PURPOSES: The maximum amount of alcohol consumed on any day is an important indicator of problem drinking. This study examined the ability of 2 maximum measures, taken individually and in combination, to predict alcohol-related consequences and dependence symptoms. METHOD: We analyzed data from 4,296 current drinkers who participated in the 2000 National Alcohol Survey (N10) and provided complete data on several alcohol measures. Items included the most they drank in any day in the past 12 months assessed by 2 questions: a standard question with a 6-level categorical response format and an open-ended (continuous) question referencing the occasion that you had the most to drink that followed items on drinking contexts. RESULTS: In a within-subjects analysis, more than two-thirds of the respondents provided consistent responses on both questions (69.9%), 12.4% respondents reported higher maximum quantities on the open-ended question partitioned into equivalent categories, and 17.7% reported a higher maximum on the initial categorical question. In multiple logistic regression analyses predicting alcohol-related consequences, the continuous open-ended maximum or the mean of the 2 maximum items were the strongest predictors, whereas for predicting alcohol dependence symptoms, the maximum of the open-ended and closed-ended maximum items was a better predictor than each measure alone. CONCLUSION: Although categorical response formats generally capture high-quantity drinking better than open-ended formats, use of a second, alternatively formatted "most drunk" question may modestly improve ability to predict survey indicators of alcohol use disorders.
PURPOSES: The maximum amount of alcohol consumed on any day is an important indicator of problem drinking. This study examined the ability of 2 maximum measures, taken individually and in combination, to predict alcohol-related consequences and dependence symptoms. METHOD: We analyzed data from 4,296 current drinkers who participated in the 2000 National Alcohol Survey (N10) and provided complete data on several alcohol measures. Items included the most they drank in any day in the past 12 months assessed by 2 questions: a standard question with a 6-level categorical response format and an open-ended (continuous) question referencing the occasion that you had the most to drink that followed items on drinking contexts. RESULTS: In a within-subjects analysis, more than two-thirds of the respondents provided consistent responses on both questions (69.9%), 12.4% respondents reported higher maximum quantities on the open-ended question partitioned into equivalent categories, and 17.7% reported a higher maximum on the initial categorical question. In multiple logistic regression analyses predicting alcohol-related consequences, the continuous open-ended maximum or the mean of the 2 maximum items were the strongest predictors, whereas for predicting alcohol dependence symptoms, the maximum of the open-ended and closed-ended maximum items was a better predictor than each measure alone. CONCLUSION: Although categorical response formats generally capture high-quantity drinking better than open-ended formats, use of a second, alternatively formatted "most drunk" question may modestly improve ability to predict survey indicators of alcohol use disorders.
Authors: E Anne Lown; Ann C Mertens; Rachael A Korcha; Wendy Leisenring; Melissa M Hudson; Thomas K Greenfield; Leslie L Robison; Lonnie K Zeltzer Journal: Psychooncology Date: 2012-06-27 Impact factor: 3.894
Authors: Lonnie K Zeltzer; Christopher Recklitis; David Buchbinder; Bradley Zebrack; Jacqueline Casillas; Jennie C I Tsao; Qian Lu; Kevin Krull Journal: J Clin Oncol Date: 2009-03-02 Impact factor: 44.544
Authors: Thomas K Greenfield; Yu Ye; Jason Bond; William C Kerr; Madhabika B Nayak; Lee Ann Kaskutas; Raymond F Anton; Raye Z Litten; Henry R Kranzler Journal: J Stud Alcohol Drugs Date: 2014-03 Impact factor: 2.582
Authors: Thomas K Greenfield; Yu Ye; William Kerr; Jason Bond; Jürgen Rehm; Norman Giesbrecht Journal: Int J Environ Res Public Health Date: 2009-12-11 Impact factor: 3.390