The frequency of caffeine withdrawal in a population-based survey and in a controlled, blinded pilot experiment

Reports of symptoms when regular caffeine consumption is stopped have appeared in the medical literature, but the frequency and significance of this phenomenon have remained controversial. The objective of this study was to collect information on the prevalence and severity of caffeine withdrawal in the general population and determine the incidence and type of symptoms reported on blind abrupt and gradual caffeine cessation among coffee drinkers reporting past episodes of caffeine-withdrawal symptoms. A community-based telephone survey was followed by a stratified, randomized, double-blind controlled study. Participants included 11,112 persons spontaneously calling to inquire about studies not related to caffeine and 57 regular caffeine users selected from among the callers because of self-reported caffeine withdrawal symptoms. Gradual or abrupt withdrawal from caffeine was compared to continuation of the same caffeine level. In a survey of 11,112 persons, 61% reported daily caffeine consumption, and 11% of the caffeine consumers reported symptoms upon stopping caffeine. Among the regular caffeine users, only 0.9% of males and 5.5% of females reported symptoms significant enough to interfere with normal activities when they abruptly stopped caffeine. A group of those reporting withdrawal symptoms were randomly assigned to three subsamples. In the group subjected to abrupt withdrawal (N = 18), 6 (33.3%) reported symptoms (e.g., headaches and tiredness). Including decreases in functional ratings, a total of 7 of the 18 (38.8%) could be considered to have experienced caffeine withdrawal. The gradual withdrawal group (N = 20) reported minimal if any caffeine withdrawal symptoms. A third group (N = 18) was kept on a level dose of caffeine for comparison. When participants are unaware of the caffeine-withdrawal focus of the study, these results suggest that both the frequency and severity of caffeine-withdrawal symptoms are much lower than found in some previous reports and that clinically significant symptoms may be uncommon events among the general population. (C) 1999 the American College of Clinical Pharmacology.