Association of Liver Fibrosis Markers with Mortality Outcomes in Patients with Chronic Kidney Disease and Coronary Artery Disease: Insights from the NHANES 1999–2018 Data

Chronic kidney disease (CKD) and coronary artery disease (CAD) are substantial worldwide health issues that frequently coexist, profoundly impacting morbidity and mortality rates. When CAD coexists with CKD, patients typically experience a more aggressive disease trajectory and a heightened risk of cardiovascular events [1, 2]. This underscores the urgent need for effective risk stratification tools to enhance clinical decision-making and optimize patient outcomes. The combination of CAD and CKD, commonly known as cardiorenal syndrome, involves complex pathophysiological interactions that drive disease progression and lead to poorer outcomes [3]. Key mechanisms include inflammation, oxidative stress, fibrosis, and endothelial dysfunction, all of which can impair both cardiac and renal function [4]. Therefore, identifying reliable biomarkers that can predict adverse outcomes in CAD patients with CKD is crucial.

Liver function tests, specifically aspartate aminotransferase (AST) and alanine aminotransferase (ALT), have become a focus among the diverse biomarkers studied for evaluating disease such as type 2 diabetes, CAD, and thrombocytopenia syndrome [5, 6]. These tests serve as important indicators of systemic inflammation and metabolic disturbances [7]. The ratio of AST to ALT, an uncomplicated but insightful metric, has been associated with the severity of fibrosis in nonalcoholic fatty liver disease (NAFLD) and could potentially serve as a predictor in other chronic diseases [8]. The Nonalcoholic Fatty Liver Score (NFS), Forns index, and the Fibrosis-4 Index (FIB-4) are recognized noninvasive instruments utilized to assess the severity of NAFLD and liver fibrosis, respectively [7, 9, 10]. These indexes include elements like age, body mass index (BMI), diabetes condition, and liver enzyme measurements, thus expanding their usefulness beyond just evaluations related to the liver.

Recent studies indicate that hepatic dysfunction, as evidenced by elevated AST/ALT ratios and fibrosis scores, may reflect systemic inflammation and metabolic imbalances that contribute to adverse outcomes in patients with CAD and CKD [11]. Nevertheless, the particular function of these biomarkers in forecasting results for patients suffering from simultaneous CKD and CAD is still not thoroughly investigated. Much of the existing research has concentrated on individual diseases or their isolated components, leaving a gap in understanding their combined effects and prognostic value in this complex patient population.

Taking into account these factors, the current research intends to explore the predictive value of FIB-4, NFS, Forns index and the AST/ALT ratio in CKD and CAD patients. Our goal, through the evaluation of a substantial group from the National Health and Nutrition Examination Survey (NHANES) covering the period 1999–2018, is to determine if these biomarkers can offer a significant understanding of patient prognosis. This could potentially enable early risk categorization and specific interventions to enhance results.

The Centers for Disease Control and Prevention (CDC) continuously conducts the National Health and Nutrition Examination Survey (NHANES), a project that evaluates the health, diet, and lifestyle of individuals in the USA, and this research made use of its data [12, 13]. NHANES utilizes a multilevel likelihood sampling technique to guarantee that its results are nationally representative. From the time it was established in 1999, NHANES has gathered extensive information, such as physical check-ups and interviews, from individuals who gave their written approval. The National Center for Health Statistics Ethics Review Board has sanctioned the procedure of gathering data and releasing public documents. The NHANES database consists of five key sections: questionnaire, nutritional, physical examination, laboratory, and demographic data. DataDryad (https://doi.org/10.5061/dryad.d5h62) provides access to the datasets.

NHANES data from survey cycles spanning 1999–2018 were conducted to analysis. Our study incorporated participants who were 20 years or older at the time of registration. We utilized exclusion standards to eliminate persons with incomplete data on vital parameters, such as AST, ALT, platelet count, albumin, cholesterol, gamma glutamyl transpeptidase (GGT), age, BMI, diabetes condition, creatinine concentrations, and life expectancy. Patients with over 30% missing covariable data (including smoking status, BMI, alcohol consumption, hypertension, hyperlipidemia, diabetes, history of heart attack, and estimated glomerular filtration rate [eGFR] values.) were also excluded. The CDC’s Institutional Review Board for research involving human subjects granted permission for this study.

