Abstract
Results are reported from two prospective longitudinal studies of heavy drinkers and alcoholics who have received treatment. The present study focuses on the drinking habits of subjects who have been interviewed four times after the end of treatment. The data uncover substantial changes in heavy drinkers' and alcoholics' consumption level over time, and these observed changes are only partly explainable in terms of measurement error and short-term fluctuations. Changes show signs of accumulation over time. This suggests that heavy drinkers' and alcoholics' consumption level is very unstable over longer periods. The signs of systematic progression are weak. No convincing evidence for loss of control is found, and neither is there convincing evidence for a strong and persistent progression towards heavier drinking. At all levels substantial changes are found, and these are nearly equally strong in both directions. It is concluded that the observed pattern of change more resembles an indeterministic (or stochastic) process than a systematic natural history of a disease.