1496例矮小症病因分析及基于IGF-1水平生长激素缺乏症诊断预测模型的建立  

Analyzing the etiology of 1 496 children with short stature and establishing the predict model of growth hormone deficiency using IGF-1 levels

作　　者：李晓静[1] 罗飞宏[1] 裴舟[1] 奚立[1] 叶蓉[1] 赵诸慧[1] 陆炜[1] 章淼滢[1] 郑章乾[1] 

机构地区：[1]复旦大学附属儿科医院内分泌遗传代谢科,上海201102

出　　处：《临床儿科杂志》2012年第12期1110-1115,共6页Journal of Clinical Pediatrics

摘　　要：目的探讨矮小症患儿的病因及胰岛素样生长因子(IGF)-1与生长激素(growth hormone,GH)水平之间的关系,建立基于IGF-1水平的简易GH缺乏(GHD)诊断预测模型。方法矮小症住院患儿1 496例,采用胰岛素低血糖法和精氨酸法测定GH分泌状态,根据体格检查及实验室检查分析病因;Logistic逐步多元回归模型建立基于IGF-1的GHD诊断预测模型。结果 GHD659例(44.05%),特发性矮小504例(33.69%),家族性矮小165例(11.03%),体质性青春期发育延迟35例(2.33%),余为甲状腺功能减低症、宫内发育迟缓、Turner综合征、多种垂体激素缺乏症等引起的矮小症。比较GHD与非GHD两组患儿潜在的影响因素,体质量、BMI、身高-SDS、ALT/AST/AKP、TG/Tch、IGF-1、IGFBP-3等的差异有统计学意义(P<0.05)。设GHD的概率为P,经Logistic逐步多元回归模型拟合如下:LN[P/(1-P)]=-2.0193+0.0683×年龄+0.1439×BMI+0.021×ALT-0.0021×IGF-1-0.1526×IGFBP-3。结论内分泌疾病是矮小症最多见的病因,但GH激发试验的水平与多种体格和生化指标有关,因此GHD诊断需要综合考虑;基于IGF-1拟合的简易模型对GHD诊断有较准确的预测价值,可用于门诊筛查。Objective To analysis the etiology of 1 496 short stature children and to establish the predict model of growth hormone deficiency using IGF-1 levels. Methods Anthropological and physical examination data were collected in 1 496 short children. Growth hormone levels were determined via Insulin tolerance test and Arginine test. A prediction model of growth hormone deficiency （GHD） was established by Logistic stepwise multivariate regression based on IGF-1 level. Results 659 cases （44.05%） were diagnosed as GHD, 504 cases （33.69%） were idiopathic short stature （ISS）, 165 cases （11.03%） were familial short stature, 35 cases （2.33%） were constitutional delay of growth and puberty. The rests were attributed to other rea sons, such as hypothyroidism, intrauterine growth retardation, Turner syndrome, multiple pituitary hormone deficiency and so on. There existed significantly statistical difference in body weight, BMI, height SDS, ALT/AST/AKP, TG/Tch, IGF-1, IGFBP-3 level between GHD and non-GHD groups. If the probability for growth hormone deficiency P, the Logistic stepwise multivari- ate regression model fitting model is as follows： LN[P/（1-p）] = -2.0193 ＋ 0.0683 × age ＋ 0.1439 ×BMI ＋ 0.021 × ALT-0.0021 ×IGF-1-0.152 6 × IGFBP-3. Conclusions Endocrine disease is the most common cause of dwarfism, which is most commonly seen in GHD; GH levels were affected by multiple factors including anthropological indexes and biochemical status, it is high ly suggests that serious cautions should be taken when diagnosing GHD. GHD can be briefly and accurately predicted based onIGF-1 value.

关 键 词：矮小症 生长激素缺乏症 病因 预测模型 儿童 

分 类 号：R725.8[医药卫生—儿科]