The effects of a caffeine placebo and suggestion on blood pressure, heart rate, well-being and cognitive performance.

We studied the effect of suggestion and different instructions in a balanced placebo design. One hundred and fifty-nine subjects were randomized into a 2*4 factorial design. All subjects except a control group received a caffeine placebo. Subjects were randomized to a condition which was identical to an earlier study, or received an information about scientifically proven effects of caffeine (factor 1). The second factor varied instructions: subjects were either made to expect coffee, no coffee or were in a double blind condition and were told either coffee or placebo would be applied. Dependent measures were blood pressure, heart rate, well-being and a cognitive task. There was one main effect of the instruction factor (P=0.03) on diastolic blood pressure, with the group 'told caffeine' reporting significantly smaller decrease in diastolic blood pressure than controls and subjects in the double blind condition. There were no other main effects on both the instruction or suggestion factor, and no interactions. Contrary to the literature, instruction effects were very small. This was apparently due to the fact that placebo-caffeine in the dose used in this study--one cup of strong 'coffee'--did not produce expectancy effects strong enough to affect the parameters measured. It is concluded that the placebo-caffeine research paradigm is not suitable for researching instruction effects in Germany, and that reported effects should be reproduced with tighter controls.