Benzodiazepines are widely used for the treatment of anxiety and insomnia as well as a variety of other disorders. In most countries they are available on medical prescription. Problems of dependence and abuse have been recognized for some time, although the dependence potential of these drugs remains a matter of contention.14 In addition to their anxiolytic and hypnotic effects, benzodiazepines produce sedation, muscle relaxation, and cognitive and psychomotor impairment.
A number of studies using animals have shown that caffeine and ben-zodiazepines share a mutually antagonistic relationship. In a study of locomotor activity in mice, Kaplan et al.15 found increases following caffeine administration and decreases following alprazolam administration. Coad-ministration of the two drugs revealed that caffeine and alprazolam each blocked the behavioral effects of the other. Also using mice, Kuribara et al.16 observed a reduction in the locomotor stimulatory effects of caffeine by doses of diazepam that had no effect when administered alone. The same group also found that caffeine reduced the impairment in avoidance behavior produced by diazepam.17 Additionally, caffeine has been shown to block a diazepam-induced reduction in skeletal muscle tone.18
This antagonistic relationship may hold for certain behavioral effects only. De Angelis et al.19 found that the effects of caffeine were reversed by desmeth-yldiazepam and chlor-desmethyldiazepam in the hole-board and spontaneous motor activity tests in mice. In contrast, benzodiazepine coadministration enhanced the increase in open-field exploratory behavior induced by caffeine. It should be noted that in this latter test some benzodiazepine doses produced increased exploratory activity when administered alone. Similarly, caffeine reversed the depressant effects of diazepam on walking and rearing in the open field test, but failed to attenuate the decrease in ambulation.20
In two very detailed studies of the effects of caffeine and alprazolam on the performance of rats on differential reinforcement of low-rate (DRL) behavior, Lau and Wang21 and Lau et al.22 observed additive effects of the two drugs. Both drugs disrupted DRL performance when administered alone. The effects of combinations were then predicted based on these findings. It was shown that there was no significant variation from a model of simple additivity. Importantly, there was no evidence of antagonism.
These same authors earlier studied another type of behavior that is impaired by both caffeine and benzodiazepines: fine motor control. In studies of acute23 and chronic24 administration of caffeine and midazolam to rats, the combination produced a greater degree of impairment than either drug alone. Together, the findings of this group clearly demonstrate that the ben-zodiazepines and caffeine do not have a simple antagonistic relationship, particularly where their effects are similar when administered alone.
Several studies have been concerned with anxiolytic/anxiogenic effects of caffeine-benzodiazepine combinations. Using the social interaction test of anxiety in rats, Baldwin and File25 found that the anxiogenic effects of caffeine could be reversed by chlordiazepoxide. Tang et al.26 used the consumption of 1.5% NaCl solution by rehydrating rats as a measure of anxi-olysis. Caffeine increased consumption of the solution at low doses (an anxiolytic effect), but decreased it at higher doses. Clonazepam had a clear anxiolytic effect which was attenuated by caffeine. Using the elevated plus maze, Bhattacharya et al.27 observed anxiolytic effects following lorazepam and anxiogenic effects when caffeine was administered. Lorazepam clearly attenuated the anxiogenic effects of caffeine. A final model of anxiety is the reversal of suppression induced by a punishing stimulus. Coffin and Spealman28 showed that the food-reinforced behavior of squirrel monkeys, suppressed by response produced electric shocks, was increased by chlor-diazepoxide. Coadministration of caffeine either had no effect or magnified the increase. These findings would appear to contradict those described above using different methods and species, but confirm an earlier report from the same laboratory.29 They also contrast with the findings from Polc et al.18 using similar methodology in rats, who found that caffeine blocked the anticonflict effect of diazepam.
Finally, two studies of the discriminative effects of chlordiazepoxide and caffeine showed that each is able to block the discriminative effects of the other. Holloway et al.30 found that chlordiazepoxide was able to block the discriminative effects of caffeine in a dose-related manner. However, it should be noted that the reversal was incomplete: doses of caffeine large enough to markedly decrease response rates did not completely abolish the caffeine discriminative cue. Pentobarbital failed to alter the caffeine discriminative cue, suggesting that the chlordiazepoxide-induced changes may be specific to benzodiazepines and not all sedative compounds. Gauvin et al.31 showed a parallel shift to the right in the chlordiazepoxide dose-response curve following caffeine coadministration. However, after comparing the effects of caffeine with those of a benzodiazepine antagonist, the authors concluded that the effects of caffeine were due to perceptual masking rather than receptor interaction.
