【摘要】 目的 识别非酒精性脂肪性肝病发病的影响因素，并建立其发病风险的预测模型。方法 22 702例受试者按7∶3被随机分配到训练集（n=15 892）和验证集（n=6 810）。采用Lasso-Cox回归构建预测非酒精性脂肪性肝病发病的列线图。采用受试者操作特征曲线下面积评估模型的区分度。结果 体重指数、高血压、糖尿病、高尿酸血症、丙氨酸转氨酶、甘油三酯对非酒精性脂肪性肝病发病的影响有统计学意义并构建列线图。在训练集中，2、3、4年曲线下面积分别为0.72、0.73、0.73。在验证集中，2、3、4年曲线下面积分别为0.72、0.74、0.76。结论 本研究建立的列线图模型对4年内非酒精性脂肪性肝病的发病率具有较好的预测价值。更多还原

【Abstract】 Objective To identify risk factors with the onset of nonalcoholic fatty liver disease and develop a prediction model for its incidence. Methods 22 702 participants were randomly assigned to the training and the validation set at 70%(n =15 892) and 30%(n = 6 810). A nomogram model for nonalcoholic fatty liver disease was constructed using Lasso-Cox regression. The area under the receiver operator characteristic curve(AUC) was used to assess model discriminability. Results body mass index, hypertension, diabetes mellitus,hyperuricemia, alanine aminotransferase and triglycerides were statistically significant for the onset of nonalcoholic fatty liver disease and were used to construct the nomogram. In the training set, AUC was 0.72, 0.73and 0.73 at 2, 3, 4 years, respectively. In the validation set, AUC was 0.72, 0.74 and 0.76 at 2, 3, 4 years, respectively. Conclusions The nomogram had a good predictive value for the incidence of nonalcoholic fatty liver disease within 4 years.更多还原