COMBINE Manuscript Proposal Form
For Administrative Use OnlyMS # B07 / Date Received: 4/2/04 / Date Reviewed:
Priority Set: / Scheduled for Review: 4/12/04 / Action:
Title: Psychosocial distress as a function of the consequences of drinking behavior and as a predictor of readiness to change
Proposed Writing Group Members: Tim Mueller, Michelle Hansten, Carolyn Cichanski,
Jennifer Williams and Stephanie O’Malley
Proposal:
Introduction: Increased psychosocial distress at baseline (perceived stress, psychiatric symptoms) appears to be predictors of worse drinking outcomes in treatment studies. It is possible that increased psychosocial distress is a reflection of the increased adverse consequences of alcohol use that occurs prior to entry into a treatment study. Furthermore, increased psychosocial stress may serve as a motivating stimulus to people seeking treatment, reflected in their readiness to change drinking behaviors.
Hypotheses:
2. Increased “alcohol involvement” will predict increased psychosocial stress.
3. Increased psychosocial stress will predict increased readiness to change.
Approach:
Controlling for baseline demographic variables (gender, age, SES, race, etc), test the hypothesis that increased alcohol involvement as reflected in high scores on the DrInC, high levels of alcohol use on the Form-90, high DSM symptom count, will predict increased psychosocial stress as reflected in the POMS, DSS, and BSI.
Again controlling for baseline demographic variables, test the hypothesis that increased psychosocial stress will predict more advanced levels of motivation for changing drinking behaviors. It is possible that there will not be enough variability in the measure of the stage of change at baseline and we may need to look at individual items in the scale.
Baseline psychosocial stress is less interesting as a sole focus then is an examination of the predictive power of psychosocial stress at baseline and subsequent treatment outcomes and the interplay of psychosocial stress during the study and proximate drinking behaviors. However this baseline analysis will set the stage for the later projects.
Data analysis:
1. Basic information about sample and variables
a. Sample descriptive information on age (M;SD), gender, educational level, ethnicity, marital status, employment, problem duration (M;SD), history of prior alcohol treatment, past history of other substance abuse or dependence, past or current history of psychiatric disorders.
b. Distribution (Mean, Median, SD) of the overall sample on the baseline POMS, DSS, BSI.Form-90 percent days abstinent (PDA), drinks per drinking day (DDD), and heavy drinking days (HDD); the total and subscale scores on the DrInC; total score on the ADS, and the number of SCID criteria endorsed.
2. Increased “alcohol involvement” will predict increased psychosocial stress.
Two primary sets of multivariate regression analyses will be involved in testing this hypothesis:
a. In the first set, three separate regression analyses will be run, one each for the total scores on the POMS, DSS, and BSI as the dependent variables. Variables to be entered into the regression include age, gender, ethnicity, total score on the DrInC, drinks per drinking day (DDD) and percent days abstinent (PDA) from the Form-90, total score on the Alcohol Dependence Scale (ADS), and the number of criteria endorsed on the alcohol section of the SCID.
b. The second set of analyses is similar to the ones above, except that a multivariate index or indices of psychosocial stress will serve as the dependent variable(s). These indices will be empirically derived by entering the total scores from the BSI, POMS, and PSS into a principal components analysis with varimax rotation (if more than one component is derived). Depending on the number of derived components derived, one or more regressions analyses will be run. As above, the predictor variables in the regression will include age, gender, ethnicity, total score on the DrInC, drinks per drinking day and percent days abstinent from the Form-90, total score on the Alcohol Dependence Scale (ADS), and the number of criteria endorsed on the alcohol section of the SCID. The derived psychosocial stress variable(s) will serve as the dependent variable(s).
3. Increased psychosocial stress will predict increased readiness to change.
This stepwise regression analysis will have the continuous readiness score from the URICA as the dependent variable (Contemplation + Action + Maintenance – Precontemplation). Age, gender, ethnicity, and the total scores from the BSI, POMS, and PSS will be entered into the regression as predictor variables.