The influence of dosage time of midazolam on its pharmacokinetics and effects in humans

The influence of dosage time of midazolam on its pharmacokinetics and effects on the central nervous system were investigated in six healthy volunteers, with pharmacokinetic‐pharmacodynamic modeling. Each volunteer received single oral doses of 15 mg midazolam on four separate occasions: 8 AM, 2 PM, 8 PM, and 2 AM. An almost significant circadian variation was found in elimination half‐life, shortest at 2 PM (1.26 ± 0.47 hours, mean ± SD) and longest at 2 AM (1.57 ± 0.44 hours) ( p = 0.05). Drug effects measured were α activity of the electroencepalograph and P100 latency of the visual‐evoked response. The maximum drug effect (E max ) model described the concentration‐effect relationship, extended with either a threshold drug concentration or a sigmoidicity parameter. A significant circadian variation was found in baseline α activity: highest at 8 AM (109% ± 19% of the 24‐hour mean) and lowest at 2 AM (80% ± 12%). For α activity the drug concentration at half‐maximum effect of both threshold E max model and sigmoid E max model showed lower values at 8 AM and 2 AM and higher values at 2 PM and 8 PM. However, these differences were either not significant ( p = 0.10, threshold model) or on the verge of statistical significance ( p = 0.05, sigmoid model). No circadian variation was found in the parameters describing the effect on the visual‐evoked response. We conclude that the sensitivity of the central nervous system to midazolam, as reflected in α activity, possibly shows a circadian variation. Clinical Pharmacology and Therapeutics (1991) 50 , 16–24; doi: 10.1038/clpt.1991.99