| Literature DB >> 15530719 |
Elizabeth E Epstein1, Erich Labouvie, Barbara S McCrady, Janine Swingle, Julie Wern.
Abstract
This study refines an empirically derived drinking pattern classification system [Addict. Behav. 20 (1995) 23] and assesses its concurrent and predictive validity in a new sample of alcohol-dependent adults in treatment. Drinking data were collected from 195 adults (133 men) at baseline and for 52 weeks postbaseline using the Timeline Follow-back (TLFB) method. Ninety-three percent of the sample were classified into one of four drinking patterns: binge (n=13, 6.5%), episodic (n=41, 21%), sporadic (n=17, 9%), or steady (n=111, 57%). The steady drinking group showed substantial variability in drinking intensity and was divided into steady/high intensity (n=67, 34%) and steady/low intensity (n=44, 23%) subgroups. With age and gender controlled, the five subgroups did not differ on baseline employment or marital status, but differed on a measure of relationship functioning. Binge and steady/high groups reported the most severe alcohol-use histories. Steady/low intensity drinkers had later age of onset of alcohol problems. The five-pattern classification demonstrated predictive validity; steady (heavy and low intensity) drinkers most dramatically reduced the frequency of their drinking at 26 weeks postbaseline, while binge drinkers reduced their drinking intensity. Binge and steady/heavy drinkers were more likely than steady/low and sporadic drinkers to be abstinent at follow up. The final classification into five pattern groups shows promise of clinical utility for treatment-seeking samples.Entities:
Mesh:
Year: 2004 PMID: 15530719 DOI: 10.1016/j.addbeh.2004.03.040
Source DB: PubMed Journal: Addict Behav ISSN: 0306-4603 Impact factor: 3.913