The primary outcome measures were all-cause and cardiovascular disease (CVD) mortality. The National Center for Health Statistics paired NHANES data with death records from the National Death Index to acquire mortality data [14]. Participants who could not be linked to death data during the follow-up period were considered alive. The evaluation was conducted on both general death rates and those related to CVDs, with the latter being categorized according to the International Classification of Diseases 10th Revision (ICD-10) codes I00-I09, I11, I13, I20-I51, and I60-I69. We collected demographic characteristics, lifestyle factors (smoking and alcohol consumption), educational attainment, race, and laboratory parameters, including glycated hemoglobin (HbA1c), liver enzymes, renal function, lipids, and electrolytes.

The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation was utilized to compute the eGFR [15], where a CKD diagnosis is determined by an eGFR value was lower than 60 mL/min per 1.73 m². The presence of CAD was confirmed through a self-declared doctor’s diagnosis of coronary heart disease, angina, or myocardial infarction. Hypertension was characterized by a systolic blood pressure exceeding 140 mm Hg, a diastolic blood pressure surpassing 90 mm Hg, a self-proclaimed diagnosis, or the utilization of drugs to lower high blood pressure. Diabetes was determined through self-reported physician diagnosis or a fasting glucose level greater than 126 mg/dL.

The Forns index was calculated using the equation [10]: Forns index = 7.811 – 3.131 × ln (platelets × 10⁹/L) + 0.781 × ln (GGT, U/L) + 3.467 × ln (age, years) – 0.541 × (cholesterol, mmol/L).

Continuous variables were presented as the median and interquartile ranges, and category variables were described as the frequencies and percentage. The baseline characteristics were compared using analysis of the Kruskal-Wallis rank sum test for continuous variables and the chi-square (χ²) test for categorical variables.

The multivariate Cox proportional hazards regression models to evaluate hazard ratios (HRs) and 95% confidence intervals (CIs), specifically designed to adjust for confounding variables. Model 1 was unadjusted, Model 2 was adjusted for age, sex, race, and education level, and Model 3 included additional adjustments for smoking status, BMI, alcohol consumption, hypertension, hyperlipidemia, diabetes, history of heart attack, and eGFR values. It is worth mentioning that since the FIB-4 and Forns index calculation incorporates the age variable, while the NFS includes age, BMI, and diabetes status, age was not included in the variable adjustments for FIB-4 and Forns index, and age, BMI, and diabetes status were not included in the variable adjustments for NFS. We explored the dose-response relationship between liver fibrosis scores (LFSs) and mortality using restricted cubic spline (RCS) analysis. Kaplan-Meier survival curves were generated to evaluate the impact of LFSs on death risk, utilizing inflection points from the RCS as cutoff points for stratifying patients into low- or high-LFS value groups.

The evaluation of each LFS’s forecasting capacity was conducted through the area under the curve (AUC) of the receiver operating characteristic (ROC). Analyses of subgroups were conducted to examine the correlation between the ratio of AST to ALT and mortality from all-causes, including cardiovascular deaths. R version 4.3.2 was utilized to perform all the analyses, considering a two-tailed p value <0.05 as statistically significant.

Initially, the research incorporated 101,316 individuals from the NHANES database, spanning the years 1999–2018 (Fig. 1). Out of the group, 1,338 adults who were 20 years or older were recognized as having both CKD and CAD. The final analysis included 1,192 participants, with 508 (42.6%) reported as deceased and 684 (57.4%) as survivors. The baseline attributes of these participants are encapsulated in Table 1. The median age of deceased patients was 78 years, while survivors had a median age of 72 years. Age, HDL, FIB-4, and NFS increased with increasing AST/ALT ratios, while BMI, waist, HbA1c, hemoglobin, and percentage of diabetes mellitus decreased with increasing AST/ALT ratios. Remarkably, there were considerable variations in the AST/ALT ratio, FIB-4 index, NFS, Forns index, age, ethnicity, gender, educational background, BMI, red blood cell count, creatinine levels, and the incidence of congestive heart failure between those who survived and those who did not (p < 0.05, online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000543500).

