The evidence for acute tolerance to human alcohol intoxication (the Mellanby effect): a systematic review

Objective: To establish the evidence for acute tolerance to human alcohol intoxication (the Mellanby effect [1]), which has been proposed to diminish the effects of a given blood alcohol concentration (BAC) during the descending part of the BAC-time curve. Methods: Multiple databases were searched using text words ‘‘tolerance,’’ ‘‘ascending’’, ‘‘descending’’ or ‘‘Mellanby’’ with Medline term ‘‘exp *alcohol/’’ or ‘‘exp *drinking behaviour/’’ or equivalent. Full text articles were retained for analysis if they dealt with acute (within dose) alcohol tolerance in human subjects and provided quantitative data on both the ascending and descending limbs of the BAC-time curve. Results: Of 384 unique articles identified and screened, 125 full text articles were assessed and 19 met criteria for analysis. Most studies were small, median 10 (range 4–28) subjects per group. Doses of alcohol and rates of administration differed. All effects are dependent on drinking history and the degree of intoxication. We distinguished eight major outcome domains (physiological effects, hand-eye co-ordination, perception, decision-making, mental arithmetic and reasoning tasks, verbal skills, memory, subjective alcohol effects), and these were assessed by at least 24 different methods.[2,3] Ratings at a given concentration (C) were better (closer to sobriety) at Cdown (descending) than at Cup (ascending) for subjects’ mean time for maze and peg-board tasks, arithmetic ability, and abstraction. Subjectively, those studied felt less drunk, and were twice as willing to drive at Cdown as at Cup. By contrast, cognitive tasks, error performance, inhibitory control, visual memory and performance in a simulated driving task were worse at Cdown. All values p < 0.05. Conclusion: The Mellanby effect is most firmly established for subjective intoxication. Confidence in ability to drive increases while measured driving ability falls. Objective measures of impairment are likely to be more robust than a person’s own account.