Trend of smoking and psychosocial predictors among youth

描述（由申请人提供）：从20世纪70年代中期到1992年，美国青年吸烟的流行率稳步下降。然而，这些比率从1992年开始上升，并于1997年达到顶峰。在过去五年中，再次观察到青年吸烟率稳步下降。在加州也观察到了类似的模式。在这项研究中，我们将使用来自加州两次青少年烟草调查的档案数据的二次分析，以确定这种倒V形趋势是否适用于各种吸烟行为（例如，30-加州青少年每天和每周的吸烟行为）。我们还将研究不同年龄组和出生队列的吸烟趋势。此外，我们将确定烟草使用的几个风险和保护因素（例如，朋友吸烟、社会规范），并使用两种新开发的统计方法：年龄-时期-队列分析和多层次结构方程模型，研究他们在过去十年中的变化模式及其对烟草使用的影响。本项目拟分析的拟议数据集（加州青年调查和加州烟草青年调查）提供了1990年至2000年的一系列丰富的横截面数据。每个横断面样本提供了一个重复的调查烟草使用的青年在加州。该项目建议使用多元回归、逻辑回归和结构方程模型分析横截面数据。从复制的横截面样本中获得的信息提供了随时间变化的概况。此外，可以使用多层次方法对这些问题进行研究，以更好地了解烟草使用的长期趋势、与烟草使用有关的心理社会因素的变化以及这些因素对烟草使用的影响随时间的变化。该项目很重要，因为它将确定与吸烟最新趋势相关的因素，这将有助于确定烟草使用预防计划的关键组成部分，从而为青年制定有效和有针对性的预防干预计划。

DESCRIPTION (provided by applicant): From the mid-1970's to 1992, the prevalence rates of cigarette smoking among youth in the United States steadily decreased. However, these rates increased from 1992 and peaked in 1997. During the past five years, a steady decline in tobacco use among youth has again been observed. A similar pattern has also been observed in California. In this study, we will use secondary analysis of archival data from two youth tobacco surveys in California to determine whether this inverted V-shaped trend applies to various smoking behaviors (e.g., 30-day and weekly smoking behaviors) among youth in California. We will also examine the trends of cigarette use across different age groups and birth cohorts. Furthermore, we will identify several risk and protective factors of tobacco use (e.g., friends' smoking, social norms) and examine their patterns of change and impacts on tobacco use in the last ten years using two newly developed statistical approaches: age-period-cohort analysis and multilevel structural equation model. The proposed data sets (California Youth Survey and California Tobacco Youth Survey) to be analyzed in this project provide a rich array of cross-sectional data from 1990 to 2000. Each cross-sectional sample provides a replicated survey of tobacco use among youth in California. This project proposes to analyze the cross-sectional data using multiple regression, logistic regression, and structural equation models. Information obtained from the replicated cross-sectional samples provides a profile of changes across time. Furthermore, they can be studied using the multilevel approach to gain a better understanding of the long-term trend of tobacco use, changes in tobacco use-related psychosocial factors, and changes in the impacts of those factors on tobacco use across time. The project is important because it will identify factors associated with recent trends in cigarette smoking, which will help in specifying critical components for tobacco use prevention programs, leading to efficient and well-targeted prevention intervention programs for youth.