非线性混合效应模型评估脑瘫患儿粗大运动功能发育进程

目的 采用非线性混合效应模型对不同级别脑瘫患儿粗大运动功能发育进程进行分析，以期为脑瘫患儿康复管理提供依据。方法 以2000年8月至2009年12月在上海7家康复机构和特殊学校接受康复诊治的脑瘫患儿为研究对象。采用中文译本的粗大运动功能测试量表（GMFM）进行粗大运动功能评估、中文版脑瘫粗大运动功能分级系统（GMFCS）进行粗大运动功能分级，分别将各GMFCS级别患儿的GMFM-66分值使用stable limit模型分析，获得各GMFCS级别患儿达到其GMFM-66的最大值以及达到最大值的速率（该值被转化为年龄-90，表示达到GMFM-66最大值90%时的年龄）。同时与加拿大相关研究（OMG）进行比较，分析不同干预背景下脑瘫患儿的粗大运动功能发育进程差异。结果 228例脑瘫患儿进入分析，其中男性152例（66.7%），女性76例（33.3%），痉挛型双瘫87例（38.2%），痉挛型四肢瘫63例（27.6%），痉挛型偏瘫48例（21.1%），痉挛型三瘫4例（1.8%），肌张力障碍型4例（1.8%），徐动型11例（4.8%），共济失调型11例（4.8%）。共有986次符合条件的GMFM-66测试结果，平均每例患儿有4.32次测试结果。首次评估时总体研究对象的平均年龄为2.95岁。GMFCS Ⅰ～Ⅴ级的GMFM-66最大值分别为81.2、62.4、52.9、40.8和24.4分，呈显著降低趋势；年龄-90分别为3.8、2.7、2.1、2.0和1.5岁，GMFCS Ⅲ和Ⅳ级间差别不大，在其余各级间表现出明显差别。本研究的GMFCSⅠ和Ⅱ级的GMFM-66最大值低于OMG，而GMFCS Ⅲ～Ⅴ级的GMFM-66最大值与之较为接近；各GMFCS级别的年龄-90均明显低于OMG。结论 本研究GMFCS Ⅰ级和Ⅱ级脑瘫患儿尽管粗大运动功能发育更早地进入了GMFM-66最大值区域，但运动功能发育的峰值水平低于OMG，应充分重视在4岁后轻度脑瘫患儿中积极实施具有针对性的多种康复干预手段。

Assessment of gross motor function development in children with cerebral palsy by nonlinear mixed effect model

Objective To describe the patterns of gross motor development of children with cerebral palsy (CP) in each level of the Gross Motor Function Classification System (GMFCS) using nonlinear mixed effect model, as a basis for planning clinical management. Methods Patients with CP were enrolled from 7 rehabilitation centers in Shanghai form August 2000 to December 2007. Severity of CP was based solely on GMFCS level and motor function was assessed with Gross Motor Function Measure-66 (GMFM-66). The stable limit model was used to make the gross motor development curve for children in each of the 5 GMFCS levels. The stable limit model has two parameters, corresponding to limit of motor function and the rate which can transforms to age-90. Age-90 means the age at which children are expected to achieve 90% of their predicted limit in GMFM-66. In addition, the results of our study were compared with those of Canada study. Results A total of 228 children (152 males, 76 females) with CP were enrolled in the study. Types of CP in these children were spastic quadriplegia (n=63), spastic diplegia (n=87), spastic hemiplegia (n=48), athetotic （n=11）,dystonia (n=4) and ataxic (n=11). Based on a total of 986 GMFM assessments (4.32 assessments per child), distinct motor development curves were constructed. The limit of GMFM-66 in GMFCS Ⅰ-Ⅴ level was 81.2, 62.4, 52.9, 40.8 and 24.4 scores, the corresponding age-90 was 3.8, 2.7, 2.1, 2.0 and 1.5 years respectively. GMFM-66 limit in GMFCS level Ⅰ and Ⅱ of our study was lower than that in Canada study, however GMFM-66 limit in GMFCS level Ⅲ-Ⅴ was closer to that in Canada study. Moreover, the corresponding age-90 in each 5 levels of GMFCS in our study was lower than that in Canada study. Conclusions The gross motor development more quickly reached its limit in GMFCS level Ⅰ and Ⅱ, however the limit of GMFM-66 was lower than that in Canada study. More attention should be paid to the patients younger than 4 years to plan intervention.