Abstract
Objective
The level of systemic inflammation plays a critical role in influencing liver function. This study aims to investigate the impact of various composite inflammatory markers on the liver during the COVID-19 pandemic.
Methods
Data were sourced from the updated NHANES database, reflecting information collected during the COVID-19 pandemic between 2021 and 2023. Based on complete blood count (CBC) data, we identified six composite inflammatory markers: neutrophil-to-lymphocyte ratio (NLR), derived neutrophil-to-lymphocyte ratio (dNLR), monocyte-to-lymphocyte ratio (MLR), neutrophil-monocyte-to-lymphocyte ratio (NMLR), systemic inflammation response index (SIRI), and systemic immune-inflammation index (SII). We described the baseline characteristics of the study population and preliminarily explored the effects of inflammation levels on liver stiffness and fatty liver using linear regression models under various fitting conditions. Furthermore, Spearman correlation analysis was performed to assess the relationships between covariates and the primary variables, and restricted cubic splines (RCS) were applied to examine the nonlinear relationships between variables. Finally, structural equation modeling (SEM) was used to investigate the mediating role of inflammatory markers in modulating the effects of different variables on liver outcomes.
Results
A total of 4,165 participants were included in the present study. After adjustment for a comprehensive set of covariates and potential confounders, SIRI demonstrated relatively stable associations with liver-related outcomes compared to other composite inflammatory indices, exhibiting a positive correlation with both median liver stiffness (MLS) and the controlled attenuation parameter (CAP). RCS analysis revealed a significant non-linear relationship between SIRI and CAP (P < 0.05). In the final mediation analysis, SIRI was identified as a mediator in the relationship between age, body mass index (BMI), and MLS. Similarly, SIRI mediated the effects of age, BMI, hemoglobin, and platelet count on CAP, further highlighting its central role in inflammation-related hepatic alterations.
Conclusions
Our findings indicate that during the COVID-19 pandemic, SIRI exerted a significant impact on liver health, demonstrating particular utility in the early detection of hepatic fibrosis and steatosis. These results underscore the potential of SIRI as a valuable marker in the prevention and management of liver diseases, warranting greater attention in clinical and public health settings.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-025-24174-1.
Keywords: COVID-19, Inflammatory level, Cirrhosis, Fatty liver, SIRI
Introduction
Cirrhosis is a chronic, progressive, and irreversible liver disease characterized by hepatocellular damage and replacement by fibrotic tissue, ultimately leading to structural distortion and functional impairment of the liver. The etiology of cirrhosis is multifactorial, with the most common causes including viral hepatitis, chronic alcohol consumption, non-alcoholic fatty liver disease (NAFLD), autoimmune liver disorders, biliary tract diseases, inherited metabolic conditions, and exposure to certain hepatotoxic drugs or toxins [1, 2].Fatty liver disease is a pathological condition characterized by the abnormal accumulation of lipids, primarily triglycerides, within hepatocytes, leading to hepatic structural and functional alterations. Fatty liver disease is broadly classified into alcoholic fatty liver disease (AFLD) and NAFLD. The etiology of fatty liver Disease is multifactorial and includes excessive alcohol consumption, obesity, type 2 diabetes or insulin resistance, dyslipidemia, unhealthy dietary patterns characterized by high caloric, high-fat, and high-sugar intake, a sedentary lifestyle, certain pharmacological agents (such as corticosteroids, chemotherapeutic drugs, and specific antidepressants), as well as rapid weight loss or malnutrition [3, 4].Cirrhosis and fatty liver disease are both prevalent hepatic disorders that, despite differences in pathophysiological mechanisms, etiologies, and clinical manifestations, are interconnected. Fatty liver disease, if left untreated, may progress to cirrhosis over time. Maintaining a healthy body weight and preventing obesity are crucial preventive strategies [5]. Lifestyle interventions, including regular physical activity and a balanced diet, play a pivotal role in mitigating disease onset and progression. Furthermore, routine medical check-ups and liver function assessments are essential for monitoring disease status and facilitating early intervention [6].
Liver enzyme markers, which reflect the extent of hepatocellular injury, along with bilirubin levels and lipid profiles that assess hepatic metabolism and bile excretion, are closely associated with liver function [7]. In addition to these parameters, systemic inflammation may also play a crucial role in liver health [8, 9]. The complete blood count (CBC) is a widely utilized hematological test that provides valuable insights into overall blood health and aids in the detection of various conditions, including infections, anemia, inflammation, and hematologic disorders [10]. By analyzing the three major types of blood cells—red blood cells, white blood cells, and platelets, along with their associated indices, the CBC serves as a fundamental tool for clinical diagnosis and health monitoring.In liver diseases such as cirrhosis and fatty liver disease, inflammation plays a pivotal role in both the onset and progression of the disease, potentially driving the development of liver fibrosis [11]. Therefore, comprehensive assessment of systemic inflammatory status is crucial for the early detection and management of liver disorders. Unlike CBC parameters, we have evaluated six composite inflammatory indices to provide a more nuanced understanding of systemic inflammation: the neutrophil-to-lymphocyte ratio (NLR), the derived neutrophil-to-lymphocyte ratio (dNLR), the monocyte-to-lymphocyte ratio (MLR), the neutrophil-monocyte-to-lymphocyte ratio (NMLR), the systemic inflammation response index (SIRI), and the systemic immune-inflammation index (SII) [12]. These indices offer valuable insights into the complex inflammatory milieu associated with liver disease and may serve as potential biomarkers for disease monitoring and risk stratification.
The intricate interplay between SARS-CoV-2 and the liver involves both direct viral insult and host-related modulation of disease progression. Emerging evidence suggests that SARS-CoV-2 gains hepatic entry primarily via cholangiocytes, which express high levels of the ACE2 receptor, leading to cholestasis [13]. In parallel, the excessive release of pro-inflammatory cytokines such as interleukin-6 (IL-6) activates hepatic stellate cells, contributing to systemic cytokine storm and accelerating fibrogenesis [14]. Furthermore, the hepatotoxicity of several COVID-19 therapeutic agents—including lopinavir/ritonavir and certain antibiotics—may independently induce or exacerbate hepatic fibrosis [15]. These mechanisms underscore the importance of vigilant inflammation monitoring and liver fibrosis surveillance not only in individuals with pre-existing liver disease, but also in the general population in the post-pandemic era [16].
Liver Ultrasound Transient Elastography (LUTE) is a non-invasive imaging technique widely utilized to assess liver stiffness, serving as an indirect measure of liver fibrosis and the severity of liver disease [17]. In this study, we selected the median liver stiffness (MLS) and the median controlled attenuation parameter (CAP) as outcome indicators to evaluate the extent of liver fibrosis and hepatic steatosis within the study population. Six distinct composite inflammatory indices were employed to analyze their associations with liver fibrosis and fatty liver disease.Prior to the commencement of the study, we formulated three primary hypotheses: [1] composite inflammatory indices are strongly associated with the two selected outcome indicators; [2] composite inflammatory indices are correlated with covariates and confounding factors; and [3] composite inflammatory indices may mediate the effects of other factors on liver health.Through this investigation, we aim to identify potential predictive factors and diagnostic biomarkers that could facilitate the prevention and progression of liver diseases. Our findings are expected to provide valuable insights for clinical health screening and contribute to the development of more effective preventive strategies for liver-related disorders.
