Gasser T, Sheehy A, Largo R H
Universität Zürich, Switzerland.
Ann Hum Biol. 2001 Jul-Aug;28(4):395-402. doi: 10.1080/03014460010013025.
This is a methodical investigation into the problem of estimating parameters for the pubertal spurt (PS). The variability involved in determining the timing, intensity and duration of the PS for height, leg height and biiliac width is estimated via a realistic simulation. Further, a decomposition of adolescent growth into a component due to the pubertal peak and one due to ongoing prepubertal velocity is evaluated.
Data for 120 boys and 112 girls are available from 4 weeks to adulthood. The curve-fitting method is kernel estimation for distance, velocity and acceleration.
The age of peak velocity and the age of stopping of the PS are well determined. In contrast, the age of onset of the PS is less well determined. Intensity is less variable for the parameter peak velocity than for maximal acceleration. It is feasible to decompose adolescent growth into a component due to the pubertal peak and one due to ongoing prepubertal growth.
Nonparametric curve-fitting methods which do not rely on a parametric growth model can be successfully used to extract individual characteristics of the PS.
这是一项对青春期生长突增（PS）参数估计问题的系统性研究。通过逼真的模拟来估计确定身高、腿长和髂嵴间宽度的PS的时间、强度和持续时间所涉及的变异性。此外，还评估了将青少年生长分解为因青春期高峰引起的成分和因青春期前持续生长速度引起的成分。
有120名男孩和112名女孩从4周龄到成年期的数据。曲线拟合方法是对距离、速度和加速度进行核估计。
速度峰值年龄和PS停止年龄确定得很好。相比之下，PS开始年龄的确定程度较差。速度峰值参数的强度变化比最大加速度的变化小。将青少年生长分解为因青春期高峰引起的成分和因青春期前持续生长引起的成分是可行的。
不依赖参数化生长模型的非参数曲线拟合方法可成功用于提取PS的个体特征。
