Hwang Jing-Shiang, Nadziejko Christine, Chen Lung Chi
Institute of Statistical Science, Academia Sinica, Taipei, Taiwan.
Inhal Toxicol. 2005 Apr;17(4-5):199-207. doi: 10.1080/08958370590912761.
Normal mice (C57) and mice prone to develop atherosclerosis (ApoE-/-) were implanted with electrocardiograph (EKG), core body temperature, and motion transmitters were exposed daily for 6 h to Tuxedo, NY, concentrated ambient particles (CAPs) for 5 day/wk during the spring and summer of 2003. The series of 5-min EKG monitoring and body-temperature measurements were obtained for each animal in the CAPs and filtered air sham exposure groups. Our hypothesis was that chronic exposure could cause cumulative health effects. We used our recently developed nonparametric method to estimate the daily time periods that mean heart rates (HR), body temperature, and physical activity differed significantly between the CAPs and sham exposed group. CAPs exposure most affected heart rate between 1:30 a.m. and 4:30 a.m. With the response variables being the average heart rate, body temperature, and physical activity, we adopted a two-stage modeling approach to obtain the estimates of chronic and acute effects on the changes of these three response variables. In the first stage, a time-varying model estimated daily crude effects. In the second stage, the true means of the estimated crude effects were modeled with a polynominal function of time for chronic effects, a linear term of daily CAPs exposure concentrations for acute effects, and a random component for unknown noise. A Bayesian framework combined these two stages. There were significant decreasing patterns of HR, body temperature, and physical activity for the ApoE-/- mice over the 5 mo of CAPs exposure, with smaller and nonsignificant changes for the C57 mice. The chronic effect changes of the three response variables for ApoE-/- mice were maximal in the last few weeks. There was also a significant relationship between CAPs exposure concentration and short-term changes of heart rate in ApoE-/- mice during exposure. Response variables were also defined for examining fluctuations of 5-min heart rates within long (i.e., 3-6 h) and short time periods (i.e., approximately 15 min). The results for the ApoE-/- mice showed that heart-rate fluctuation within the longer periods increased to 1.35-fold by the end of exposure experiment, while the heart-rate fluctuation within 15 min decreased to 0.7-fold.
正常小鼠(C57)和易患动脉粥样硬化的小鼠(ApoE -/-)植入了心电图(EKG)设备,同时记录核心体温,并通过运动传感器监测。在2003年春夏季节,小鼠每周5天、每天暴露于纽约州塔克西多的浓缩环境颗粒物(CAPs)中6小时。在CAPs组和过滤空气假暴露组中,对每只动物进行了一系列5分钟的EKG监测和体温测量。我们的假设是,长期暴露可能会导致累积健康影响。我们使用最近开发的非参数方法来估计CAPs组和假暴露组之间平均心率(HR)、体温和身体活动存在显著差异的每日时间段。CAPs暴露对凌晨1:30至4:30之间的心率影响最大。以平均心率、体温和身体活动作为响应变量,我们采用两阶段建模方法来估计对这三个响应变量变化的慢性和急性影响。在第一阶段,一个随时间变化的模型估计每日的粗略影响。在第二阶段,用时间的多项式函数对估计的粗略影响的真实均值进行建模以得到慢性影响,用每日CAPs暴露浓度的线性项来表示急性影响,并用一个随机成分来表示未知噪声。一个贝叶斯框架将这两个阶段结合起来。在CAPs暴露的5个月期间,ApoE -/-小鼠的心率、体温和身体活动有显著下降趋势,而C57小鼠的变化较小且不显著。ApoE -/-小鼠的三个响应变量的慢性影响变化在最后几周达到最大。在暴露期间,ApoE -/-小鼠的CAPs暴露浓度与心率的短期变化之间也存在显著关系。还定义了响应变量来检查长时间段(即3 - 6小时)和短时间段(即约15分钟)内五分钟心率的波动情况。ApoE -/-小鼠的结果表明,在暴露实验结束时,较长时间段内的心率波动增加到1.35倍,而15分钟内的心率波动下降到0.7倍。