文摘
The purpose of this dissertation is to analyze the interaction between various psychological, environmental, household and genetic factors and body mass index BMI) growth during the youths teens and twenties. The first essay examines the socio-economic factors that are correlated with BMI among NLSY97 respondents at two points in time, 1997 and 2010. The youth were 12 to 17 in the first survey year and 25-31 in the last. Covariates of importance in the final year were surprisingly similar to those in the first. Common socio-economic factors had only modest impacts on the respondents BMI in either year. Consistent with other studies, maternal BMI had a strong, positive effect on adolescent BMI, but also on BMI of the respondents in their late twenties. The impact was systematically larger for biological mothers, which could suggest a genetic factor. The second essay assesses the determinants of annual BMI growth rates over the entire 14 year period. The strong maternal BMI effect emerges in the annual growth rate estimates as well. The general results are robust to the use of individual and time fixed effects, though th the fixed effect approach does not permit estimation of maternal BMI only available in 1997) on BMI growth. Growth rates decline with age, as one might expect, and more so at higher levels of initial BMI. Conversely BMI growth rates are less for respondents with initially higher BMI, a result that increases in absolute value as the respondent ages. The modest impacts of the usual socio-economic factors on BMI change between 1997 and 2010 noted in the first essay are confirmed in this annual growth rate data. The last essay focuses on the decision-making process that underpins these BMI relationships. Weight control is considered as an expected utility maximization decision subject to the usual budget constraint in food and other goods. Of special interest is the impact of the youths perception of his or her weight-to-height BMI) status. In the first survey year, males and females appear to use different metrics for judging their height to weight proportionality. Males appear to judge themselves by standards, that are used for adults of the same BMI, females align themselves more closely with the CDC youth standard, which factors in expected growth in BMI until adulthood. Perhaps not surprisingly, males become heavier higher BMI) by 2010. Misperceptions of weight category become systematic by 2010, with obese young 25-31) men and women reluctant to define themselves as "very overweight." This bias can be found in the 1997 data as well, but is less obvious because of the far fewer obese respondents in the first year. Misperceiving weight has deleterious consequences---those who underestimate increase weight faster and those who overestimate increase weight slower---compared to accurate perceivers.
