Alcohol consumption and problems: The relevance of drinking patterns

Background: Data on risks and benefits associated with drinking patterns provide the scientific basis for moderate, low-risk drinking guidelines. Illustrated are methods to investigate and adjust for heterogeneity in relations between three-dimensional drinking patterns and 41 alcohol problems assessed among current regular drinkers in the 1988 National Health Interview Survey. Methods: Three dimensions of mean drinking patterns, (i.e., usual quantities, heavy drinking rates (days of 5+ drinks/drinking days x 100), and drinking frequencies) were estimated in overlapping subsets of the population reporting each of the 41 problems, and mean usual quantities and heavy drinking rates were plotted against frequencies. Respondents were categorized into drinking problem groups associated with comparable mean drinking patterns; and main and interactive effects of age and sex on drinking patterns were examined by conducting three regression analyses within each group, with quantity, frequency, and heavy drinking rates as dependent variables, respectively. Results: Analyses revealed substantial heterogeneity in relations between drinking patterns and alcohol problems. Respondents having only minor problems drank on average two days a week, usually had 2.6 drinks, and drank heavily 12-13 days a year. Whereas, those having minor and severe problems drank an average of 3.5 days a week, usually had 4.7 drinks, and drank heavily 58 days a year. Within each problem group, usual quantity and frequency were higher among males than females, but the greatest gender differences were seen in heavy-drinking rates. Age-related differences in drinking patterns were striking. Usual quantity and heavy-drinking rates associated with problems decreased with age, whereas drinking frequency increased. Conclusions: Findings demonstrated the importance of assessing and adjusting for heterogeneity in relations between drinking patterns and alcohol problems when aggregating and interpreting such data, (e.g., when assessing alcohol dependence criteria or evaluating guidelines for moderate drinking), and illustrated new methods for doing so.