Methods and materials
Study participants
This study utilized data from the National Health and Nutrition Examination Survey (NHANES), a comprehensive health and nutrition database jointly administered by the Centers for Disease Control and Prevention (CDC) and the National Center for Health Statistics (NCHS). We analyzed data collected during the COVID-19 pandemic, specifically from the period 2021–2023, and included the following variables: sex, age, race, body mass index (BMI), median liver stiffness (MLS), controlled attenuation parameter (CAP), alcohol use, hepatitis B virus (HBV), high-density lipoprotein (HDL), total cholesterol (TCHOL), CBC, red blood cell (RBC) folate, and High-Sensitivity C-reactive protein (HS-CRP).
Inclusion and exclusion criteria were applied as follows: Initially, a total of 11,933 participants were enrolled in the NHANES dataset collected during the COVID-19 pandemic period (2021–2023). All individuals were included in the initial analysis and subsequently subjected to rigorous data screening procedures. Data for BMI were missing in 3,462 cases, LUTE data were missing in 5,234 cases, HBV data were unavailable for 1,241 individuals, alcohol use data were missing in 7,011 cases, HDL and TCHOL data were absent for 5,043 participants, leukocyte and erythrocyte counts were missing in 4,340 individuals, lymphocyte, monocyte, and neutrophil data were incomplete for 4,351 participants, RBC folate data were missing for 4,427 individuals, and HS-CRP data were unavailable for 4,651 participants. Ultimately, 4,165 participants with complete data across all study variables were included in the final analysis(Fig. 1).In addition, Supplementary file 1 outlined the complete data selection process.
Fig. 1.
Flowsheet
It is important to note that NHANES data are publicly available and can be accessed for free via the NCHS website, allowing researchers to download both raw and processed datasets for further analysis.
Inflammatory derived indicators
From the CBC, we obtained measurements of neutrophil count, lymphocyte count, total leukocyte count, and monocyte count. Using these four hematological parameters, we calculated six inflammation-derived indices: NLR = neutrophil count/lymphocyte count, dNLR = neutrophil count/(leukocyte count - lymphocyte count), MLR = monocyte count/lymphocyte count, NMLR = (monocyte count + neutrophil count)/lymphocyte count, SIRI = neutrophil count * monocyte count/lymphocyte count, SII = platelet count * neutrophil count/lymphocyte count.
Liver ultrasound transient elastography
The primary objective of the NHANES liver ultrasound transient elastography(LUTE) component is to provide objective measurements for two critical liver conditions: liver fibrosis (scarring of the liver) and hepatic steatosis (fat accumulation in the liver). Liver fibrosis is assessed using FibroScan, a device that employs ultrasound and Vibration-Controlled Transient Elastography (VCTE™) to determine liver stiffness. The device also simultaneously measures ultrasound attenuation associated with hepatic steatosis and records the controlled attenuation parameter (CAP™) as a marker of liver fat content. LUTE is a well-established non-invasive modality for rapid assessment of hepatic fibrosis severity and steatosis degree [18]. In the current study, both liver ultrasound indicators were considered as outcome variables, and their associations with various inflammation-derived indices were examined.
Critically, we retained both LUTE outcomes as continuous variables to maximize analytical precision. Categorization would risk substantial information loss through reduced granularity, obscuring within-group heterogeneity and potentially masking non-linear associations—while simultaneously constraining the flexibility of statistical modeling.In addition, as widely used clinical measures, MLS values below 7.0 kPa are generally considered within the normal range. For CAP, values below 238 dB/m indicate normal liver status, whereas those exceeding 290 dB/m are indicative of severe hepatic steatosis [19, 20].
Analysis & design
In addition to examining the relationships between the primary variables and outcome measures of interest in this study, we also categorized several covariates. Following the guidelines established by the World Health Organization (WHO), BMI was classified into the following categories: <18.5, 18.5–24.9, 25.0–29.9, and ≥ 30. For alcohol consumption, participants’ drinking frequency over the past year was assessed using questionnaire data, with drinking frequency categorized into five groups: “Never”,“Daily”,“Weekly”, “Monthly” and “Yearly”.
NHANES data are collected through a complex, multi-stage probability design, including stratification, cluster sampling, and weight adjustments. The purpose of these weights is to ensure that the sample accurately represents the non-institutionalized population of the United States. However, in the present study, our focus is not on the representativeness of the participants but rather on the relationships between specific variables, such as inflammation-derived indices and liver ultrasound markers. In conducting the analysis, we opted not to apply weights, particularly given that the data generation process was already appropriately designed. Our primary interest lies in the internal structure of the model and hypothesis testing, rather than ensuring sample representativeness.
First, only participants with complete data were included in this study. Next, independent samples t-tests and chi-square tests were used to compare baseline characteristics between the HBV and non-HBV groups, depending on the variable type. Linear regression analyses were conducted to examine the relationships between six different inflammation-related variables and two liver ultrasound outcome markers. Spearman’s rank correlation was employed to assess pairwise associations between variables. To explore the non-linear relationships between inflammation indices and outcomes, restricted cubic splines were applied while adjusting for different models. Structural equation modeling (SEM), a multivariate statistical technique that combines factor and path analysis, was utilized to investigate the complex interrelationships between variables. SEM allows for the simultaneous analysis of multiple dependent and independent variables, incorporating both latent and observed variables. We employed SEM to explore the mediating effects of inflammation-derived indices on covariates and outcome variables. All statistical analyses were performed with a significance threshold of P < 0.05.
Result
Baseline characteristics of participants
Based on data from the NHANES database (2021–2023), presented in Table 1, a total of 4,165 participants were included, consisting of 1,957 males (47%) and 2,208 females (53%). Among these participants, 27.9% were under 40 years of age, 28.8% were aged 40–59 years, 39.8% were aged 60–79 years, and 3.6% were aged 80 or older. The participants were classified into two groups: the HBV group (54 participants) and the non-HBV group (4,111 participants), and intergroup differences were assessed. Significant differences were observed in terms of age and race between the groups (P < 0.05), while baseline characteristics for other variables were balanced, with intergroup differences likely attributed to random fluctuation (P > 0.05). Supplementary files 2 and 3 present quartile-based stratifications by MLS and CAP levels, respectively, detailing additional associations between covariates and clinical outcomes.
Table 1.
