【摘要】 目的:构建基于血清胰岛素样生长因子-1(IGF-1)的儿童生长激素缺乏症(GHD)预测模型,并评估其在临床筛查及分层治疗中的应用价值。方法:纳入2024年7月至2025年3月中山市小榄人民医院收治的矮小症患儿200例,行系统病因学评估、生长激素(GH)激发试验及血清IGF-1检测。根据GH峰值分为GHD组与非GHD组,采用logistic回归构建多因素预测模型,并评价其诊断效能;通过交叉验证进行内部验证。结果:GHD组IGF-1水平低于非GHD组(P<0.05),IGF-1<100 ng~/m L为主要危险因素(OR=3.92,95%CI 2.17～6.67)。骨龄延迟、丙氨酸氨基转移酶(ALT)、天冬氨酸氨基转移酶(AST)、碱性磷酸酶(ALP)、游离三碘甲状腺原氨酸(FT3)、游离甲状腺素(FT4)、促甲状腺激素(TSH)等与GHD存在统计学相关性。本模型曲线下面积(AUC)为0.93,灵敏度89.2%,特异度85.0%;交叉验证AUC为0.91,提示模型具有良好稳定性。在此基础上构建评分分层方案,不同风险水平在临床干预策略上呈差异趋势。结论:基于IGF-1联合多指标构建的预测模型可提高GHD的早期识别能力,可为分层及个体化治疗提供参考,具有一定临床应用价值。更多还原

【Abstract】 Objective: To construct a diagnostic predictive model for children with growth hormone deficiency(GHD) based on serum insulin-like growth factor-1(IGF-1) and to evaluate its potential value in clinical screening and stratification treatment. Methods: A total of 200 children with short stature admitted into Xiaolan People’s Hospital of Zhongshan from Jul. 2024 to Mar. 2025 were enrolled. All participants underwent comprehensive etiological evaluation, growth hormone( GH) stimulation tests, and serum IGF-1 measurement. Patients were classified into the GHD group and non-GHD group according to GH peak levels. A multivariable prediction model was established by using logistic regression, and its diagnostic performance was assessed. Internal validation was performed by using cross-validation. Results: Serum IGF-1 levels were significantly lower in the GHD group than those in the non-GHD group(P<0. 05). IGF-1<100 ng~/m L was identified as a major risk factor for GHD(OR = 3. 92, 95% CI 2. 17 to 6. 67). Delayed bone age, alanine aminotransferase(ALT), aspartate aminotransferase( AST), alkaline phosphatase( ALKP), free triiodothyronine( FT3), free thyroxine(FT4), and thyroid stimulating hormone(TSH) showed statistical correlation with GHD. The model achieved an area under the curve(AUC) of 0. 93, with a sensitivity of 89. 2% and a specificity of 85. 0%. The cross-validation AUC was 0. 91, indicating that the model had good stability. On this basis, a stratification scheme was constructed, and different risk levels showed a trend of differences in clinical intervention strategies. Conclusion: The prediction model constructed based on IGF-1 combined with multiple indicators can enhance the early identification ability of GHD, and provide reference for stratification and individualized treatment, with certain clinical application value.更多还原