The Overlap in Predicting Alcohol Outcome for Two Measures of the Level of Response to Alcohol
Version of Record online: 22 DEC 2008
Copyright © 2008 by the Research Society on Alcoholism
Alcoholism: Clinical and Experimental Research
Volume 33, Issue 3, pages 563–569, March 2009
How to Cite
Schuckit, M. A., Smith, T. L., Trim, R., Fukukura, T. and Allen, R. (2009), The Overlap in Predicting Alcohol Outcome for Two Measures of the Level of Response to Alcohol. Alcoholism: Clinical and Experimental Research, 33: 563–569. doi: 10.1111/j.1530-0277.2008.00870.x
- Issue online: 19 FEB 2009
- Version of Record online: 22 DEC 2008
- Received for publication July 31, 2008; accepted October 29, 2008.
- Level of Response;
Background: Two different measures have been used to establish a person’s level of response (LR) to alcohol as a risk factor for alcohol use disorders. LR values established by the alcohol challenge protocol and the Self-Report of the Effects of Ethanol (SRE) questionnaire usually correlate at 0.3 to 0.4, up to 0.6. However, it is not clear how this correlation relates to the ability of each measure to predict alcohol outcomes. This paper evaluates that overlap.
Methods: Sixty-six Caucasian males (mean age = 22 years) from 2 protocols participated in alcohol challenges with 0.75 ml/kg of ethanol, filled out the SRE, and were followed with a structured interview ∼5 years later. The relationship between the subjective feelings of intoxication at the time of peak breath alcohol levels from the alcohol challenge and the SRE score for a time early in the drinking career were evaluated regarding predicting the drinks per occasion in the 6 months prior to follow-up.
Results: Cross-sectional correlations between alcohol challenge and SRE LR’s ranged from −0.25 (p < 0.05) to −0.32 (p = 0.02) for the full sample, and the 2 LR measures correlated with drinking at follow-up (−0.26 and 0.41, respectively). The SRE measure was more robust than the challenge in a regression analysis predicting the outcome in the context of other baseline predictors (e.g., drinking at baseline). As much as 60% of the ability of the more well established (gold standard) alcohol challenge LR to predict outcome was shared with the SRE. The alcohol challenge accounted for as much as 44% of the ability of the SRE to predict outcome.
Conclusions: The SRE-generated LR overlapped considerably with the alcohol challenge LR in the ability to predict future heavier drinking.