Baseline data of the population
| Characteristic | Total(4165) | Non-HBV(4111) | HBV(54) | P |
|---|---|---|---|---|
| Sex (%) | 0.811 | |||
| Male | 1957 (47.0) | 1933 (47.0) | 24 (44.4) | |
| Female | 2208 (53.0) | 2178 (53.0) | 30 (55.6) | |
| Age(%) | 0.002 | |||
| <40 | 1160 (27.9) | 1157 (28.1) | 3 (5.6) | |
| 40-59 | 1199 (28.8) | 1181 (28.7) | 18 (33.3) | |
| 60-79 | 1658 (39.8) | 1629 (39.6) | 29 (53.7) | |
| 80 &>80 | 148 (3.6) | 144 (3.5) | 4 (7.4) | |
| Race(%) | 0.049 | |||
| Mexican American | 277 (6.7) | 274 (6.7) | 3 (5.6) | |
| Other Hispanic | 396 (9.5) | 385 (9.4) | 11 (20.4) | |
| Non-Hispanic White | 2630 (63.1) | 2602 (63.3) | 28 (51.9) | |
| Non-Hispanic Black | 432 (10.4) | 428 (10.4) | 4 (7.4) | |
| Other Race - Including Multi-Racial | 430 (10.3) | 422 (10.3) | 8 (14.8) | |
| BMI(mean (SD), kg/m2) | 0.557 | |||
| <18.5 | 55 (1.3) | 55 (1.3) | 0 (0.0) | |
| 18.5-24.9 | 1089 (26.1) | 1072 (26.1) | 17 (31.5) | |
| 25.0-29.9 | 1350 (32.4) | 1331 (32.4) | 19 (35.2) | |
| ≥30 | 1671 (40.1) | 1653 (40.2) | 18 (33.3) | |
| MLS(mean (SD), Kpa) | 6.39 (6.74) | 6.39 (6.75) | 6.27 (6.37) | 0.903 |
| CAP(mean (SD), dB/m) | 262.32 (63.08) | 262.31 (63.08) | 262.93 (62.91) | 0.943 |
| Alcoho use(last year,%) | 0.366 | |||
| Never | 674 (16.2) | 661 (16.1) | 13 (24.1) | |
| Daily | 392 (9.4) | 388 (9.4) | 4 (7.4) | |
| Weekly | 1201 (28.8) | 1186 (28.8) | 15 (27.8) | |
| Monthly | 788 (18.9) | 782 (19.0) | 6 (11.1) | |
| Yearly | 1110 (26.7) | 1094 (26.6) | 16 (29.6) | |
| HDL(mean (SD), mmol/L) | 1.42 (0.38) | 1.42 (0.38) | 1.44 (0.39) | 0.721 |
| TCHOL(mean (SD), mmol/L) | 4.89 (1.09) | 4.89 (1.09) | 4.84 (1.08) | 0.734 |
| Leukocyte(mean (SD),103 cell/uL) | 6.86 (1.97) | 6.86 (1.97) | 6.71 (2.28) | 0.568 |
| Lymphocyte(mean (SD),103 cell/uL) | 2.01 (0.69) | 2.01 (0.68) | 2.07 (1.26) | 0.520 |
| Monocyte(mean (SD),103 cell/uL) | 0.54 (0.18) | 0.54 (0.18) | 0.52 (0.18) | 0.303 |
| Neutrophils(mean (SD),103 cell/uL) | 4.07 (1.54) | 4.07 (1.55) | 3.88 (1.28) | 0.362 |
| Erythrocyte(mean (SD),103 cell/uL) | 4.68 (0.46) | 4.68 (0.46) | 4.62 (0.46) | 0.381 |
| Hemoglobin (mean (SD), g/dL) | 14.00 (1.41) | 13.99 (1.41) | 14.04 (1.33) | 0.805 |
| Platelet(mean (SD),103 cell/uL) | 255.57 (65.35) | 255.76 (65.21) | 241.09 (74.52) | 0.101 |
| RBC Folate(mean (SD), nmol/L) | 1213.71 (540.84) | 1214.14 (542.48) | 1180.65 (398.56) | 0.651 |
| HS-CRP(mean (SD), mg/L) | 3.64 (6.91) | 3.65 (6.94) | 3.43 (5.05) | 0.817 |
| SII(mean (SD),103 cell/uL) | 567.21 (344.51) | 567.71 (344.90) | 529.43 (314.12) | 0.417 |
| SIRI(mean (SD),103 cell/uL) | 1.24 (0.94) | 1.24 (0.94) | 1.13 (0.59) | 0.402 |
| NMLR(mean (SD)) | 2.50 (1.22) | 2.50 (1.22) | 2.49 (1.19) | 0.929 |
| MLR(mean (SD)) | 0.29 (0.13) | 0.29 (0.13) | 0.29 (0.12) | 0.792 |
| dNLR(mean (SD)) | 0.83 (0.05) | 0.83 (0.05) | 0.84 (0.05) | 0.785 |
| NLR(mean (SD)) | 2.21 (1.12) | 2.21 (1.12) | 2.20 (1.10) | 0.951 |
Linear regression of MLS and CAP
Linear regression analyses of the two outcome measures, MLS and CAP, were conducted under different model adjustments, as presented in Tables 2 and 3. To further explore the impact of inflammation-derived indices as both categorical and continuous variables on the outcomes, these indices were Divided into four equal quantiles based on the 25th, 50th, and 75th percentiles (Q1, Q2, Q3, Q4).
Table 2.
Linear regression between different composite inflammatory markers and MLS
| Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|
| Beta(95%CI) | P-value | Beta(95%CI) | P-value | Beta(95%CI) | P-value | |
| SII | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | −1.09(−1.76,−0.42) | 0.002 | −0.89(−1.56,−0.23) | 0.008 | −0.28(−1.02,0.47) | 0.464 |
| Q3 | −1.69(−2.50,−0.87) | <0.001 | −1.35(−2.17,−0.54) | 0.001 | −0.15(−1.18,0.88) | 0.777 |
| Q4 | −2.75(−3.72,−1.77) | <0.001 | −2.37(−3.35,−1.38) | <0.001 | −0.35(−1.77,1.06) | 0.626 |
| Continuous | −0.005(−0006,−0.004) | <0.001 | −0.005(−0.006,−0.004) | <0.001 | −0.006(−0.008,−0.004) | <0.001 |
| SIRI | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 0.78(0.09,1.48) | 0.027 | 0.33(−0.35,1.02) | 0.340 | 0.53(−0.16,1.23) | 0.131 |
| Q3 | 1.41(0.55,2.28) | 0.001 | 0.73(−0.13,1.59) | 0.097 | 1.02(0.15,1.90) | 0.021 |
| Q4 | 2.80(1.72,3.87) | <0.001 | 1.70(0.62,2.78) | 0.002 | 2.11(1.01,3.21) | <0.001 |
| Continuous | 0.998(0.513,1.484) | <0.001 | 0.603(0.118,1.088) | 0.015 | 0.573(0.070,1.077) | 0.003 |
| NMLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 0.11(−1.37,1.59) | 0.883 | −0.11(−1.57,1.34) | 0.880 | −0.01(−1.48,1.46) | 0.990 |
| Q3 | 0.40(−1.58,2.36) | 0.694 | 0.27(−1.66,2.21) | 0.783 | 0.03(−1.93,1.99) | 0.976 |
| Q4 | 1.00(−1.54,3.54) | 0.440 | 0.70(−1.81,3.20) | 0.586 | 0.37(−2.17,2.90) | 0.777 |
| Continuous | −6.603(−49.74,36.54) | 0.764 | −6.407(−48.75,35.94) | 0.767 | −7.495(−50.37,35.38) | 0.732 |
| MLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | −0.66(−1.31,0.01) | 0.050 | −0.49(−1.14,0.16) | 0.136 | −0.64(−1.29,0.01) | 0.054 |
| Q3 | −1.05(−1.85,−0.25) | 0.010 | −0.86(−1.66,−0.07) | 0.034 | −1.05(−1.85,−0.25) | 0.011 |
| Q4 | −0.74(−1.70,−0.22) | 0.133 | −0.44(−1.41,0.52) | 0.366 | −0.75(−1.72,0.21) | 0.126 |
| Continuous | 6.93(−36.24,50.10) | 0.753 | 7.904(−34.46,50.27) | 0.715 | 9.125(−33.79,52.04) | 0.677 |
| dNLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 0.09(−0.55,0.72) | 0.790 | 0.17(−0.46,0.79) | 0.605 | 0.04(−0.59,0.67) | 0.908 |
| Q3 | 0.30(−0.35,0.94) | 0.373 | 0.34(−0.30,0.98) | 0.301 | 0.13(−0.52,0.77) | 0.705 |
| Q4 | −0.12(−0.88,0.64) | 0.764 | 0.08(−0.67,0.83) | 0.836 | −0.35(−1.11,0.41) | 0.365 |
| Continuous | −1.900(−7.582,3.782) | 0.512 | −1.330(−6.961, 4.302) | 0.644 | −1.545(−7.217,4.128) | 0.594 |
| NLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 0.60(−0.86,2.06) | 0.422 | 0.71(−0.73,2.14) | 0.335 | 0.30(−1.15,1.75) | 0.683 |
| Q3 | 0.78(−1.15,2.71) | 0.430 | 0.84(−1.06,2.74) | 0.388 | 0.30(−1.64,2.23) | 0.763 |
| Q4 | 1.01(−1.48,3.50) | 0.426 | 1.26(−1.18,3.71) | 0.312 | 0.18(−2.33,2.68) | 0.891 |
| Continuous | 7.916(−35.24,51.07) | 0.719 | 7.743(−34.62,50.11) | 0.720 | 9.185(−33.71,52.08) | 0.675 |
Model 1: Adjusted for SII, NMLR, MLR, dNLR, NLR
Model 2: Adjusted for SII, NMLR, MLR, dNLR, NLR, Sex, Age, Race, BMI, Alcohol use
Model 3: Adjusted for SII, NMLR, MLR, dNLR, NLR, Erythrocyte, Hemoglobin, Platelet, RBC Folate, HS-CRP, HDL, TCHOL
Table 3.