Our analysis revealed a linear relationship between the FIB-4 index, NFS, Forns index and AST/ALT values with the probabilities of all-cause and CVD death risk among participants with CAD and CKD, as determined by the multivariable RCS model (online suppl. Fig. 1, 2). Specifically, elevated levels of FIB-4, NFS, Forns index, and AST/ALT suggest a heightened risk of mortality. Participants were divided into two categories – low-level and high-level, using cutoff values of 1.87 for FIB-4, 0.11 for NFS, 5.90 for Forns index and 1.24 for AST/ALT according to the RCS findings.

The Kaplan-Meier curve analysis revealed an increase in both all-cause and CVD mortality rates in groups with high FIB-4, NFS, Forns index and AST/ALT levels compared to those with lower levels. This was evidenced by a marked disparity in cumulative HRs (log-rank p < 0.001, Figure 2; online suppl. Fig. 3). Moreover, the Cox regression analysis (as shown in Table 2) revealed that FIB-4, NFS, Forns index, and AST/ALT, after thorough adjustments, were positively correlated with a heightened risk of mortality from all-causes. The HR were 1.188 (95% CI: 1.108–1.274), 1.145 (95% CI: 1.069–1.227), 1.142 (1.081–1.201), and 1.316 (95% CI: 1.056–1.639), respectively. In terms of the risk of death from CVD, comparable relationships were noted, exhibiting HR of 1.133 (95% CI: 1.007–1.275), 1.155 (95% CI: 1.024–1.303), 1.208 (1.109–1.316), and 1.636 (95% CI: 1.203–2.224).

The predictive capacity of FIB-4, NFS, Forns index, and AST/ALT was evaluated through the receiver operating characteristic (ROC) analysis (online suppl. Fig. 4). The accuracy in forecasting all-cause mortality in CKD and CAD patients was similar for all four biomarkers, with AUC values of 0.686, 0.684, 0.682 and 0.687, respectively. For CVD mortality prediction, FIB-4, NFS and AST/ALT outperformed Forns index slightly, with AUC values of 0.632, 0.628, 0.636 compared to 0.626 for Forns index. Although AST/ALT showed the greatest predictive potential for both overall and CVD mortality, the discrepancies between AST/ALT and FIB-4 were not statistically meaningful. Nonetheless, a notable disparity was observed between AST/ALT and NFS in predicting all-cause mortality as well as AST/ALT and Forns index (p < 0.05, DeLong’s test, online suppl. Table 2).

To further validate the relationship between AST/ALT levels and both all-cause and CVD mortality, we performed subgroup analyses stratified by age, sex, smoking status, kidney function, and health conditions (diabetes, hypertension, hyperlipidemia, and history of heart attack). We found that eGFR levels and history of heart attack interacted with the relationship between AST/ALT and all-cause mortality. Interestingly, this correlation remained consistent among the majority of participants, excluding those who were 65 years old or younger and those suffering from serious kidney damage (eGFR ≤15 mL/min per 1.73 m²) (refer to Figure 3; online suppl. Fig. 5).

This research offers an exhaustive analysis of the correlation between markers of liver fibrosis, specifically, FIB-4, NFS, Forns index, and the AST/ALT ratio, and the survival rates in individuals suffering from CAD and CKD in the NHANES group. The results of our study suggest a positive correlation between these indicators, especially the AST/ALT ratio, and mortality due to all-causes and CVD in people suffering from CAD and CKD. Notably, the AST/ALT ratio emerged as the most straightforward and valuable biomarker for predicting mortality in this patient population.

