Midazolam 12 mg is moderately counteracted by 250 mg caffeine in man

Caffeine (Caf) counteracts various effects of benzodiazepines (BZDs). Since the effects of zolpidem, a short-acting atypical GABA(A)-BZD agonist, were not antagonized by Caf, we studied an interaction between Caf and midazolam (Mid) in healthy volunteers.

In Study 1, 108 healthy students divided to 6 parallel groups were given Mid 12 mg (capsule) and Caf 125 and 250 mg (in decaffeinated coffee), alone and in combinations in the double-blind placebo-controlled manner. Objective and subjective tests were done before and at 45 and 90 min after intake. Ranked delta-values (changes from baseline) were analyzed by one-way contrast ANOVA and Scheffe's tests. In Study 2, six healthy subjects took Mid 15 mg (tablet) with and without Caf 300 mg. The dynamic effects were analyzed as in Study 1 and the plasma concentrations were assayed.

In Study 1, learn effects after placebo (ad + 15%) were seen for letter cancellation and digit symbol substitution tests. Midazolam alone significantly (p < 0.05 vs. delta-placebo) reduced letter cancellation and digit symbol substitution, lowered flicker fusion, impaired digit learning and caused subjective calmness on VAS. Caffeine alone did not differ from placebo objectively, yet improved quick-wittedness and contentedness on VAS. In the combinations, Mid + Caf 125 mg differed from placebo objectively as Mid alone, whereas Mid + Caf 250 mg did not. Mid + Caf 250 mg differed from Mid on digit substitution, but did not differ from Mid+Caf 150 mg in impairing memory and causing subjective sedation. In Study 2, Mid 15 mg caused sedation and Caf 300 mg increased plasma Mid at 45 min. Mid + Caf did not differ from Mid alone objectively, but did so subjectively on VAS (p > 0.05).

In conclusion, in a parallel group study, sedative effects of Mid 12 mg were only moderately antagonized by Caf 250 mg but not by Caf 125 mg. In a cross-over study, a weak interaction was found subjectively but not in objective measures.