探讨正常胎儿肾脏大小与其生物学生长参数的关系，从而定量胎儿肾脏大小的Z评分实际参考范围。

随机选取2015年9月至11月在浙江省立同德医院行胎儿超声产前检查的孕妇403名，采用二维超声对403名孕龄为18~40周的胎儿进行研究。获得胎儿的孕周（GA）、双顶径（BPD）及股骨长径（FL）等正常胎儿的生物学生长参数，通过胎儿肾脏长轴切面测量胎儿左侧肾脏长径与横径。以胎儿肾脏长径与横径作为因变量，GA、BPD及FL为自变量，采用回归分析，确定Z评分最佳模型。对得到的残差（实际测量值与从方程得出的预测值之间的差异）行残差加权回归分析。以FL、BPD及GA作为自变量，分别以预测的左侧肾脏长径与横径为因变量建立胎儿肾脏的Z评分参考范围。

直线回归方程反应胎儿GA、BPD、FL与胎儿左侧肾脏长径、左侧肾脏横径之间呈显著正相关（r=0.8761、0.8818、0.8797、0.8604、0.8723、0.8643，P均<0.01）。随GA、BPD、FL的增长，左侧肾脏长径与横径也显著增长，GA、BPD、FL与左侧肾脏长径的直线回归方程分别为：Y=0.094 22＋0.1088X，Y=0.132 57＋0.4349X，Y=0.6664＋0.491 05X；GA、BPD、FL与左侧肾脏横径的直线回归方程分别为：Y=0.060 58X-0.051 98，Y=0.262 19X-0.197 55，Y=0.294 02X＋0.136 19。标准差（s）随自变量增大而增大，也呈显著正相关（r=0.925、0.934、0.915、0.908、0.914、0.922，P均<0.01）。建立直线回归模型，左侧肾脏长径的s与GA、BPD、FL直线回归方程分别为：Y=0.025＋0.005 58X；Y=0.097 09＋0.010 45X；Y=0.046 17＋0.025 65X；左侧肾脏横径的s与FL、BPD、GA的直线回归方程分别为：Y=0.001 13＋0.002 24X；Y=0.003 67＋0.011 55X；Y=0.005 07＋0.015 89X。根据以上方程求得预测值代入公式：左侧肾脏长径与横径的Z评分值=（实际测量左侧肾脏长径或横径测值-预测左侧肾脏长径或横径测值）/预测s，求得Z评分，范围在±3为正常。

采用胎儿GA、BPD、FL计算胎儿左侧肾脏长径、横径的Z评分正常参考范围，为正常发育的胎儿肾脏提供参考依据，在胎儿肾脏相关畸形的诊断方面具有潜在的应用价值。

To develop Z-scores reference ranges for kidney in normal fetuses from the measurements of gestational age (GA), biparietal diameter (BPD) or femur length (FL) using two-dimension ultrasound.

A retrospective cross-sectional study of 403 singleton normal fetuses was performed. The gestation age range was from 18 to 40 weeks. Fetal biologic parameters included GA and BPD and FL based on menstrual age. Left kidney length and left kidney width of normal fetus were measured. Normal Z-score ranges were developed for left kidney length or left kidney width using GA, BPD and FL as independent variables.

Linear regression model was the best description of the data in each case and correlation between left kidney length or left kidney width and independent variables were excellent (r=0.8761, 0.8818, 0.8797, 0.8604, 0.8723, 0.8643). All P values were less than 0.01. The equations were as follows, Y=0.094 22+ 0.1088X, Y=0.132 57+ 0.4349X, Y=0.6664+ 0.491 05X, Y=0.060 58X-0.051 98, Y=0.262 19X-0.197 55, Y=0.294 02X+ 0.136 19. Heteroscedasticity of standard deviation (s) with increasing independent variables also could be modeled with a simple linear (r=0.925, 0.934, 0.915, 0.908, 0.914, 0.922), All P values were less than 0.01. The equations were as follows, Y=0.025+ 0.005 58X, Y=0.097 09+ 0.010 45X, Y=0.046 17+ 0.025 65X, Y=0.001 13+ 0.002 24X, Y=0.003 67+ 0.011 55X, Y=0.005 07+ 0.015 89X. Then we got Z-scores based on the equation, Z-score= (actual measurement data of left kidney length or left kidney width-predictive data of left kidney length or left kidney width) predictive s.

Normal reference ranges and Z-scores for fetal left kidney length and left kidney width had been provided. These normative data may be useful tools for assessment of fetal kidney length and kidney width, especially had potential applications in malformations of fetal kidney.