This competitive revision seeks to expand on the parent grant to provide an in-depth and rigorous psychometric evaluation of nicotine dependence (ND) symptoms by using an innovative method (Integrative Data Analysis: IDA) to pool the two study data sets used in the parent grant and, using the pooled data, apply a novel statistical method (Moderated Nonlinear Factor Analysis: MNLFA) to evaluate the psychometric properties of ND symptoms from three commonly used measures of nicotine dependence: the DSM-IV, the Nicotine Dependence Syndrome Scale (NDSS), and the modified Fagerstrom Tolerance Questionnaire (FTQ). In the past, combining substance use data sets to obtain more powerful tests of hypotheses due to increases in sample size, heterogeneity, and base rates of ND, has been prohibitive primarily because of differences across studies in the ways in which constructs are measured. However, IDA is a method for pooling data sets despite differences in study populations, design, and measures. IDA permits considerably more powerful, more comprehensive, and more rigorous studies than can be achieved by studying single study data sets independently. Furthermore, new statistical methods (MNLFA) have been developed that permit tests of similarity (invariance) in ND symptom properties (in terms of how the symptoms relate to the underlying ND construct and their likelihood of endorsement at different levels of the construct) across studies and measures, as well as permitting a simultaneous evaluation of invariance as a function of individual risk factors. The proposed research extends the parent study in several ways. First, it proposes two new aims: 1) to conduct a rigorous psychometric analysis of ND symptoms to test invariance across studies on the pooled data set and 2) to test invariance of these symptoms as a function of depression and alcohol use disorders. This psychometric evaluation will inform how differences in measures of ND and these individual risk factors influence the meaning and subsequent scoring of an ND construct. Finally, a third aim is to use a combined ND symptom score for each individual, resulting from the models tested in the first two aims, to cross validate the specific aims of the parent grant regarding the relation between smoking exposure and nicotine sensitivity. The purpose is to conduct a more powerful test of the parent study hypotheses by using this new ND score that has a common metric across studies despite differences in measurement and other study characteristics and that takes into account potential differences in the psychometric properties of the symptoms as a function of depression and alcohol use disorders. This application has significant potential to advance the goals of OppNet through application of new and innovative methodological and statistical methods to advance and optimize the measurement of ND by identifying ND symptoms that are invariant across studies despite differences in study design and measures. Results can inform the measurement of ND and the extent to which symptoms from different ND measures can be combined to develop an optimal measure of ND.
This research will provide a greater understanding of the measurement of nicotine dependence symptoms in young smokers. The proposed research is expected to inform further development of a brief but highly accurate assessment to identify young smokers who are at greatest risk for continued and heavier smoking and to inform targeted interventions for these young smokers.
Showing the most recent 10 out of 19 publications