Linear regression between different composite inflammatory markers and CAP
| Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|
| Beta(95%CI) | P-value | Beta(95%CI) | P-value | Beta(95%CI) | P-value | |
| SII | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 2.70(−3.53,8.93) | 0.396 | 6.16(0.95,11.36) | 0.020 | 0.12(−6.32,6.57) | 0.970 |
| Q3 | −1.38(−8.93,6.17) | 0.720 | 3.84(−2.54,10.22) | 0.238 | −3.87(−12.80,5.06) | 0.396 |
| Q4 | −2.37(−11.41,6.66) | 0.606 | 2.67(−5.04,10.37) | 0.497 | −8.63(−20.90,3.64) | 0.168 |
| Continuous | −0.005(−0.016,0.006) | 0.317 | 0.001(−0.007,0.011) | 0.764 | −0.022(−0.041,−0.003) | 0.026 |
| SIRI | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 20.91(14.48,27.33) | <0.001 | 8.31(2.92,13.69) | 0.003 | 11.18(5.18,17.18) | <0.001 |
| Q3 | 32.03(24.02,40.04) | <0.001 | 12.20(5.46,18.94) | <0.001 | 15.13(7.59,22.67) | <0.001 |
| Q4 | 57.60(47.62,67.58) | <0.001 | 27.15(18.70,35.60) | <0.001 | 31.57(22.08,41.06) | <0.001 |
| Continuous | 23.91(19.36,28.46) | <0.001 | 14.02(1.02,17.85) | <0.001 | 10.51(6.11,14.92) | <0.001 |
| NMLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | −1.19(−14.89,12.52) | 0.865 | −4.94(−16.32,6.442) | 0.395 | 3.87(−8.84,16.59) | 0.550 |
| Q3 | −7.73(−25.94,10.48) | 0.406 | −8.19(−23,31,6.92) | 0.288 | −0.49(−17.43,16.46) | 0.955 |
| Q4 | −2.40(−25.96,21.16) | 0.842 | −6.33(−25,90,13.24) | 0.526 | 6.59(−15.37,28.55) | 0.556 |
| Continuous | −272.1(−676.7,132.4) | 0.187 | −252.8(−587.5,81.96) | 0.139 | −333.2(−707.9,41.58) | 0.081 |
| MLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | −11.20(−17.28,−5.12) | <0.001 | −4.47(−9.54,0.60) | 0.084 | −6.41(−12.06,−0.76) | 0.026 |
| Q3 | −20.56(−27.99,−13.13) | <0.001 | −12.21(−18.44,−5.97) | <0.001 | −13.38(−20.31,−6.44) | <0.001 |
| Q4 | −36.66(−45.55,−27.76) | <0.001 | −22.77(−30.31,−15.23) | <0.001 | −24.25(−32.62,−15.87) | <0.001 |
| Continuous | 164.1(−240.8,568.9) | 0.4727 | 174.7(−161.1,509.6) | 0.307 | 277.8(−97.20,652.9) | 0.147 |
| dNLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 2.66(−3.23,8.55) | 0.376 | 3.39(−1.51,8.30) | 0.175 | 3.04(−2.42,8.50) | 0.275 |
| Q3 | 0.05(−5.97,6.07) | 0.987 | −0.002(−5.00,5.00) | 0.999 | −0.17(−5.77,5.43) | 0.953 |
| Q4 | −3.53(−10.59,3.52) | 0.326 | 0.56(−5.31,6.44) | 0.851 | −3.45(−10.03,3.14) | 0.305 |
| Continuous | −10.12(−63.40,43.16) | 0.71 | 3.49(−41.02,48.00) | 0.878 | 5.43(−44.14,55.01) | 0.830 |
| NLR | ||||||
| Q1 | Ref | Ref | Ref | |||
| Q2 | 1.10(−12,42,14.63) | 0.873 | 2.15(−9.07,13.37) | 0.708 | 1.86(−10.70,14.43) | 0.771 |
| Q3 | 4.12(−13.80,22.05) | 0.652 | 3.76(−11.11,18.64) | 0.620 | 7.76(−8.99,24.50) | 0.364 |
| Q4 | −7.38(−30.43,15.67) | 0.531 | −2.88(−22.02,16.26) | 0.768 | −0.69(−22.37,20.98) | 0.950 |
| Continuous | 268.5(−136.1,673.2) | 0.193 | 249.4(−85.49,584.21) | 0.144 | 337.5(−37.41,712.4) | 0.778 |
Model 1: Adjusted for SII, NMLR, MLR, dNLR, NLR
Model 2: Adjusted for SII, NMLR, MLR, dNLR, NLR, Sex, Age, Race, BMI, Alcohol use
Model 3: Adjusted for SII, NMLR, MLR, dNLR, NLR, Erythrocyte, Hemoglobin, Platelet, RBC Folate, HS-CRP, HDL, TCHOL
When MLS was used as the outcome, in the unadjusted model, SII in its continuous form showed a negative correlation with MLS (β: −0.005, P < 0.05), while, similarly, the categorical form also exhibited a negative association (Q2–Q4: −1.09, −1.69, −2.75, P < 0.05). After adjusting for selected covariates, the results from Model 2 remained consistent. In Model 3, however, only the continuous form of SII retained statistical significance (β: −0.006, P < 0.05), while the categorical variable was no longer significant (P > 0.05).
Compared to SII, the continuous form of SIRI remained consistently and positively associated with MLS across all models (β: 0.998, 0.603, 0.573; all P < 0.05). In its categorical form, SIRI showed statistically significant associations in Model 1 across all quartiles (Q2–Q4: 0.78, 1.41, 2.80; all P < 0.05). In Model 2, the highest quartile of SIRI (Q4) retained significance (β: 1.70, P = 0.002), and in Model 3, SIRI remained influential, with Q3 and Q4 showing significant associations (β: 1.02, 2.01; P < 0.05). In contrast, MLR exhibited no statistically significant association with MLS across models, except in Q3.
Subsequently, all inflammation-derived indices were incorporated into the analysis of CAP. SII, in both its continuous and categorical forms, showed no significant association with CAP. In contrast, the continuous form of SIRI demonstrated a robust and statistically significant positive correlation with CAP across all models (β: 23.91, 14.02, 10.51; all P < 0.001), and its categorical form also consistently exhibited significant associations across quartiles. While MLR showed some degree of statistical significance, its association with CAP proved unstable under Different model specifications and covariate adjustments. Taken together, these findings underscore the stable and prominent influence of SIRI on both hepatic outcomes assessed in this study.Notably, several covariates from Models 2 and 3 were included in the final mediation analysis.Their associations—derived through linear analyses—with the six inflammation-derived indices treated as categorical variables are presented in Supplementary files 4 and 5.