The AST/ALT ratio, along with NFS, FIB-4, and Forns index, has gained attention for its potential to indicate liver injury, fibrosis, and metabolic disturbances – all of which are closely linked to cardiovascular and renal health [18, 19]. While primarily utilized in liver assessment, the AST/ALT ratio also holds promise for predicting adverse outcomes in non-hepatic conditions like CAD and CKD [8].

In addition, changes in AST and ALT levels are associated with mortality rates, with abnormal levels of ALT or AST correlating with future mortality in the population [20, 21]. Recent studies have found that a higher AST/ALT ratio is an independent predictor of 1-year mortality in patients with polymyositis/dermatomyositis-related interstitial lung disease [22]. Furthermore, an elevated AST/ALT ratio is linked to poor prognosis in various cancers, including renal cell carcinoma and oropharyngeal cancer [23]. Our research demonstrates that AST/ALT levels are independently associated with all-cause mortality and cardiovascular mortality in patients with CKD and CAD, consistent with previous literature. This may be attributed to the fact that AST and ALT serve as indicators of metabolism and inflammation related to oxidative stress [24‒26], both of which are risk factors for CV diseases and mortality. Consequently, these factors may mediate the occurrence of CV diseases and mortality. Additionally, elevated AST level are associated with the progression of atherosclerosis [27], suggesting that the mediating effect of CVD may increase the risk of death. However, the specific mechanisms warrant further investigation.

Likewise, NFS is broadly acknowledged for assessing the danger of NAFLD and its advancement to fibrosis. By applying a high cutoff value, the NFS can accurately diagnose the presence of advanced fibrosis, with a positive predictive value of up to 82% in the validation group [17]. This is especially pertinent considering the high occurrence of NAFLD among patients with CAD and CKD [28]. FIB-4, a noninvasive marker that assesses liver fibrosis using age, AST, ALT, and platelet count, has shown utility in predicting outcomes across various chronic liver diseases [29]. The FIB-4 index can effectively identify patients with severe fibrosis (F3-F4) and cirrhosis. A FIB-4 index greater than 3.25 has a positive predictive value of 82.1% for confirming significant fibrosis (F3-F4) and a specificity of 98.2% [30]. It may also reflect fibrosis in the cardiovascular system due to shared pathological pathways, enhancing its relevance in the context of CAD and CKD. The Forns index is a model and scoring system that integrates age, GGT, cholesterol, and platelet count to help identify patients without apparent liver fibrosis. It can accurately exclude the presence of significant fibrosis (F2 to F4), with a negative predictive value of 36% [10]. Patients with CKD and CAD are prone to metabolic disorders and hyperlipidemia, thus the predictive value of the Forns index in this population warrants further exploration.

The development of liver fibrosis is often linked to conditions like NAFLD or metabolic dysfunction-associated fatty liver disease (MAFLD) [31], which are progressively acknowledged as elements of metabolic syndrome. The overlap of risk factors associated with NAFLD/MAFLD and CAD underscores a significant link between liver health and cardiovascular outcomes [32]. Studies have demonstrated, for example, that even a slight presence of NAFLD can elevate the risk of CVD and overall death rate in patients suffering from type 2 diabetes [33]. Additionally, Korean national studies have identified a markedly elevated CVD risk in individuals with NAFLD/MAFLD [34]. Although liver biopsy is still considered the benchmark for evaluating the severity of liver fibrosis, noninvasive scoring methods such as FIB-4, NFS, Forns index, and the AST/ALT ratio are often used because of their ease of use [35, 36]. Nevertheless, there is still limited research investigating the link between liver fibrosis and negative results in CAD and CKD patients.