Studies with humans have generally supported the hypothesis of a mutually antagonistic relationship between caffeine and benzodiazepines. In a brief report, Mattila et al.32 indicated that caffeine blocked the impairment of cognitive skills, increase in muscle relaxation, and subjective calming produced by 10 mg diazepam. Similar results were reported in a more detailed follow-up study.33 Using the benzodiazepine triazolam, Mattila et al.34 found relatively mild sedative effects that were largely reversed by caffeine. Similarly, the alerting effects of caffeine were diminished when triazolam was coadministered. Caffeine, at doses that had little effect when administered alone, has been shown to reverse the effects of both lorazepam and triazolam on learning, performance, and ratings of sedation.3536 File et al.37 administered a battery of tests that measured psychomotor and cognitive performance, mood-state, and other factors. Lorazepam impaired psy-chomotor and cognitive performance. On some tests (e.g., digit-symbol substitution), caffeine improved performance and reversed the effect of lorazepam. Caffeine also reversed the self-reported anxiolytic effects of lorazepam. A similar battery of tests was used by Roache and Griffiths38 in a study notable for using a range of doses of both diazepam and caffeine. In a number of tasks, but not all, diazepam and caffeine each blocked the effects of the other. In addition to the type of test, the effects were dependent on the exact doses of each drug.
Somewhat contradictory findings were obtained in two studies using caffeine and diazepam by Ghoneim and colleagues. Loke et al.39 evaluated the effects of a range of doses of each drug as well as combinations of the two using a number of tests. The major effects of diazepam were on cognitive tasks and mood, while caffeine had effects on relatively few tasks. Only in the symbol cancellation test did caffeine ameliorate the impairment produced by diazepam. Ghoneim et al.40 also found little evidence of reversal of diazepam-induced impairment by caffeine.
Johnson et al.41 examined the day-after effects of the benzodiazepines flurazepam and triazolam. Caffeine alone enhanced early morning alertness, but also reversed drowsiness evident in mornings after benzodiazepine administration.
It is apparent that a number of both animal and human studies have provided evidence of a mutually antagonistic relationship between caffeine and benzodiazepines. Such a relationship is most likely to be found where caffeine has an effect opposite to the benzodiazepine. This will depend on the nature of the task investigated as well as doses of the drugs tested. There are also many instances of the failure of caffeine to reverse the effects of benzodiazepines.
There are a number of possible explanations for the interaction between caffeine and benzodiazepines. Ghoneim et al.40 examined the hypothesis that the interaction may be pharmacokinetic in nature. They found that coadmin-istration of caffeine with diazepam resulted in a 22% reduction in plasma diazepam levels. Lau and Falk24 observed some effect of alprazolam on caffeine absorption, but no other pharmacokinetic interaction between the two. The differences observed in each of the studies were of relatively small magnitude and unlikely to account for the mutually antagonistic relationships between caffeine and benzodiazepines.
A number of investigators have examined whether caffeine interacts with the benzodiazepine receptor. Recent evidence suggests that most of the effects of methylxanthines such as caffeine are due to their action as antagonists at adenosine receptors.42 Nevertheless, some evidence has accumulated showing that caffeine interacts with benzodiazepine receptors. Caffeine has weak antagonist properties at these receptors, but this occurs only at relatively high concentrations.43 Lopez et al44 have evidence that caffeine, while not interacting directly with the benzodiazepine receptor site, may alter benzodiazepine function through changes in GABAa activity. Other studies have failed to find evidence of a common substrate for caffeine and benzodiazepines.45
The studies of Lau, Falk, and colleagues on DRL performance and fine motor control in rats would appear to confirm this view. For their measures, caffeine and benzodiazepines both had similar effects, and there was no evidence of one blocking the other. Rather, the relationship was a simple additive one. Such findings argue strongly against a common neural substrate at which caffeine and the benzodiazepines have opposite actions.
Was this article helpful?