Spearman’s correlation analysis visualization
The Spearman correlation matrix heatmap is an excellent method for illustrating pairwise correlations among multiple variables. In Fig. 2, we examine the correlations between inflammation-derived variables and covariates in this study, with a particular focus on the SII and SIRI indices. The data visualization provides a clear representation of the results: Age, BMI, HS-CRP, Erythrocyte count, RBC Folate, Hemoglobin, and Platelet count all exhibit varying degrees of positive correlations with SIRI, with HS-CRP showing the strongest correlation. In contrast, HDL, Race, and Sex demonstrate negative correlations with SIRI, with Sex showing the most pronounced negative relationship. Furthermore, Age, BMI, HS-CRP, RBC Folate, Platelet count, Sex, and SII are positively correlated, with HS-CRP again displaying the strongest positive association. Conversely, HDL and Race exhibit negative correlations with SII, with the relationship with Race being notably stronger.
Fig. 2.
The relationship between six inflammation-derived indices and various covariates
Nonlinear relationship of inflammatory derived indicators
To capture the nonlinear relationships among the data in this study, we employed Restricted Cubic Splines (RCS) to analyze the associations between SIRI and MLS, as well as between SIRI and CAP. As shown in Fig. 3, the results indicate that the relationship between SIRI and MLS is not statistically significant (P > 0.05). In contrast, the analysis of SIRI with CAP revealed a significant nonlinear positive correlation. The visualized graph further elucidates the underlying connection between the variables.
Fig. 3.
The nonlinear relationships between SIRI and MLS, CAP. Panels (A), (B), and (C) Display the nonlinear associations between SIRI and MLS after adjustment for models 1, 2, and 3, respectively. Panels (D), (E), and (F) show the nonlinear relationships between SIRI and CAP after adjustment for models 1, 2, and 3, respectively.Model 1: Adjusted for SII, NMLR, MLR, dNLR, NLR.Model 2: Adjusted for SII, NMLR, MLR, dNLR, NLR, Sex, Age, Race, BMI, Alcohol use.Model 3: Adjusted for SII, NMLR, MLR, dNLR, NLR, Erythrocyte, Hemoglobin, Platelet, RBC Folate, HS-CRP, HDL, TCHOL
The mediating effect of SIRI
Based on the results of the previous correlation analysis, we incorporated six different variables that were positively correlated with SIRI into a mediation analysis to explore their relationships with MLS and CAP (Fig. 4). In the mediation analysis for MLS, we found that SIRI mediates the associations between age, and BMI (Table 4). Despite statistically significant associations (P < 0.05) for all platelet-related effect estimates, the directional opposition between the total effect and indirect effect on MLS precludes consideration of SIRI as a mediating factor. For CAP, in addition to the three aforementioned variables, we further identified that SIRI also mediates the relationship with hemoglobin and platelet (Table 5).
Fig. 4.
Mediation analysis of SIRI
Table 4.
Mediation analysis between SIRI and MLS
| variable | Effect(a) (Beta,95%) |
P-value | Effect(b) (Beta,95%) |
P-value | Indirect effect (Beta,95%) |
P-value | Total effect (Beta,95%) |
P-value |
|---|---|---|---|---|---|---|---|---|
| Age | 0.007(0.005,0.009) | <0.001 | 0.866(0.479,1.325) | <0.001 | 0.006(0.003,0.011) | 0.001 | 0.038(0.028,0.048) | <0.001 |
| BMI | 0.013(0.008,0.300) | <0.001 | 0.776(0.382,1.240) | <0.001 | 0.010(0.004,0.017) | 0.002 | 0.258(0.213,0.307) | <0.001 |
| Erythrocyte | 0.039(−0.03,0.943) | 0.288 | 0.943(0.551,1.397) | <0.001 | 0.037(−0.03,0.123) | 0.331 | 0.446(0.033,0.921) | 0.051 |
| RBC Folate | 0 | - | 0.901(0.517,1.334) | <0.001 | - | - | - | - |
| Hemoglobin | 0.048(0.026,0.072) | <0.001 | 0.937(0.539,1.408) | <0.001 | 0.045(0.021,0.075) | <0.001 | 0.135(−0.038,0.32) | 0.136 |
| Platelet | 0.002(0.001,0.002) | <0.001 | 1.039(0.638,1.513) | <0.001 | 0.002(0.001,0.003) | <0.001 | −0.01(−0.013,−0.005) | <0.001 |
Table 5.
Mediation analysis between SIRI and CAP
| variable | Effect(a) (Beta,95%) |
P-value | Effect(b) (Beta,95%) |
P-value | Indirect effect (Beta,95%) |
P-value | Total effect (Beta,95%) |
P-value |
|---|---|---|---|---|---|---|---|---|
| Age | 0.007(0.005,0.009) | <0.001 | 5.720(3.595,7.906) | <0.001 | 0.042(0.026,0.061) | 0.001 | 0.568(0.460,0.676) | <0.001 |
| BMI | 0.013(0.008,0.018) | <0.001 | 3.558(1.637,5.464) | <0.001 | 0.045(0.020,0.075) | 0.001 | 5.131(4.861,5.410) | <0.001 |
| Hemoglobin | 0.048(0.026,0.072) | <0.001 | 6.558(4.409,8.764) | <0.001 | 0.315(0.162,0.508) | <0.001 | 4.195(3.546,6.176) | <0.001 |
| RBC Folate | 0 | - | 6.231(4.208,8.392) | <0.001 | - | - | - | - |
| Erythrocyte | 0.039(−0.03,0.114) | 0.287 | 6.85(4.774,8.963) | <0.001 | 0.268(−0.203,0.804) | 0.291 | 22.139(18.161,26.203) | <0.001 |
| Platelet | 0.002(0.001,0.002) | <0.001 | 6.584(4.348,8.801) | <0.001 | 0.011(0.007,0.016) | <0.001 | 0.067(0.04,0.097) | <0.001 |
Discussion
Among 4,165 participants from NHANES, we analyzed data from 54 individuals with HBV and 4,111 individuals without HBV. We included several study variables, encompassing different data types for the same variable (categorical and continuous). To obtain more robust results, we performed multivariable linear regression analysis for the study variables across different models and identified that an increase in SIRI was associated with elevated risks of MLS and CAP. Concurrently, we conducted RCS analysis on SIRI and the two outcome measures, revealing a nonlinear positive correlation between SIRI and CAP. Additionally, we performed correlation analysis between inflammation-derived variables and covariates, identifying six different covariates positively correlated with SIRI. Based on these findings, we further conducted mediation analysis and found that SIRI acts as a mediator in the relationship between the six associated covariates and MLS, CAP. These results further support our hypothesis that inflammation-derived markers may be linked to liver fibrosis and fatty liver to some extent.
