The Placebo Effect of Caffeine on Physiological Responses
Amanda Ruscin
Behavioral Neuroscience 2009
This study was designed to investigate whether the psychological effects of a placebo could influence natural physiological responses. GSR, EEG, and EOG of subjects were measured during a reading task and an object location task during two separate testing days: one control day and one day in which they received placebo caffeine pills. There was no significant difference in GSR or EEG measures between the placebo caffeine or control conditions. GSR was significantly higher in the object location task than in the reading task for both conditions. There were also significantly more saccadic eye movements recorded during the placebo caffeine condition than in the control condition.
Introduction:
Caffeine is the most widely used central nervous system stimulant (Barry, et al., 2005). Once it is ingested, it is absorbed by the gastrointestinal tract and enters the brain. It is commonly known that ingesting caffeine fights fatigue by increasing vigilance, alertness, sustained attention, and arousal (Barry, et al., 2005). However, caffeine also causes increases in individuals’ skin conductance level (Barry, et al., 2005), EEG alpha frequency (which corresponds to a decrease in EEG power) (Barry, et al., 2007), and saccadic eye movements (Miller, 2008). Skin conductance level, also known as galvanic skin response (GSR), is the measure of global sweat gland activity. Sweat gland activity increases with arousal, providing a good measure of the effect of caffeine on the body. EEG, or an electroencephalogram, is a measure of the electrical activity within the brain. As arousal is increased, more neural signals are sent in the brain, increasing electrical activity and subsequent EEG alpha frequency. Finally, EOG, or an electrooculogram, is a measure of saccadic eye movements, which are quick, reflexive eye movements which serve to constantly refresh the image in focus. A recent study has shown that the ingestion of caffeine increases saccadic eye movements (Miller, 2008).
A placebo is any “fake” drug or treatment that has no physiological effect on the body. Research has shown that these inert substances can actually produce various effects within the body, as long as the patient believes strongly enough in its efficacy. It is estimated that about one-third of people treated with a placebo for a given medical condition show improvements based solely on their expectations of the placebo (Nordenburg, 2000). Although the exact biological mechanism underlying the placebo effect is unknown, it has been suggested that beliefs are related to an individual’s physiology through sympathetic and parasympathetic nervous system activation of the limbic system of the brain (an area which has been shown to regulate emotion) (Lanotte, et al., 2005).
While the placebo effect has been demonstrated using many drugs, little research has been conducted investigating the placebo effect of caffeine. However, a recent study by Anderson and Horne demonstrated that decaffeinated coffee, when told that it was caffeinated, resulted in improved performance and increased reaction times (Anderson & Horne, 2008). Therefore, the purpose of this study is to investigate whether the placebo effect of caffeine has similar results as caffeine on other measures. Subjects are tested during two separate days, one in which they complete a reading and object location task, and another in which they complete similar tasks after consuming pills that they were told were caffeine pills. It is hypothesized that by telling subjects they are ingesting caffeine pills, measures of GSR, EEG frequency, and EOG will show global increases.
Methods:
Subjects: The subjects of this study consisted of 7 18-21 year-old female Wofford College students.
Materials: Two Vitamin E pills were given to each subject with water on the experiment day. The pills were kept in an energy/caffeine pill bottle which subjects could see. Utilizing a computer, subjects completed a reading task, consisting of a paragraph explaining either how to change a tire or how to change the oil in a car, and an object location task, consisting of a PowerPoint presentation of random movements of a small blue ball (6 2-second vertical and 6 2-second horizontal movements with 18 seconds of the stationary ball between each). A BIOPAC MP30 acquisition unit, leads, and electrodes were utilized to collect saccadic eye movement data during each task for each participant. A BIOPAC MP30 unit, head cap, and corresponding leads were utilized to collect brain electrical activity during each task for each participant. Finally, a BIOPAC MP30 unit, leads, and electrodes were utilized to collect galvanic skin response data during each task for each participant.
Procedure: Subjects were each tested on two days, after signing a consent form on the first day. During one day of testing, subjects were given two placebo caffeine pills to take with water. Subjects viewed the experimenter removing the pills from a bottle of energy pills and were subsequently instructed that the pills they were ingesting were double the recommended dose and that they were liquid-release, requiring only 15 minutes to take effect. During this time, the subject read a series of three sham abstracts, detailing hypothetical experiments that found that caffeine increased saccadic eye movements, EEG frequency, and galvanic skin response. During this time, subjects gained a knowledge of the physiological responses that caffeine should induce in the human body. The experimenter then placed an EEG cap and electrodes on the subject, following the BIOPAC instruction manual. The subject was instructed to keep their head and hand as still as possible. They then completed a reading task (how to change a tire), followed by an object location task (random movements of a blue ball), during which EEG (of electrodes FP1, FP2, F3, and F4 on the EEG cap), horizontal EOG, and palm GSR were measured. During the control testing day, the subject was recorded in the same manner while completing the reading task (how to change oil) and an object location task (a different series of random ball movements); however, the subject was not provided with any pills to ingest. The day in which the subject received the placebo caffeine pill was counterbalanced, meaning that four subjects received the placebo on their first day of testing, while the other three subjects were control-tested on their first day of testing. Counterbalancing ensures that the order in which subjects are tested (control or placebo first) does not influence the results. EEG data was analyzed with a repeated-measures ANOVA; GSR and EOG data were measured using paired-samples t-tests. Significance was defined as p < 0.05.
