Identifying the factors associated with why people with an alcohol or drug use disorder improve and maintain improvement in their substance use is critical to understanding the change process. Given that a significant number of individuals with substance use problems seek assistance from formal or informal sources, the role of help-seeking in the change process is itself an important area for study. Little empirical research on these areas has been done using large-scale national datasets, and that work has not examined behavior change longitudinally. The current proposal uses cross sectional and longitudinal data from Waves 1 and 2 of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) to examine these issues. The first objective of this project is to identify the factors associated with changes in substance use among individuals with a lifetime diagnosis of abuse or dependence for either alcohol, drugs, or both alcohol and drugs (N = 12,086). It is hypothesized that select personal and environmental variables will be associated with change, that abstinence will be associated with greater perceived severity of the problem, and that improvement among individuals with an alcohol and drug disorder will be associated with more legal consequences as compared to individuals with only an alcohol disorder. For this first objective Markov modeling will be used to test the predictive power of personal and environmental factors on improvement. The second objective of this project is to identify the factors associated with help-seeking among these same individuals. It is hypothesized that personal and environmental factors will also be associated with help-seeking, that specific factors will predict use of 12- step assistance versus non-12-step assistance, and that personal factors will be a better predictor of help-seeking among individuals with an alcohol use disorder only while environmental factors will be a better predictor of help-seeking among individuals with both alcohol and drug use disorders. For this second objective latent variable modeling will be used to test the predictive power of personal and environmental factors on help-seeking. The third objective is to identify the factors associated with maintaining improvement from Wave 1 to Wave 2 of NESARC among individuals with a lifetime substance use disorder who no longer met criteria for diagnosis in the year preceding Wave 1 (N = 9,027). It is hypothesized that physical health, psychosocial functioning, and negative life events will predict maintenance of change over time. For this third objective Markov modeling will be used to predict maintenance of improvement from Wave 1 to Wave 2. The results of this study will provide new information about the initiation and maintenance of behavior change, and may contribute significantly to informing public policy, prevention, and treatment efforts.
