Background: United States (US) epidemiological studies suggest that for every 5-8 who start drinking alcoholic beverages, at least one drinker will develop an alcohol dependence (AD) syndrome within the first 10 years after onset of drinking (Lopez-Quintero et al., 2011; Wagner & Anthony, 2002). Recently, we described a multiparametric functional analysis approach for new research to estimate these transition probabilities with a one-dimensional function (1D; Vsevolozhskaya & Anthony, 2015). Here, we demonstrate extension of this analysis to two-dimensional (2D) functions that combine information about number of recent drinking days and number of drinks on the typical drinking day.
Methods: Data are from the United States National Survey on Drug Use and Health (NSDUH) Restricted-use Data Analysis System, 2002-2011, with nationally representative samples of newly incident drinkers and rapid-onset AD syndromes ascertained via standardized audio computer self interviews, completed for surveys of non-institutionalized civilian US citizens, age 12 years and older. Drinking history, including DSM-IV AD status, were assessed via the standardized computer-assisted interview assessments. The 2D functional estimates are based on a non-linear parametric Hill equation evaluated for (1) number of drinking days in 30 days just before NSDUH assessment, and (2) typical number of drinks on recent drinking days.
Results: Among newly incident drinkers with just one drink per drinking day, the estimated AD risk ranges from more or less 1% among infrequent drinkers with a single drinking day per month (95% bootstrap confidence interval, CI: 0.7, 1.0), upward to about 3% among daily drinkers (95% CI: 1.4, 3.7). Among newly incident drinkers with ~2 drinks per drinking day, estimated AD risk is much larger among daily drinkers (21.4%; 95% CI = 5, 21). Across subgroups defined by 3, 4, and 5 or more drinks per day, the estimated AD risk is larger, as can be seen clearly for those who have progressed to daily drinking: 31% for 3 drinks, 84% for 4 drinks, 90% for 5+ drinks, respectively, with some degree of CI overlap. However, among infrequent drinkers, with no more than one drinking day per month, the estimated AD risk does not appreciably differ from 1% irrespective of the number of drinks consumed per typical drinking day.
Conclusions: Via the multiparametric functional analysis approach extended beyond the number of drinks per typical drinking day, this evidence helps clarify that AD risk apparently is relatively constant and quite limited when newly incident drinking is limited to no more than one drinking day per month. When newly incident drinkers are observed within 12 months after drinking onset, there is substantial increase in AD risk among daily drinkers, provided the typical number of drinks per day increases from 1 to 5+ drinks. This study is novel in its focus on newly incident drinkers and variations in risk of developing alcohol dependence soon after drinking onset. A new agenda for research AD risk among newly incident drinkers can be built upon this initial platform of new evidence, particularly if family history and individual-level genomic characteristics can be assessed and brought into play in future national surveys of this type.
Vsevolozhskaya, Olga A. and Anthony, James C., "Estimated Probability of Becoming Alcohol Dependent: Extending a Multiparametric Approach" (2015). Biostatistics Presentations. 2.