Nonalcoholic fatty liver disease (NAFLD) is rarely detected at early stage, and often found when merged with metabolic disorders. Currently, the predictive value of noninvasive indicators for NAFLD with metabolic syndrome (MS) has been rarely reported.

To assess the predictive value of monocyte to high-density lipoprotein cholesterol ratio (MHR) for NAFLD with MS in community residents, and to evaluate the associated factors using multivariate regression tree analysis.

From June to December 2020, among 9 812 permanent community residents (>45 years) who underwent physical examination in Songjiang District Sijing Community Health Center, 5 727 eligible individuals were selected, including 4 652 with MS, diabetes, hypertension and dyslipidemia and other metabolic disorders〔1 948 with no NAFLD and MS (controls) , 1 248 with NAFLD (NAFLD cases) , 1 456 with both NAFLD and MS (MAFLD cases) 〕, and 1 075 without metabolic disorders (healthy controls) . General data, anthropometric parameters, blood pressure, biochemical parameters and routine blood test results of all participants were collected. Spearman rank correlation analysis was performed to assess the correlation of MHR and lipid accumulation product (LAP) index with NAFLD and MS. ROC analysis was conducted to analyze the optimal cutoff value of noninvasive indicators〔including neutrophil to lymphocyte ratio (NLR) , platelet to lymphocyte ratio (PLR) , lymphocyte to monocyte ratio (LMR) and MHR〕for NAFLD with MS. Factors associated with the predictive value of above-mentioned noninvasive indicators for NAFLD with MS were identified using multivariate regression tree analysis.

The prevalence rates for metabolic disorders and NAFLD in this group were 76.61% (7 517/9 812) and 27.59% (2 704/9 812) , respectively. The controls, NAFLD cases, MAFLD cases, and healthy controls had significant differences in female ratio, mean age, BMI, waistline, hipline, waist-to-hip ratio, systolic and diastolic blood pressure, fasting blood glucose, total cholesterol, triacylglycerol, low-density lipoprotein cholesterol, HDL-C, total bilirubin, serum uric acid, ALT, AST, ALT/AST ratio, GGT, eGFR, white blood cell count, neutrophil count, monocyte count, lymphocyte count, platelet count, NLR, PLR, LMR, MHR, alcoholic fatty liver disease to NAFLD index, LAP index, and atherogenic index of plasma, and distribution of waistline and BMI, as well as prevalence of hypertension and dyslipidemia (P<0.01) . Spearman rank correlation analysis showed that, MHR was positively correlated with the grouping (r_s=0.342, P<0.001) ; LAP index was also positively correlated with the grouping (r_s=0.580, P<0.001) . ROC analysis found that, in the prediction of NAFLD with MS, the AUC of NLR was 0.528〔95%CI (0.511, 0.545) 〕with 74.66% sensitivity, and 31.52% specificity when the optimal cutoff value was chosen as 2.192, that of PLR was 0.581〔95%CI (0.564, 0.598) 〕with 59.82% sensitivity, and 53.95% specificity when the optimal cut-off value was chosen as 115.470, and that of LMR was 0.546〔95%CI (0.529, 0.563) 〕with 51.79% sensitivity, and 56.62% specificity when the optimal cut-off value was chosen as 0.193, and that of MHR was 0.695〔95%CI (0.679, 0.711) 〕with 69.51% sensitivity, and 60.63% specificity when the optimal cut-off value was chosen as 0.292. Multivariate regression tree analysis revealed that sex, BMI, waistline, triacylglycerol, and fasting blood glucose were associated with the predictive value of MHR for NAFLD with MS.

The value of MHR for predicting NAFLD with MS in physical examinees in the community is relatively high, but it may be significantly influenced by sex, BMI, waistline, triacylglycerol and fasting blood glucose.

非酒精性脂肪性肝病（NAFLD）早期难以发现，常在出现代谢性疾病时才能被发现，但目前关于非侵入性指标对NAFLD合并代谢综合征预测价值的研究报道较少见。

探讨非侵入性指标单核细胞与高密度脂蛋白胆固醇比值（MHR）对NAFLD合并代谢综合征的预测价值，并进行多因素Logistic决策树分析。

2020年6—12月，共有9 812例年龄>45周岁的社区常住居民在上海市松江区泗泾镇社区卫生服务中心进行健康体检，其中符合纳入与排除标准的目标代谢性疾病（包括代谢综合征、糖尿病、高血压、血脂异常等）患者4 652例，无代谢性疾病患者1 075例（健康对照组）；4 652例目标代谢性疾病患者中无NAFLD及代谢综合征者1 948例（对照组），仅有NAFLD者1 248例（NAFLD组），NAFLD合并代谢综合征者1 456例（MAFLD组）。比较四组受试者一般资料、人体测量学指标、血压、生化指标及血常规检查结果；MHR、脂质蓄积指数（LAP）与分组的相关性分析采用Spearman秩相关分析。绘制受试者工作特征（ROC）曲线以确定非侵入性指标中性粒细胞与淋巴细胞比值（NLR）、血小板与淋巴细胞比值（PLR）、淋巴细胞与单核细胞比值（LMR）、MHR预测社区居民NAFLD合并代谢综合征的最佳截断值，并进行多因素Logistic决策树分析。

本研究在9 812例进行健康体检的年龄>45周岁的社区常住居民中共发现目标代谢性疾病患者7 517例、NAFLD患者2 704例，目标代谢性疾病、NAFLD患病率分别为76.61%（7 517/9 812）、27.59%（2 704/9 812）。四组受试者女性比例、年龄、体质指数（BMI）、腰围、臀围、腰臀比、收缩压、舒张压、空腹血糖、总胆固醇、三酰甘油、低密度脂蛋白胆固醇（LDL-C）、高密度脂蛋白胆固醇（HDL-C）、总胆红素、血尿酸、丙氨酸氨基转移酶（ALT）、天冬氨酸氨基转移酶（AST）、ALT/AST、γ-谷氨酰转肽酶（GGT）、估算肾小球滤过率（eGFR）、白细胞计数、中性粒细胞计数、单核细胞计数、淋巴细胞计数、血小板计数、NLR、PLR、LMR、MHR、酒精性脂肪性肝病/非酒精性脂肪性肝病指数（ANI）、LAP、血浆致动脉硬化指数（AIP）、腰围分类、BMI分类、高血压及血脂异常发生情况比较，差异均有统计学意义（P<0.01）。Spearman秩相关分析结果显示，MHR、LAP与研究对象分组均呈正相关（r_s值分别为0.342、0.580，P<0.001）。ROC曲线分析结果显示，非侵入性指标NLR、PLR、LMR、MHR预测社区居民NAFLD合并代谢综合征的ROC曲线下面积分别为0.528〔95%CI（0.511，0.545）〕、0.581〔95%CI（0.564，0.598）〕、0.546〔95%CI（0.529，0.563）〕、0.695〔95%CI（0.679，0.711）〕，灵敏度分别为74.66%、59.82%、51.79%、69.51%，特异度分别为31.52%、53.95%、56.62%、60.63%，最佳截断值分别为2.192、115.470、0.193、0.292。多因素Logistic决策树分析结果显示，性别、BMI分类、腰围分类、三酰甘油、空腹血糖对MHR预测社区居民NAFLD合并代谢综合征有影响。

MHR对社区居民NAFLD合并代谢综合征的预测价值较高，而性别、BMI分类、腰围分类、三酰甘油、空腹血糖是影响MHR预测社区居民NAFLD合并代谢综合征准确性的重要变量。