Short-term effects of intravenous benzodiazepines on autonomic neurocardiac regulation in humans: a comparison between midazolam, diazepam, and lorazepam.

Objectives To evaluate the effects of intravenously applied diazepam, lorazepam, and midazolam on autonomic neurocardiac regulation assessed by standardized measurements of heart rate variability. Design Prospective, randomized clinical study. Setting University teaching hospital. Patients Forty-five patients, who underwent a gastroscopy, were randomly assigned to intravenous premedication with midazolam (5 mg), diazepam (10 mg), or lorazepam (4 mg). Six subjects refused an injection and served as nonpremedicated controls. Interventions Serial recordings of the 5-min resting heart rate variability were obtained before and 15 and 30 mins after premedication. Seven benzodiazepine-treated patients received intravenous flumazenil (0.5 mg). Measurements and Main Results The average doses applied were 0.07 mg/kg for midazolam, 0.13 mg/kg for diazepam, and 0.06 mg/kg for lorazepam. Fifteen minutes after intravenous benzodiazepines were administered, we found an increase in resting heart rate and a reduction of vagal tone compared with baseline in all three benzodiazepine-treated subgroups. Multivariate analysis (covariate age) of the changes in heart rate variability indices over the experimental course revealed a significant reduction in absolute high-frequency power with midazolam or diazepam compared with nonpremedicated subjects. Moreover, midazolam-treated subjects showed a significantly larger reduction in relative high-frequency power not only compared with nontreated subjects, but also compared with lorazepam- or diazepam-treated subjects. Vagal tone remained reduced compared with baseline even 30 mins after benzodiazepine application, however, the resting heart rate decreased toward baseline levels. After flumazenil administration, there was a linear correlation between an increase in high-frequency power and a corresponding decrease in resting heart rate. Conclusions Benzodiazepines can influence autonomic neurocardiac regulation in man, probably through their interaction with the γ-aminobutyric acidA-receptor chloride ion channel complex. The pattern of findings suggests that intravenous midazolam, diazepam, and lorazepam influence human autonomic neurocardiac regulation in a biphasic way. First, they cause a reduction of central vagal tone, and second, they may decrease the cardiac pacemaker directly. Flumazenil completely abolished the autonomic neurocardiac regulation effects of benzodiazepines.