作者: Hui-Qi Qu , Quan Li , Megan L. Grove , Yang Lu , Jen-Jung Pan
DOI: 10.1016/J.ARCMED.2012.08.005
关键词:
摘要: Background and Aims Elevated alanine aminotransferase (ALT >40 IU/mL) is a marker of liver injury but provides little insight into etiology. We aimed to identify stratify risk factors associated with elevated ALT in randomly selected population high prevalence (39%), obesity (49%) diabetes (30%). Methods Two machine learning methods, the support vector (SVM) Bayesian logistic regression (BLR), were used capture community cohort 1532 adults from Cameron County Hispanic Cohort (CCHC). A total 28 predictor variables prediction models. The recently identified genetic rs738409 on PNPLA3 gene was genotyped using Sequenom iPLEX assay. Results four major for fasting plasma insulin level resistance, increased BMI body weight, triglycerides non-HDL cholesterol, diastolic hypertension. In spite highly significant association females, role model minimal, compared other epidemiological factors. Age drug alcohol consumption not independent determinants this analysis. Conclusions most strongly are components metabolic syndrome point nonalcoholic fatty disease (NAFLD). This population-based identifies likely cause without requirement individual pathological diagnosis diseases. Use such can greatly contribute approach prevention disease.