摘要
目的:根据不同思路建立Actigraph加速度传感器能耗预测模型并加以验证,为更好地应用加速度传感器监测体力活动提供依据。方法:以男女大学生各30人为实验组,进行静坐、看书、整理书桌、扫地、慢走(4 km/h)、快走(6 km/h)和慢跑(8 km/h)共7项活动;另以男女大学生各10人作为验证组,进行约4 h日常体力活动。分别以Actigraph GT3X加速度传感器和Cosmed K4b2气体代谢分析仪监测垂直轴加速度记数(accelerometry counts,AC)和能耗。采用线性回归法,根据实验组数据建立3个以AC为自变量的能耗预测模型。以验证组数据,采用配对t检验和Altman-Bland图验证上述3个模型和Freedson模型的有效性。结果:本研究建立的分段线性模型为:如AC<1630 counts/min,则METs=1.419+0.005644×AC-5.927×10-6×AC2+1.993×10-9×AC3;如AC≥1630 counts/min,则METs=1.818+0.000752×AC。经验证,使用Actigraph加速度传感器监测日常体力活动时,可应用上述模型推算体力活动总能耗。
Objective To develop and validate the energy expenditure prediction models for monitoring physical activity. Methods Experiment group (30 male and 30 female college students)underwent 7 activities, including sitting, reading, desk cleaning, floor sweeping, slow walking (4 km/h), fast walking (6 km/h) and jogging (8 km/h). Validation group ( 10 male and 10 female college students)underwent 4-hour daily physical activity. Accelerometry counts (AC)were detected by Actigraph GT3X accelerometer and oxygen consump- tion by energy expenditure device Cosmed K4b2. Three prediction models for energy expenditure were devel- oped based on the data from experiment group using the linear regression, then they were compared to the da- ta from validation group and validated by paired t -test and Altman-Bland figure analysis. Energy expendi- ture was calculated using Freedson formula(METs = 1.439008+0.000795xAC ). Results The following piece- wise regression model can effectively estimate 4-hour energy expenditure : METs = 1.419 + 0.005644 x AC - 5.927 x 106x AC2+ 1.993 x 10-9x AC3' when AC 〈 1630 counts/min; METs -- 1.818 + 0.000752 x AC, when AC≥1630 counts/min. Conclusion Our piecewise regression models can be used to calculate energy expendi- ture when physical activity was monitored by accelerometer.
出处
《中国运动医学杂志》
CAS
CSCD
北大核心
2013年第11期961-965,共5页
Chinese Journal of Sports Medicine
基金
浙江省教育厅科研项目(Y201226077)
关键词
加速度传感器
能耗
预测模型
体力活动
accelerometer, energy expenditure, prediction model, physical activity