A model to predict the onset of non-alcoholic fatty liver disease within 2 years in elderly adults

Diagnostic accuracy of blood biomarkers and non-invasive scores for the diagnosis of NAFLD and NASH: Systematic review and meta-analysis

INTRODUCTION AND OBJECTIVES

Fatty liver disease is an important public health problem. Early diagnosis is critical to lower its rate of progression to irreversible/terminal stages. This study aimed to evaluate the accuracy of non-invasive prediction scores for fatty liver disease (NAFLD and NASH) diagnosis in adults.

MATERIALS AND METHODS

A search was conducted in 10 databases, a qualitative synthesis of 45 studies, and quantitative analysis of the six most common scores. There were 23 risk scores found for NAFLD diagnosis and 32 for NASH diagnosis. The most used were Fatty Liver Index (FLI), aspartate aminotransferase (AST) to Platelet Ratio Index, Fibrosis-4 Index (FIB-4), AST/alanine aminotransferase (ALT) ratio, BARD score, and NAFLD fibrosis score (NFS).

RESULTS

The results from the meta-analysis for FLI: Area under the curve (AUC) of 0.76 (95% Confidence Interval [CI] 0.73, 0.80), sensitivity 0.67 (CI 95% 0.62, 0.72) and specificity 0.78 (CI 95% 0.74, 0.83). The AST to Platelet Ratio Index: AUC 0.83 (CI 95% 0.80, 0.86), sensitivity 0.45 (95% CI 0.29, 0.62), and specificity of 0.89 (95% CI 0.83, 0.92). The NFS: AUC of 0.82 (CI 95% 0.78, 0.85), sensitivity 0.30 (CI 95% 0.27, 0.33) and specificity 0.96 (CI 95% 0.95,0.96).

CONCLUSIONS

The FLI for NAFLD and AST to Platelet Ratio Index for NASH were the risk scores with the highest prognostic value in the included studies. Further research is needed for the application of new diagnostic risk scores for NAFLD and NASH.

Adiposity, metabolomic biomarkers, and risk of nonalcoholic fatty liver disease: a case-cohort study

BACKGROUND

Globally, the burden of obesity and associated nonalcoholic fatty liver disease (NAFLD) are rising, but little is known about the role that circulating metabolomic biomarkers play in mediating their association.

OBJECTIVES

We aimed to examine the observational and genetic associations of adiposity with metabolomic biomarkers and the observational associations of metabolomic biomarkers with incident NAFLD.

METHODS

A case-subcohort study within the prospective China Kadoorie Biobank included 176 NAFLD cases and 180 subcohort individuals and measured 1208 metabolites in stored baseline plasma using a Metabolon assay. In the subcohort the observational and genetic associations of BMI with biomarkers were assessed using linear regression, with adjustment for multiple testing. Cox regression was used to estimate adjusted HRs for NAFLD associated with biomarkers.

RESULTS

In observational analyses, BMI (kg/m2; mean: 23.9 in the subcohort) was associated with 199 metabolites at a 5% false discovery rate. The effects of genetically elevated BMI with specific metabolites were directionally consistent with the observational associations. Overall, 35 metabolites were associated with NAFLD risk, of which 15 were also associated with BMI, including glutamate (HR per 1-SD higher metabolite: 1.95; 95% CI: 1.48, 2.56), cysteine-glutathione disulfide (0.44; 0.31, 0.62), diaclyglycerol (C32:1) (1.71; 1.24, 2.35), behenoyl dihydrosphingomyelin (C40:0) (1.92; 1.42, 2.59), butyrylcarnitine (C4) (1.91; 1.38, 2.35), 2-hydroxybehenate (1.81; 1.34, 2.45), and 4-cholesten-3-one (1.79; 1.27, 2.54). The discriminatory performance of known risk factors was increased when 28 metabolites were also considered simultaneously in the model (weighted C-statistic: 0.84 to 0.90; P  < 0.001).

CONCLUSIONS

Among relatively lean Chinese adults, a range of metabolomic biomarkers are associated with NAFLD risk and these biomarkers may lie on the pathway between adiposity and NAFLD.

Nine Markers to Predict Nonalcoholic Fatty Liver Disease for a Chinese Diabetic Population

Background: Nonalcoholic fatty liver disease (NAFLD) increases the possibility to suffer from liver or cardiovascular disease. Although hepatic biopsy is well acknowledged as the standard diagnosis, it is difficult to implement because of its intrusiveness and cost concerns. Moreover, overweight people or diabetic patients are always NAFLD-positive, but not absolute. Therefore, to distinguish whether a diabetic case has NAFLD via nonintrusive indicators is of great significance for further interventions. Objective: With 8499 diabetic patients hosted by Shanghai Sixth People’s Hospital, we try to rank the impacts of multiple routine indicators (features) on NAFLD, and further predict NAFLD within this diabetic population. Methods: We first rank dozens of related features according to their contributions in NAFLD prediction, and then we prune several trivial features to simplify the prediction. Additionally, three classification algorithms are considered and compared, e.g., C4.5, Naïve Bayes and Random Forest. Results: The experiment shows that Random Forest outperforms the rest (accuracy 85.1%, recall 90.98% and AUC 0.631). Conclusions: We find that the top nine markers together can effectively tell NAFLD out of this diabetic population. They are triglyceride (TG), low density lipoprotein (LDL), insulin (INS), hbA1C, high-density lipoprotein (HDL), fasting plasma glucose (FPG), age, total cholesterol (TC) and duration.