Inflammatory responses are tightly linked to the pathogenesis and progression of liver cirrhosis, serving as a central mechanism driving the disease [21]. In cirrhosis, normal hepatocytes are progressively replaced by fibrous tissue, resulting in severe impairment of liver architecture and function [22]. Upon repeated liver injury, the immune system activates an inflammatory cascade aimed at clearing damaged cells and pathogens [23]. However, in chronic liver disease, this inflammatory response remains unresolved, leading to persistent tissue injury and a continuous cycle of repair [24]. Chronic inflammation, in turn, drives the activation of fibroblasts and hepatic stellate cells (HSCs), which underpins the development of liver fibrosis [25]. Moreover, inflammatory cells—such as neutrophils, monocytes, and lymphocytes—accumulate at sites of injury within the liver. These immune cells release a variety of pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukins IL-6, IL-1, and others, further exacerbating local inflammation [26].Neutrophils, essential components of the innate immune system, primarily mediate acute inflammatory responses and antimicrobial defense. In the context of chronic liver injury, however, overactivated neutrophils release substantial quantities of inflammatory mediators and reactive oxygen species, driving sustained hepatocellular damage. This not only exacerbates tissue injury but also promotes the activation of HSCs, thereby accelerating the process of liver fibrosis and contributing to the worsening of cirrhosis.Lymphocytes—comprising T cells, B cells, and natural killer (NK) cells—also play pivotal roles in the immune response to liver cirrhosis [27]. T cells, in particular, are central to the progression of chronic inflammation and immune-mediated liver injury. Through sustained activation, they drive the inflammatory processes that perpetuate liver damage [28].Monocytes, key players of the innate immune response, are integral to the onset and progression of cirrhosis, as well as the associated immune reactions [29]. These cells contribute to fibrosis through cytokine secretion and HSCs activation, while also modulating the chronic inflammatory environment in the liver. A deeper understanding of the roles of these immune cells in cirrhosis could provide valuable insights into the mechanisms driving disease progression, offering potential therapeutic targets aimed at mitigating fibrosis and inflammation.
Similarly, inflammation plays a critical role in the onset and progression of fatty liver disease, particularly during the transition from simple steatosis to non-alcoholic steatohepatitis [30]. As the disease progresses, the liver becomes infiltrated by various immune cells, which not only release pro-inflammatory cytokines but also secrete fibrogenic factors such as transforming growth factor-beta (TGF-β) and platelet-derived growth factor (PDGF), contributing to tissue remodeling and fibrosis [31]. The activation of these factors recruits and activates HSCs, leading to the deposition of extracellular matrix components, including collagen, thereby exacerbating fibrosis. Additionally, the activation of neutrophils and T cells further amplifies the local inflammatory response, driving hepatocellular injury and accelerating disease progression [32].
The role of inflammation-derived indices in cirrhosis and fatty liver disease has garnered increasing attention, given the strong link between these diseases and systemic inflammation. By analyzing inflammation-derived indices from CBC data, it is possible to assess the onset, progression, and prognosis of liver diseases [33]. SIRI, an inflammation marker derived from the combined counts of neutrophils, lymphocytes, and platelets in CBC, has emerged as a comprehensive index reflecting systemic inflammation levels [34].
SIRI is a composite biomarker derived from peripheral neutrophil, monocyte, and lymphocyte counts, offering a more comprehensive reflection of systemic inflammatory status. It holds distinct diagnostic value, particularly in early disease stages where conventional markers may fail to capture the full extent of immune activation. By integrating multiple leukocyte subtypes, SIRI enables more precise assessment and facilitates risk stratification. Although historically less studied than the widely used SII, SIRI has recently gained recognition for its potential in prognostic evaluation across a range of diseases. In patients with advanced heart failure, SIRI outperformed SII in predicting adverse outcomes [35]. Similarly, in cerebrovascular disease, SIRI demonstrated a unique association with stroke risk [36]. Moreover, elevated SIRI levels have been linked to increased cardiovascular mortality in individuals with chronic kidney disease [37]. In the present study, SIRI showed a more robust and comprehensive association with both liver stiffness and hepatic steatosis compared to SII, further underscoring its potential as a sensitive marker of inflammation-related hepatic alterations.Moreover, SIRI has demonstrated potential in assisting the diagnosis of severe pancreatitis, predicting the prognosis of bacterial peritonitis, and even forecasting outcomes in pancreatic and lung cancers [38, 39].Thus, SIRI holds great promise as an early diagnostic tool, as well as a means for evaluating disease activity and predicting prognosis across various conditions.
In our analysis, we systematically investigated the potential associations between a range of inflammation-derived indices and two non-invasive markers of liver status, while rigorously adjusting for multiple covariates to ensure the robustness and reliability of our findings. Initial linear and non-linear regression analyses identified SIRI as a particularly influential factor in relation to both hepatic outcomes. Notably, HS-CRP exhibited strong correlations with both SIRI and SII, a finding that is readily interpretable given its established role as a conventional marker of systemic inflammation. To mitigate the risk of multicollinearity in subsequent analyses, HS-CRP was excluded from the mediation models. We then employed structural equation modeling (SEM) to disentangle the direct and indirect effects of SIRI on liver parameters. The results revealed a complex network of associations between SIRI and various biological variables, offering valuable insights into early prevention and therapeutic strategies in clinical settings. Notably, our findings indicate that SIRI mediates associations between several baseline characteristics—including age, BMI, hemoglobin, and platelet count—and clinical outcomes. Age-dependent progression of inflammaging, characterized not by diminished acute inflammation but by a chronic, low-grade systemic inflammatory state, underlies the pathogenesis of multiple age-related disorders [40]. Similarly, elevated BMI promotes a metabolically driven, low-grade chronic inflammation central to obesity-related comorbidities [41]. The relationship between hemoglobin and inflammation exhibits bidirectional complexity: inflammatory states induce hemoglobin depletion (inflammation-induced anemia), while aberrant hemoglobin levels may reciprocally modulate inflammatory cascades [42]. Crucially, our results reveal significant positive correlations among hemoglobin levels, systemic inflammation, and fatty liver disease.
Throughout the COVID-19 pandemic, dynamic fluctuations in systemic inflammation have shaped both disease pathophysiology and the datasets derived from this period. Acute SARS-CoV-2 infection elicits a potent immune response, often culminating in a cytokine storm in severe cases. The uncontrolled release of pro-inflammatory mediators such as IL-6, IL-1β, and TNF-α can drive extensive tissue injury, precipitating acute respiratory distress syndrome (ARDS) and multiorgan failure [43].Beyond the acute phase, a subset of recovered individuals exhibit prolonged immune dysregulation, characterized by persistent low-grade inflammation, autoimmune manifestations, and endothelial dysfunction with a heightened thrombotic risk [44, 45]. As the pandemic subsides, longitudinal assessment of post-recovery inflammatory dynamics and hepatic function may provide critical insights into the long-term immunopathological sequelae of COVID-19.
Although NHANES was not specifically designed to address COVID-19, the most recent cycles include health examination data collected during the pandemic, providing a valuable resource for assessing its indirect impact on population health. While NHANES currently lacks variables directly related to SARS-CoV-2 infection, the data collected under the influence of the pandemic remains of substantial research value. It offers high-quality support for investigating the inflammation–liver axis and serves as a powerful platform for understanding and addressing health conditions that emerged or were exacerbated during this period. Longitudinal comparisons of health indicators across pre-, mid-, and post-pandemic cycles may further elucidate the pandemic’s effects on nutrition, metabolism, systemic inflammation, and liver function, and represent a critical direction for future validation studies.