Our research revealed notable links between FIB-4, NFS, Forns index, and the AST/ALT ratio and both all-cause and cardiovascular death in individuals suffering from CAD and CKD. Remarkably, the forecasted values of these biological markers exhibited minimal statistical variation, implying that the AST/ALT ratio might be a more favored diagnostic method because of its easy availability and simple quantification. While a universally accepted threshold for the AST/ALT ratio has yet to be established, our study identified a significant increase in mortality risk for individuals with a ratio exceeding 1.24. Moreover, eGFR levels and the history of heart attacks emerged as critical predictors for mortality in this patient group [37, 38]. The analysis of our subgroups suggested that these elements influence the correlation between the AST/ALT ratio and total mortality in patients with CAD and CKD. Furthermore, in patients without a history of heart attack, the AST/ALT ratio may have a more pronounced effect on mortality risk. However, since statistical tests did not show significant effect modification, this conclusion should be interpreted with caution, highlighting the need for further research to confirm this observation.

The study’s strengths include its substantial sample size, comprehensive adjustments for confounding variables, and the evaluation of multiple biomarkers. The results are poised to enhance understanding of the management of CKD and CAD comorbidity, contributing to clinical practice guidelines.

Despite these strengths, our study has limitations. The retrospective design restricts our ability to infer causal relationships between the biomarkers and mortality outcomes. Therefore, the associations noted should be interpreted cautiously and validated through prospective research. Although we adjusted for numerous variables, residual confounding from unmeasured factors remains a possibility. For example, lifestyle elements like diet and physical activity, which significantly affect mortality risk, were not thoroughly assessed in the NHANES dataset. Additionally, although the NHANES cohort is nationally representative, it may not encompass all patients with CAD and CKD – particularly those with more severe disease – limiting the generalizability of our findings. Ultimately, the dominance of people with American ancestry in the NHANES group may limit the relevance of our findings to different communities. Future studies should include diverse ethnic groups to ensure broader relevance. Additionally, liver biopsy is the gold standard for assessing liver fibrosis. Due to limitations in publicly available data, we were unable to further investigate the association between biopsy-diagnosed liver fibrosis and prognosis in patients with CKD with CAD. There are also several noninvasive liver fibrosis markers, such as the Fibro index, BARD score, and proteomic and glycomic profiles, which have shown good predictive value for liver fibrosis. However, given the use of public databases and considerations of clinical practicalities, we did not conduct further analysis on these markers in this study. Instead, this research compared the relationship between four commonly used and representative liver fibrosis biomarkers and the prognosis of patients with CKD and CAD.

In summary, this study emphasizes the significance of hepatic fibrosis markers, especially the AST/ALT ratio, in evaluating survival outcomes for patients with concurrent CAD and CKD. The observed associations suggest that early identification and management of hepatic fibrosis could improve outcomes in this high-risk population. Future research should delve into the mechanisms linking hepatic fibrosis to adverse outcomes and investigate potential interventions to mitigate these risks.

We acknowledge all participants in the NHANCE research team for survey design and data collection.

A Statement of Ethics is not applicable because this study is based exclusively on published literature. All methods were carried out according to relevant guidelines and regulations.

The authors have no conflicts of interest to declare.

This work was supported by the National High-Level Hospital Clinical Research Funding (2024-NHLHCRF-YS-01 and 2024-NHLHCRF-JBGS-WZ-06), Beijing Research Ward Construction Clinical Research Project (2022-YJXBF-04-03), National Natural Science Foundation of China (No. 82270352), Capital’s Founds for Health Improvement and Research (No. 2022-1-4062), and Chinese Society of Cardiology’s Foundation (No. CSCF2021B02).

K.D., J.Z., and Z.Y. contributed to the study design. Z.Y., E.X., and Z.G. contributed to data collection, manuscript writing, data processing, and figure mapping. K.D. and J.Z. contributed to the data proofreading. Z.Y. contributed to formal analysis; and writing – original draft preparation, Y.G. and Z.H. contributed to review and editing. All authors have read and agreed to the published version of the manuscript.

Zixiang Ye and Enmin Xie contributed equally to this work and share the first authorship.

The data that support the findings of this study are openly available from the National Health and Nutrition Examination Survey (NHANES) at https://www.cdc.gov/nchs/nhanes/index.htm.