Loss of control, heavy drinking and drinking problems in a longitudinal study.

Abstract

The importance of loss of control over drinking in the development of alcoholism was tested using data from a 4 yr longitudinal study of 615 white men, aged 21-59, living in San Francisco [California, USA] when initially interviewed. Of the 582 drinkers, 36 had lost control at either time 1 (1st interview) or time 2 (2nd interview, 4 yr later) or at both. Of these 36 men, only 24 were defined as heavy drinkers (took at least 12 drinks on an occasion about once a month or 8 drinks on an occasion twice a week or 4 drinks on an occasion 3 or 4 times/wk). Problems with family or friends, fears of becoming an alcoholic, bout drinking and seeking medical or counseling help for alcohol problems bore no clear relation to loss of control, nor was it highly correlated with other problems or indicators of trouble in social functioning. Of the 18 persons who reported loss of control at time 1, 14 (78%) had 1 or more drinking-related problems at time 2, but the same number said that they no longer suffered from loss of control at time 2. Variables listed as problems at time 1 (bout drinking, coping behavior, loss of control, interpersonal problems, accidents, problems with police, job, finances or spouse) were analyzed using the Automatic Interaction Detection (AID) method to determine the best predictors of problems at time 2. The presence of drinking problems at time 1 was the best indication that an individual would have problems at time 2, although not necessarily the same problems. No dominant pattern of problems at time 1 could predict specific observations for time 2. Drinkers who had problems at time 1 but not at time 2 were younger and had lower incomes and occupational status than those who had problems throughout the 4 yr study. About 10% of both groups expressed personal fears of becoming alcoholics; and those out of trouble at time 2 reported generally fewer alcohol problems of all types at time 1. AID analysis revealed that the problem variables at time 1 explained about 34% of the variance in conditions at time 2, although loss of control specifically accounted only for about 7%, and was not necessary at time 1 for an individual to have serious problems at time 2. Moreover, by time 2, 69% of those without problems had reduced their drinking since time 1, while only 29% of those still with problems had done so. Loss of control was not necessarily part of an unfolding process of alcoholism, nor a 1-way gate to worse problems. Like other drinking problems, it seemed to come and go over time, and was only 1 problem among others to be used in predicting future drinking problems.