Certainly, this study does have some limitations: [1] Despite NHANES using a complex stratified sampling method to represent the U.S. population, sampling bias may still exist. Certain populations, such as the homeless or institutionalized individuals, may be underrepresented, which could limit the generalizability of the findings; [2] NHANES data is primarily cross-sectional, providing a snapshot of the population at a single point in time. While associations between variables can be identified, causality cannot be established, nor can changes over time be examined; [3] Despite efforts to adjust for confounding variables, NHANES data may still be subject to unmeasured or residual confounding. This is particularly relevant when studying the relationship between health behaviors, nutrition, and chronic diseases, where many factors are interrelated [4]. Notably, while LUTE has been validated as a non-invasive assessment tool and is widely applied in clinical practice, it should not serve as the sole basis for clinical diagnosis. Its interpretation must be integrated with other diagnostic evaluations to ensure comprehensive clinical decision-making [5]. The findings of this study underscore the impact of inflammation on liver health and highlight a compelling association between SIRI levels and hepatic status. However, it is important to recognize that COVID-19 may elevate systemic inflammation across populations, potentially influencing the observed associations. Further investigations are warranted to elucidate the broader health implications during the COVID-19 pandemic.
In this study, we conducted our analysis based on complete data, excluding all cases with missing values. Our primary focus was to explore potential associations between clinical findings within this specific subset, rather than the broader U.S. population. We believe that when the study population is targeted at a specific subgroup, unweighted analysis may be more appropriate, as weighting adjustments could introduce unnecessary complexity.
Conclusion
In summary, our data analysis yielded several findings with potential clinical relevance. Inflammation-derived indices were significantly associated with hepatic fibrosis and steatosis, supporting their role as contributory factors in liver pathology. Among these markers, SIRI demonstrated superior and more consistent associations compared to SII and MLR, showing strong correlations with both non-invasive hepatic biomarkers and a nonlinear positive relationship with CAP. Furthermore, we explored key variables closely linked to SIRI and conducted mediation analyses, revealing its intermediary role in the observed associations. Future studies should aim to validate these findings in larger, population-based prospective cohorts to further elucidate the clinical utility and mechanistic significance of inflammation-based indices such as SIRI in liver disease.
Supplementary Information
Supplementary Material 1: Supplementary file 1. Data filtering process.
Supplementary Material 2: Supplementary file 2. Baseline demographic characteristics stratified by MLS.
Supplementary Material 3: Supplementary file 3. Baseline demographic characteristics stratified by CAP.
Supplementary Material 4: Supplementary file 4. Supplementary data regarding the associations between selected covariates in Models 2 and 3 and the six inflammation-derived indices treated as continuous variables.
Supplementary Material 5: Supplementary file 5. Supplementary data on the associations between selected covariates in Models 2 and 3 and the six inflammation-derived indices analyzed as categorical variables.
Authors’ contributions
All authors contributed to the preparation of the manuscript. Li and Luo primarily drafted the manuscript, while He provided overall direction and guidance.
Funding
No financial support or funding was received for this study.
Data availability
The availability and access information regarding the data and materials used in this study can be requested from the corresponding author of this article upon request.
Declarations
Ethics approval and consent to participate
This study only uses data from public databases and does not contain any identifiable personal information.All data can be obtained on the official website(https://www.cdc.gov/nchs/nhanes).
Consent for publication
Not Applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jinhao Li and Yunzhao Luo contributed equally to this work.
References
- 1.Adrià J, Elisa P, Pere GJMC. Liver Cirrhosis: ancient disease, new challenge. 2024;164(5).
- 2. Somnay K, Wadgaonkar P, Sridhar N, Roshni P, Rao N, Wadgaonkar R. Liver Fibrosis Leading to Cirrhosis: Basic Mechanisms and Clinical Perspectives. Biomedicines. 2024;12(10). 10.3390/biomedicines12102229 [DOI] [PMC free article] [PubMed]
- 4.Engin A. Nonalcoholic Fatty Liver Disease and Staging of Hepatic Fibrosis. Advances in experimental medicine and biology. 2024;1460:539–74. [DOI] [PubMed]
- 5.Bawiskar N, Acharya S, Kumar S. Fatty liver disease - non alcoholic to metabolic - A transition of concepts!! Journal of family medicine and primary care. 2024;13(8):2857-62. [DOI] [PMC free article] [PubMed]
- 6.Li B, Zhang C, Zhan YT. Nonalcoholic Fatty Liver Disease Cirrhosis: A Review of Its Epidemiology, Risk Factors, Clinical Presentation, Diagnosis, Management, and Prognosis. Canadian journal of gastroenterology & hepatology. 2018:2784537. [DOI] [PMC free article] [PubMed]
- 7.Huang YH, Chan C, Lee HW, Huang C, Chen YJ, Liu PC, et al. Influence of Nonalcoholic Fatty Liver Disease With Increased Liver Enzyme Levels on the Risk of Cirrhosis and Hepatocellular Carcinoma. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association. 2023;21(4):960–9. [DOI] [PMC free article] [PubMed]
- 8.Simbrunner B, Reiberger T, Mandorfer M. Author's reply: The relationship between systemic inflammation and liver fibrogenesis in patients with cirrhosis. Liver international : official journal of the International Association for the Study of the Liver. 2023;43(2):520-1. [DOI] [PubMed]
- 9.Dirchwolf M, Ruf AE. Role of systemic inflammation in cirrhosis: From pathogenesis to prognosis. World journal of hepatology. 2015;7(16):1974-81. [DOI] [PMC free article] [PubMed]
- 10.Ward PC. The CBC at the turn of the millennium: an overview. Clinical chemistry. 2000;46(8 Pt 2):1215-20. [PubMed]
- 11.Taru V, Szabo G, Mehal W, Reiberger T. Inflammasomes in chronic liver disease: Hepatic injury, fibrosis progression and systemic inflammation. Journal of hepatology. 2024;81(5):895-910. [DOI] [PMC free article] [PubMed]
- 12.Liu K, Tang S, Liu C, Ma J, Cao X, Yang X, et al. Systemic immune-inflammatory biomarkers (SII, NLR, PLR and LMR) linked to non-alcoholic fatty liver disease risk. Frontiers in immunology. 2024;15:1337241. [DOI] [PMC free article] [PubMed]
- 14.Duan T, Xing C, Chu J, Deng X, Du Y, Liu X, et al. ACE2-dependent and -independent SARS-CoV-2 entries dictate viral replication and inflammatory response during infection. Nat Cell Biol. 2024;26(4):628–44. [DOI] [PubMed] [Google Scholar]
- 15.Zhang C, Shi L, Wang FS. Liver injury in COVID-19: management and challenges. Lancet Gastroenterol Hepatol. 2020;5(5):428–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Marjot T, Moon AM, Cook JA, Abd-Elsalam S, Aloman C, Armstrong MJ, et al. Outcomes following SARS-CoV-2 infection in patients with chronic liver disease: an international registry study. J Hepatol. 2021;74(3):567–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhang GL, Zhao QY, Lin CS, Hu ZX, Zhang T, Gao ZL. Transient Elastography and Ultrasonography: Optimal Evaluation of Liver Fibrosis and Cirrhosis in Patients with Chronic Hepatitis B Concurrent with Nonalcoholic Fatty Liver Disease. BioMed research international. 2019;2019:3951574. [DOI] [PMC free article] [PubMed]
