegylated interferon alfa-2b and ribavirin on liver fibrosis in patients with chronic hepatitis C. Gastroenterology. 2002;122(5):1303–1313. doi: 10.1053/gast.2002.33023 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request. All data requests require review and approval by the Medical University of South Carolina (MUSC) Institutional Review Board (IRB). Data are located in controlled access data storage at MUSC.