Association between the fasting blood glucose to high-density lipoprotein cholesterol ratio and non-alcoholic fatty liver disease in Chinese adults: a secondary analysis

BackgroundElevated fasting blood glucose (FBG) and decreased high-density lipoprotein cholesterol (HDL-C) are individually associated with non-alcoholic fatty liver disease (NAFLD). This study investigated whether their composite ratio (GHR) is independently associated with NAFLD and whether this association varies across subgroups.MethodsThis secondary analysis included 1,592 Chinese adults (40–70 years) from routine health checks. NAFLD was diagnosed via ultrasound. GHR was calculated as FBG divided by HDL-C. Multivariable logistic regression assessed the GHR–NAFLD association. Receiver operating characteristic (ROC) curves compared the discriminative ability of GHR with that of FBG and HDL-C alone using DeLong's test. Subgroup and interaction analyses were performed.ResultsNAFLD prevalence was 61.1%. After full adjustment, each 1-unit increase in GHR was associated with 66% higher NAFLD odds [odds ratio (OR) = 1.66, 95% confidence interval (CI): 1.02–2.69]. Across quartiles, a significant positive trend persisted (p for trend = 0.017). Compared with the lowest quartile, the fully adjusted ORs were 1.36 (p = 0.076) for quartile 2, 1.73 (95% CI: 1.16–2.60, p = 0.008) for quartile 3, and 1.74 (95% CI: 1.03–2.93, p = 0.037) for quartile 4. In exploratory subgroup analyses, a borderline interaction with diabetes was observed (P for interaction = 0.046). GHR was associated with NAFLD in non-diabetic participants (OR = 2.17, 95% CI: 1.13–4.18) but not in diabetic participants. The AUC for GHR (0.684) exceeded that of FBG (0.640, p = 0.001) and HDL-C (0.637, p < 0.001). The fully adjusted model achieved an AUC of 0.799 (95% CI: 0.777–0.821).ConclusionGHR was independently associated with NAFLD and showed better discrimination than FBG or HDL-C alone. However, adding GHR to the fully adjusted model did not significantly improve discrimination, and the possible stronger association in non-diabetic individuals requires confirmation. Its incremental value beyond established clinical risk factors is therefore limited, and GHR should not replace comprehensive risk assessment.