Halliday Timothy J, Kwak Sally
University of Hawaii Manoa, Honolulu, HI 96822, United States.
Econ Hum Biol. 2009 Jul;7(2):181-90. doi: 10.1016/j.ehb.2009.05.002. Epub 2009 May 14.
Despite the urgent public health implications, relatively little is yet known about the effect of peers on adolescent weight gain. We describe trends and features of adolescent BMI in a nationally representative dataset and document correlations in weight gain among peers. We find strong correlations between own body mass index (BMI) and peers' BMI's. Though the correlations are especially strong in the upper ends of the BMI distribution, the relationship is smooth and holds over almost the entire range of adolescent BMI. Furthermore, the results are robust to the inclusion of school fixed effects and basic controls for other confounding factors such as race, sex, and age. Some recent research in this area considers whether or not adolescent weight gain is caused by peers. We discuss the econometric issues in plausibly estimating such effects while accounting for growth spurts and difficulties in defining adolescent obesity. While our work identifies correlations between adolescent BMI and peers' BMI, it is not intended to and cannot fully address the existence of endogenous peer effects.
尽管存在紧迫的公共卫生影响，但关于同龄人对青少年体重增加的影响，目前所知相对较少。我们在一个具有全国代表性的数据集中描述了青少年BMI的趋势和特征，并记录了同龄人之间体重增加的相关性。我们发现自己的体重指数（BMI）与同龄人的BMI之间存在很强的相关性。尽管这种相关性在BMI分布的较高端尤为明显，但这种关系是平滑的，并且几乎在青少年BMI的整个范围内都成立。此外，在纳入学校固定效应以及对种族、性别和年龄等其他混杂因素进行基本控制后，结果依然稳健。该领域最近的一些研究探讨了青少年体重增加是否由同龄人引起。我们讨论了在合理估计此类影响时的计量经济学问题，同时考虑到生长突增以及定义青少年肥胖的困难。虽然我们的研究确定了青少年BMI与同龄人BMI之间的相关性，但并非旨在也无法完全解决内生同龄人效应的存在问题。