Results:
EEG:
The average standard deviations of the EEG alpha power for electrodes FP1 (left), FP2 (right), F3 (left), and F4 (right) during the reading task and during three one-minute segments during the object tracking task were measured for each individual. A condition by electrode repeated-measures ANOVA revealed no main effect of condition (caffeine vs. control) or of electrode. There were also no interactions between components. These results suggest that the caffeine placebo had no effect on EEG measures in any of the recorded brain locations. On the other hand, if one examines the means of the averages of the two left electrodes and the two right electrodes during the reading and object location task across the control and caffeine conditions, one sees that the mean EEG alpha power is decreased in both hemispheres during the caffeine condition, relative to the control condition (although this difference was not statistically significant). Because a decrease in EEG alpha power corresponds to an increase in frequency, this would be the expected trend if the caffeine placebo had an effect on EEG.
Figure 1. Average EEG power of the left and right hemispheres during the control and caffeine conditions
GSR:
Average peak to peak GSR was calculated for each individual during the reading task and during the object location task during control and caffeine conditions. Two paired samples t-tests revealed no significant difference between GSR during the reading task in either condition or during the object location task in either condition. A third t-test was conducted with the average GSR during both tasks in the control condition versus the average in the caffeine condition. This t-test also did not reveal significance. Each individual’s baseline GSR is unique; because of this, each individual’s peak to peak GSR data was normalized to indicate the percent change from the average of reading and object data. A condition (control versus caffeine) by type of movement (reading versus tracking) repeated measures ANOVA was conducted. There was no main effect of condition found; however, a main effect of movement was revealed [F (1,6) = 8.897, p < 0.05]. This data suggests that the caffeine placebo did not have a significant effect on GSR in either the reading or tracking condition. It also indicates, as shown in Figure 2, that GSR is globally increased in the object location task, relative to the reading task.
Figure 2. Average GSR during the reading and tracking tasks of the control and caffeine conditions
* indicates statistical significance
EOG:
The average duration of time each individual spent looking at the dot in each of the six horizontal movements was measured; the average peak to peak of each of these six movements was also measured. Using these two values, the average speed of the movement was calculated for each individual. A paired samples t-test revealed no significant difference in the average speed between each condition, as shown if Figure 3. This suggests that the caffeine placebo did not have an effect on the speed of eye movement.
Figure 3. Average speed of the horizontal eye movement in the object tracking task during the control and caffeine condition
The average number of saccades during the reading task was measured for each individual during each condition. A paired samples t-test revealed a significant increase in the number of saccades during the caffeine condition, relative to the control condition t(6)=0.05, p<0.05. This suggests that the caffeine placebo did increase the number of saccadic eye movements, as shown in Figure 4.
Figure 4. Average number of saccadic eye movements during the reading task in the control and caffeine condition
* indicates statistical significance
Discussion:
A significantly higher number of saccadic eye movements were recorded during the reading task in the caffeine condition than during the reading task in the control condition. This supports the hypothesis that a caffeine placebo would have similar effects as actual caffeine and that the number of saccades would increase. It is interesting that the caffeine placebo produced an effect on saccadic eye movements, given that saccades are completely reflexive movements which one cannot consciously control. Because the placebo had an effect on such an autonomic physiological response, it is possible that further investigation of the effect of caffeine and caffeine placebos on saccadic eye movements would yield valuable information about the mechanism behind placebo effects.
The hypothesis that EEG frequency and GSR would increase with the caffeine placebo was not supported, due to a lack of significance within the data. It is highly possible that a greater sample size would have produced significance in the EEG data, given that the current data appears to show a general trend in decreasing power (and subsequently, increasing frequency) in the caffeine condition. The GSR recording device did not work properly for all subjects, providing a possible reason that significance was not obtained between conditions in this measure. The fact, therefore, that significant GSR differences did appear between tasks speaks to the power of this statistic. It is possible that this difference occurs because individuals are asked to be vigilant in their observation of the ball and that the movement of the ball produces a stress-like response. In contrast, the paragraph the subjects were instructed to read was intended to be as mundane as possible, eliciting little to no emotional response.
Subjects were not instructed to limit actual caffeine use on testing days, which may pose a possible confound. It is possible that they consumed more actual caffeine on the day of their caffeine placebo testing, which would increase measures in the expected direction, regardless of a placebo effect. However, because the days in which subjects received each condition were counterbalanced, it is unlikely that this was a systematic difference.
Future research in this area should control for the amount of actual caffeine consumed on testing days and obtain a larger subject pool, although it is difficult given the fact that subject must come in for two separate days of testing. Also, the strength of the placebo effect of caffeine on saccadic eye movements should be further investigated, given the possible beneficial information it could provide about underlying placebo effect mechanisms.
References:
Anderson, C. & Horne, J.A. (2008). Placebo response to caffeine improves reaction time
performance in sleepy people. Human Psychopharmacology: Clinical and Experimental, 23(4). 333-336.
Barry, R.J., Clarke, A.R., Johnstone, S.J., & Rushby, J.A. (2007). Timing of caffeine’s impact on
autonomic and central nervous system measures: Clarification of arousal effects. Biological Psychology, 77(3). 304-316.
Barry, R.J., Rushby, J.A., Wallace, M.J., Clarke, A.R., Johnstone, S.J., & Zlojutro, I. (2005).
Caffeine effects on resting-state arousal. Clinical Neurophysiology, 116. 2693-2700.
Lanotte, M., Lopiano, L., Torre, E., Bergamasco, B., Colloca, L., & Benedetti, F. (2005).
Expectation enhances autonomic responses to stimulation of the human subthalamic limbic region. Brain, Behavior, and Immunity, 19(6). 500-509.
Nordenberg, T. (2000). The healing power of placebos. FDA Consumer Magazine, 34(1).