- 18.Barr RG, Ferraioli G, Palmeri ML, Goodman ZD, Garcia-Tsao G, Rubin J, et al. Elastography Assessment of Liver Fibrosis: Society of Radiologists in Ultrasound Consensus Conference Statement. Radiology. 2015;276(3):845–61. [DOI] [PubMed] [Google Scholar]
- 19.Castéra L, Vergniol J, Foucher J, Le Bail B, Chanteloup E, Haaser M, et al. Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C. Gastroenterology. 2005;128(2):343–50. [DOI] [PubMed] [Google Scholar]
- 20.Karlas T, Petroff D, Sasso M, Fan JG, Mi YQ, de Lédinghen V, et al. Individual patient data meta-analysis of controlled Attenuation parameter (CAP) technology for assessing steatosis. J Hepatol. 2017;66(5):1022–30. [DOI] [PubMed] [Google Scholar]
- 22.Bernardi M, Moreau R, Angeli P, Schnabl B, Arroyo V. Mechanisms of decompensation and organ failure in cirrhosis: From peripheral arterial vasodilation to systemic inflammation hypothesis. Journal of hepatology. 2015;63(5):1272-84. [DOI] [PubMed]
- 23.Engelmann C, Clària J, Szabo G, Bosch J, Bernardi M. Pathophysiology of decompensated cirrhosis: Portal hypertension, circulatory dysfunction, inflammation, metabolism and mitochondrial dysfunction. Journal of hepatology. 2021;75 Suppl 1(Suppl 1):S49-s66. [DOI] [PMC free article] [PubMed]
- 24.Ramakrishna G, Rastogi A, Trehanpati N, Sen B, Khosla R, Sarin SK. From cirrhosis to hepatocellular carcinoma: new molecular insights on inflammation and cellular senescence. Liver cancer. 2013;2(3-4):367-83. [DOI] [PMC free article] [PubMed]
- 25.Yan M, Cui Y, Xiang Q. Metabolism of hepatic stellate cells in chronic liver diseases: emerging molecular and therapeutic interventions. Theranostics. 2025;15(5):1715-40. [DOI] [PMC free article] [PubMed]
- 26.Efremova I, Maslennikov R, Kudryavtseva A, Avdeeva A, Krasnov G, Diatroptov M, et al. Gut Microbiota and Cytokine Profile in Cirrhosis. Journal of clinical and translational hepatology. 2024;12(8):689-700. [DOI] [PMC free article] [PubMed]
- 28.Lebossé F, Gudd C, Tunc E, Singanayagam A, Nathwani R, Triantafyllou E, et al. CD8(+)T cells from patients with cirrhosis display a phenotype that may contribute to cirrhosis-associated immune dysfunction. EBioMedicine. 2019;49:258-68. [DOI] [PMC free article] [PubMed]
- 29.Tazi KA, Quioc JJ, Saada V, Bezeaud A, Lebrec D, Moreau R. Upregulation of TNF-alpha production signaling pathways in monocytes from patients with advanced cirrhosis: possible role of Akt and IRAK-M. Journal of hepatology. 2006;45(2):280-9. [DOI] [PubMed]
- 30.Khan H, Siddiqui MS. Hepatic inflammation: an important target for biomarker development in nonalcoholic fatty liver disease. Hepatobiliary surgery and nutrition. 2023;12(3):447-9. [DOI] [PMC free article] [PubMed]
- 31.Wiering L, Tacke F. Treating inflammation to combat non-alcoholic fatty liver disease. The Journal of endocrinology. 2023;256(1). 10.1530/JOE-22-0194 [DOI] [PubMed]
- 32.Peiseler M, Schwabe R, Hampe J, Kubes P, Heikenwälder M, Tacke F. Immune mechanisms linking metabolic injury to inflammation and fibrosis in fatty liver disease - novel insights into cellular communication circuits. Journal of hepatology. 2022;77(4):1136-60. [DOI] [PubMed]
- 33.Shen D, Cai X, Hu J, Song S, Zhu Q, Ma H, et al. Inflammatory Indices and MAFLD Prevalence in Hypertensive Patients: A Large-Scale Cross-Sectional Analysis from China. Journal of inflammation research. 2025;18:1623-38. [DOI] [PMC free article] [PubMed]
- 34.Zhang Z, Zhang W, Liu Z, Ou J, Sun Y, Zhang L, et al. Association between systemic inflammation markers and cardiovascular mortality in adults with metabolic dysfunction-associated steatotic liver disease. Nutrition, metabolism, and cardiovascular diseases : NMCD. 2025;35(5):103781. [DOI] [PubMed]
- 35.Wang X, Zhang S, Wang X, Shen X, Huang L. Systemic inflammatory response index is a more promising prognostic index than systemic immune inflammation index in critically ill heart failure patients: A retrospective cohort analysis of the MIMIC-IV database. Vasc Health Risk Manag. 2025;21:345–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Valibeygi A, Fardaei M, Niknejad S. Association between stroke and systemic inflammation response index (SIRI): a National health and nutrition examination survey (NHANES) study 2015–2020. BMJ Neurol Open. 2025;7(1):e000718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Gu L, Xia Z, Qing B, Wang W, Chen H, Wang J, et al. Systemic inflammatory response index (SIRI) is associated with all-cause mortality and cardiovascular mortality in population with chronic kidney disease: evidence from NHANES (2001–2018). Front Immunol. 2024;15:1338025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Efgan MG, Acar H, Kanter E, Kırık S, Duman Şahan T. Role of Systemic Immune Inflammation Index, Systemic Immune Response Index, Neutrophil Lymphocyte Ratio and Platelet Lymphocyte Ratio in Predicting Peritoneal Culture Positivity and Prognosis in Cases of Spontaneous Bacterial Peritonitis Admitted to the Emergency Department. Medicina (Kaunas, Lithuania). 2024;60(8). 10.3390/medicina60081335 [DOI] [PMC free article] [PubMed]
- 39.Pacheco-Barcia V, Custodio-Cabello S, Carrasco-Valero F, Palka-Kotlowska M, Mariño-Mendez A, Carmona-Bayonas A, et al. Systemic Inflammation Response Index and weight loss as prognostic factors in metastatic pancreatic cancer: A concept study from the PANTHEIA-SEOM trial. World journal of gastrointestinal oncology. 2024;16(2):386-97. [DOI] [PMC free article] [PubMed]
- 40.López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013;153(6):1194–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Hotamisligil GS. Inflammation and metabolic disorders. Nature. 2006;444(7121):860–7. [DOI] [PubMed] [Google Scholar]
- 42.Ganz T. Anemia of inflammation. N Engl J Med. 2019;381(12):1148–57. [DOI] [PubMed] [Google Scholar]
- 43.Tay MZ, Poh CM, Rénia L, MacAry PA, Ng LFP. The trinity of COVID-19: immunity, inflammation and intervention. Nature reviews Immunology. 2020;20(6):363-74. [DOI] [PMC free article] [PubMed]
- 44.Yanhui L, Jiangang W, Jianping G, Tao W. Post-COVID-19 syndrome and low-grade inflammation: exploring gender and occupational inequalities in a retrospective cohort study. International health. 2025;17(5):720-6. [DOI] [PMC free article] [PubMed]
- 45.Patrascu R, Dumitru CS. Advances in Understanding Inflammation and Tissue Damage: Markers of Persistent Sequelae in COVID-19 Patients. Journal of clinical medicine. 2025;14(5). 10.3390/jcm14051475 [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1: Supplementary file 1. Data filtering process.
Supplementary Material 2: Supplementary file 2. Baseline demographic characteristics stratified by MLS.
Supplementary Material 3: Supplementary file 3. Baseline demographic characteristics stratified by CAP.
Supplementary Material 4: Supplementary file 4. Supplementary data regarding the associations between selected covariates in Models 2 and 3 and the six inflammation-derived indices treated as continuous variables.
Supplementary Material 5: Supplementary file 5. Supplementary data on the associations between selected covariates in Models 2 and 3 and the six inflammation-derived indices analyzed as categorical variables.
Data Availability Statement
The availability and access information regarding the data and materials used in this study can be requested from the corresponding author of this article upon request.