Haentjens P, Massaad D, Reynaert H, Peeters E, Van Meerhaeghe A, Vinken S, Poppe K, Velkeniers B
Department of Experimental Surgery, Universitair Ziekenhuis Brussel, Belgium.
Acta Clin Belg. 2009 Nov-Dec;64(6):483-93. doi: 10.1179/acb.2009.084.
To determine the prevalence of liver steatosis among asymptomatic individuals attending an out-patient clinic for a problem of overweight, and to define the discriminatory value of several characteristics for predicting liver steatosis among them.
Consecutive patients Swith a body mass index (BMI) of > or =25 kg/m2 who consented to undergo liver ultrasound and blood tests were recruited for inclusion. Receiver operating characteristic (ROC) curves were generated and statistical indices of diagnostic performance and their corresponding 95% confidence intervals (95% CI) were computed. Logistic regression analyses were performed to determine whether a combination of characteristics could improve diagnostic accuracy.
We enrolled sixty-eight subjects (mean BMI, 37.5 kg/m2), of whom 39 (57.4%) had liver steatosis on ultrasound. Logistic regression analyses indicated that only 3 variables were significantly and independently correlated with liver steatosis: female gender, low adiponectin levels, and high insulin resistance index. A composite index for predicting liver steatosis was calculated by summing the risk factors of female gender, low adiponectin, and insulin resistance index (FAIR score). The accuracy of this score was determined by ROC analysis to be 0.85 (95% CI, 0.74-0.96; P < 0.001). The presence of two or more risk factors (FAIR score > or =2) had a sensitivity, specificity, positive predictive value, and negative predictive value of 77%, 91%, 92%, and 74%, respectively. The likelihood ratio for a positive result was 8.43.
Among asymptomatic overweight individuals attending an out-patient clinic, the prevalence of liver steatosis on ultrasound is 57%. Female gender, the insulin resistance index, and low adiponectin are significant and independent predictors of liver steatosis. A combination of these three factors allows sensitivity and specificity for non alcoholic fatty liver of 77% and 91%, respectively.
确定在因超重问题前往门诊就诊的无症状个体中肝脂肪变性的患病率,并明确几种特征对于预测他们肝脂肪变性的鉴别价值。
纳入连续的体重指数(BMI)≥25 kg/m²且同意接受肝脏超声检查和血液检测的患者。生成受试者工作特征(ROC)曲线,并计算诊断性能的统计指标及其相应的95%置信区间(95%CI)。进行逻辑回归分析以确定特征组合是否能提高诊断准确性。
我们招募了68名受试者(平均BMI为37.5 kg/m²),其中39名(57.4%)经超声检查存在肝脂肪变性。逻辑回归分析表明,仅有3个变量与肝脂肪变性显著且独立相关:女性性别、低脂联素水平和高胰岛素抵抗指数。通过将女性性别、低脂联素和胰岛素抵抗指数的危险因素相加计算出预测肝脂肪变性的综合指数(FAIR评分)。经ROC分析确定该评分的准确性为0.85(95%CI,0.74 - 0.96;P < 0.001)。存在两个或更多危险因素(FAIR评分≥2)时,敏感性、特异性、阳性预测值和阴性预测值分别为77%、91%、92%和74%。阳性结果的似然比为8.43。
在前往门诊就诊的无症状超重个体中,超声检查发现肝脂肪变性的患病率为57%。女性性别、胰岛素抵抗指数和低脂联素是肝脂肪变性的显著且独立预测因素。这三个因素的组合使非酒精性脂肪肝的敏感性和特异性分别达到77%和91%。