Demographic and Clinical Characteristics in Normal and NAFLD Groups
The data presented in Table 1 summarize the clinical characteristics of the study population, divided into three groups: Control, NASH, and NAST. NASH and NAST groups had a significantly higher proportion of hyperlipidemic patients compared to the control group (χ² = 14.23, p = 0.001). The NASH and NAST groups demonstrated higher mean body mass index (BMI) values compared to the control group (p < 0.001), indicating a greater prevalence of obesity among patients with NASH and NAST. Furthermore, the presence of diabetes mellitus was more common in the NASH and NAST groups, with a higher percentage of patients being diagnosed with DM (χ² = 5.42, p = 0.067).
Age differences across the groups were not statistically significant (t = 1.02, p = 0.313), suggesting that age is not a primary differentiating factor in the progression of NAFLD. Similarly, gender distribution did not differ significantly between the groups (χ² = 0.73, p = 0.695).
Hypertension prevalence was assessed across the groups, but no significant differences were observed (χ² = 1.44, p = 0.487). However, the trend indicated a higher occurrence of hypertension in the NASH group compared to the control and NAST groups.
There were significant differences in fibrosis and inflammation activity scores between the groups, with NASH patients having more advanced fibrosis (F2-F3) and higher activity (A2-A3) compared to the control group (p < 0.001). The distribution of steatosis grades and ultrasound findings also differed significantly among the groups, with the NASH and NAST groups showing more severe liver changes. These results highlight the progressive nature of liver disease in NAFLD and underscore the importance of early diagnosis and management.
Biochemical and Hematological Parameters Across Control, NASH, and NAST Groups
Table 2 details the biochemical and hematological parameters, showing significantly higher median FBS levels in NASH and NAST groups, indicating a tendency towards hyperglycemia. Direct bilirubin and GGT levels were elevated in NASH and NAST groups, suggesting liver dysfunction. Lipid profile analysis revealed higher total cholesterol levels in NASH and NAST groups, reflecting dyslipidemia. Hematological markers such as hemoglobin and platelet count remained stable across groups.
Alterations in Bile Acid Across NAFLD Groups
Table 3 summarizes the bile acid profiles among NHC, NAST, and nonalcoholic steatohepatitis (NASH). Primary unconjugated bile acids, CA and CDCA, were significantly elevated in both NAST and NASH compared to NHC (all p < 0.05), with no significant differences between NAST and NASH (all p > 0.05). Similarly, primary conjugated bile acids, including GCA, GCDCA, TCA, and TCDCA, were significantly higher in NASH and NAST compared to NHC (all p < 0.05). GCA and TCA levels were notably higher in NASH compared to NAST (all p < 0.05), whereas GCDCA and TCDCA did not differ significantly between NAST and NASH (all p > 0.05).
Secondary bile acids, LCA, TLCA, GUDCA, and TUDCA, were significantly elevated in NASH and NAST compared to NHC (all p < 0.05), with no significant differences between NAST and NASH (all p > 0.05). DCA, GDCA, TDCA, and UDCA did not show significant differences between either NAST or NASH and NHC (all p > 0.05).
Bile Acids in Discriminating NAFLD Patients and Control
ROC curves (Figure2) and Principal Component Analysis (PCA) (Figure3) were used to examine the capacity of bile acids in discriminating healthy from NAFLD subgroups NAST and NAST. Figure 2a-b summarize the ROC curve analysis for primary and secondary bile acids in differentiating healthy controls from NAFLD patients (NAST+NASH). The primary unconjugated bile acid CDCA showed high diagnostic potential with an AUC of 0.937, 100% specificity, and 74.7% sensitivity at a cutoff of 0.002. The primary conjugated bile acid GCDCA achieved an AUC of 0.877, 100% specificity, and 80% sensitivity at a cutoff of 2.48.
The secondary bile acids collectively demonstrated moderate diagnostic performance. LCA had an AUC of 0.864, with 96% specificity and 74.7% sensitivity at a cutoff of 0.002, while TLCA had an AUC of 0.840, with 12% specificity and 80% sensitivity at a cutoff of 0.002. Other secondary bile acids, including TUDCA and GDCA, displayed varying degrees of diagnostic accuracy, with AUCs ranging from 0.784 to 0.420, reflecting different levels of specificity and sensitivity.
Overall, primary bile acids exhibited superior diagnostic accuracy compared to secondary bile acids, underscoring their potential utility as biomarkers for distinguishing NAFLD patients from healthy controls. In the subsequent analysis comparing NAST and NASH, the ROC curves for both primary and secondary bile acids showed limited diagnostic utility, with AUC values generally indicating weak discrimination. Primary bile acids like CDCA and CA had AUCs of 0.596 and 0.575, respectively, demonstrating moderate accuracy. Sensitivity and specificity were inconsistent, with significant variability. Secondary bile acids exhibited similarly low AUC values, ranging from 0.466 for LCA to 0.571 for TLCA, indicating limited differentiation between the conditions (Figure 2c-d). These findings, suggest that the bile acids assessed are not robust biomarkers for distinguishing NAST from NASH. PCA of these bile acids across the studied groups. PCA demonstrated that the first two principal components (PC1 and PC2) accounted for 55.82% of the total variance, with eigenvalues of 5.27 and 2.54, respectively. Key bile acids, including GCDCA, CA, and CDCA, emerged as significant contributors, underscoring their potential as biomarkers for differentiating NAFLD from the control group (Figure 3).
Impact of Gender, Diabetes Status, and Lipid Profile on Bile Acid Levels in NAST and NASH Patients
Table 4 presents the analysis of bile acid levels in NAFLD (NAST and NASH) patients across gender, diabetes status, and lipid profile, highlighting several key differences. Significant differences were observed in TDCA and TUDCA levels between genders, with females showing higher levels (p = 0.033 and p = 0.037, respectively). For diabetes status, CA, CDCA, GCA, and TCDCA levels were significantly different, with the highest levels found in the DM group (p < 0.05 for all). Hyperlipidemic patients exhibit higher significant differences in bile acids CA, LCA, GCA and GUDCA than normolipidemic with p < 0.05 for all these bile acids, suggesting that lipid profile substantially impact bile acid metabolism in NAFLD. Further correlation analysis between bile acids and lipid profile in NAST and NASH revealed several significant associations between some bile acids and lipid profile parameters (Table5).
In the NAST group, CA and CDCA had strong negative correlations with LDL (CA: r=-0.57**, p<0.01; CDCA: r=-0.53**, p<0.01) and positive correlations with HDL (CA: r=0.57**, p<0.01; CDCA: r=0.56**, p<0.01). LCA and GDCA were also negatively correlated with LDL (LCA: r=-0.69**, p<0.01; GDCA: r=-0.41*, p<0.05) and positively correlated with HDL (LCA: r=0.70**, p<0.01); GDCA: r= 0.39*, p<0.05). Conversely, other bile acids did not show significant correlations with lipid components.
In the NASH group, unconjugated bile acids, CA was positively correlated with HDL (CA: r=0.42**, p<0.01) and negatively correlated with LDL (CA: r=-0.31*, p<0.05). Primary conjugated bile acids GCA, GCDCA and TCDCA were positively correlated with cholesterol (GCA: r=0.49**, p<0.01; GCDCA: r=0.38*, p<0.05; TCDCA: r=0.31*, p<0.05). Secondary bile acids, GDCA, and GDUCA were positively correlated with cholesterol (GDCA: r=0.39*, p<0.01; GDUCA: r=0.38*, p<0.05) and HDL (GDCA: r=0.46**, p<0.01; GUDCA: r=0.50**, p<0.01) and negatively correlated with LDL (GDCA: r=-0.44**, p<0.01; GDUCA: r=-0.41*, p<0.05). TUDCA positively correlated with cholesterol (r=0.37*, p<0.05) and HDL (r= 0.38*, p<0.05). Other bile acids did not show significant correlations with any lipid components.
Bile Acid Levels Across Fibrosis, Inflammation, and Steatosis Grades in NAFLD Patients
Table 6 presents a detailed comparison of bile acid levels across various stages of fibrosis, inflammation, and steatosis in patients with either NAST or NASH.
In the context of fibrosis: CA, CDCA and GCA levels were significantly higher in the F2-F3 group compared to the F0-F1 group (all p < 0.05). No significant differences were found for, in the primary conjugated GCDCA, TCA, TCDCA, and all the secondaries uncogitated and conjugated bile acid DCA, LCA, GDCA, TDCA, TLCA, UDCA, GUDCA, and TUDCA (all p > 0.05).
In the context of inflammation GCA, TCA and TLCA levels were significantly higher in the A2-A3 group compared to A0-A1 (p < 0.05). In contrast, LCA levels were notably higher in A0-A1 compared to A2-A3 but did not reach the level of significant (p > 0.05). No significant differences were found for CDCA, GCDCA, TCDCA, DCA, GDCA, TDCA, UDCA, GUDCA, and TUDCA (all p > 0.05).
In the context of steatosis, CA, GCA, TCA, TCDCA, and TLCA levels were significantly elevated in the G2-G3 group compared to G0-G1 (CA: p = 0.003; GCA: p = 0.032). DCA levels were significantly higher in G0-G1 compared to G2-G3 (p < 0.05). No significant differences were observed for CDCA, GCDCA, LCA, GDCA, TDCA, UDCA, GUDCA, and TUDCA (all p > 0.05).
Predictive Value of Bile Acids for Fibrosis, Inflammation, and Steatosis in NAFLD: Logistic Regression Analysis
Table 7 reveals the logistic regression analysis results for bile acids as predictors of fibrosis, active inflammation, and steatosis in NAFLD patients (NAST and NASH). For fibrosis the analysis identified several bile acids with significant associations with fibrosis, CA, CDCA and DCA were notable predictors, with CA showing an odds ratio of 2.05 (p = 0.02); CDCA had an odds ratio of (Exp(B) = 1.58, p = 0.04) and DCA an odds ratio of 2.06 (p = 0.04).
In terms of active inflammation, GCA was a significant predictor (Exp(B) = 1.92, p = 0.03), along with TCA (Exp(B) = 1.94, p = 0.02, and TLCA (Exp(B) = 15.95, p = 0.03). CDCA did not show a significant association with inflammation (Exp(B) = 0.49, p = 0.66).
In the context of steatosis, 1ry bile acids CA, CDCA, and GCA, were significant predictors (CA: Exp(B) = 2.62, p = 0.048; CDCA: Exp(B) = 1.25, p = 0.017; GCA: Exp(B) = 2.92, p = 0.041. Additionally,2ry Bile acid, DCA, TDCA, TLCA, UDCA showed a significant association and predictor of steatosis DCA: Exp(B) = 3.24, p = 0.042; TDCA: Exp(B) = 4.35, p = 0.04; TLCA: Exp(B) = 20, p = 0.03); and UDCA (Exp(B) = 20, p = 0.02).
These results highlight the role of specific bile acids, such as CA, CDCA and GCA, in predicting NAFLD severity, with implications for both fibrosis and steatosis. The significant associations of DCA, TDCA, and TLCA with steatosis suggest their potential importance in hepatic fat accumulation. Additionally, the significant correlations of GCA, TCA, and TLCA with inflammation suggest their potential utility in predicting liver inflammation activity.