Effect of dose on the ability of caffeine to serve as a reinforcer in humans

This study tested the effects of dose on the reinforcing effects of caffeine in humans. Eight moderate coffee drinkers were given concurrent access to decaffeinated coffee vs. decaffeinated coffee to which different doses of caffeine (25, 50, 150 and 200mg/cup) were added. Subjects were tested across several independent double-blind trials. The coffees with 25mg of caffeine were repeatedly self-administered at a rate greater than that of decaffeinated coffee in two of six subjects, the 50mg dose in four of eight subjects, the 150mg dose in three of six subjects, and the 200mg dose in none of the three subjects tested. Headaches, drowsiness and fatigue occurred with use of decaffeinated coffee in five subjects. When these symptoms occurred, there was a greater probability of self-administration of the caffeinated coffee. We conclude that doses of caffeine similar to those in tea or soda can serve as reinforcers.     

Forced-choice versus free-choice procedures: Caffeine self-administration in humans

Methodological comparisons of procedures for drug self-administration are rare. In studies examining the reinforcing effect of caffeine in humans, caffeine self-administration usually has been inferred from performance under forced-choice procedures. In the present experiment, caffeine self-administration via coffee was compared under forced-choice and free-choice conditions; i.e., when subjects were and were not required to use a minimum number of coffees. Ten moderate coffee drinkers (2–7 cups/day) were assigned to forced- and free-choice conditions using a randomized cross-over design. Under each choice condition, subjects completed six independent, double-blind trials, consisting of a 2-day exposure period followed by a 2-day test period. During exposure, subjects consumed either decaffeinated or caffeinated (100 mg/serving) coffee on day 1 and the other coffee on day 2. During the test period, subjects had concurrent access to the same decaffeinated and caffeinated coffees. Under the forced-choice condition, subjects were required to drink at least four cups of coffee per day during the test period. Under the free-choice condition, subjects did not have a minimum-cup requirement. In general, the relative rate at which subjects self-administered caffeinated versus decaffeinated coffee was similar across choice conditions, even though subjects self-administered significantly fewer cups of both coffee types under the free-choice than the forced-choice condition. These results suggest that, at least for caffeine, forced-choice and free-choice procedures produce comparable results. Whether this finding generalizes to a context in which caffeine or another drug is more robustly self-administered, remains to be determined.Supported by grant DA-04843, T32-DA07242, and Research Scientist Development Award DA-00109 (J.R.H) from the National Institute on Drug Abuse.     

Effects of Caffeine and Noise on Mood,Performance and Cardiovascular Functioning

An experiment was conducted to examine the effects of caffeine and noise on mood, mental performance and cardiovascular function. One hundred and six young adults (mean age 21·2 years) took part in the study. Subjects were assigned to one of six groups formed by combining noise/quiet and drink (caffeinated coffee, decaffeinated coffee and fruit juice) conditions. Subjects were familiarized with the tasks and then completed a pre-drink baseline session (conducted in the quiet). Subjects were then given either caffeinated coffee (1·5 mg/kg caffeine), decaffeinated coffee or fruit juice. Following consumption of the drink subjects were re-tested 1 h later, either in noise (75 dBA conglomerate noise, consisting of speech, music and machinery noise) or in quiet. The subjects exposed to noise felt more anxious and showed an increase in blood pressure. Their performance of a cognitive vigilance task also declined over time. There were no significant main effects of caffeine, although simple reaction time was quickest in the caffeinated coffee group. Caffeine did not modify the effects of noise on mood, cardiovascular functioning or sustained attention. Indeed, the only interaction between drinks and noise was found in recall and recognition memory tasks, with the caffeine/noise group having better memory performance than the decaffeinated/noise subjects. Overall, the results show that low levels of caffeine do not increase the behavioural and physiological changes observed in a stressful situation. © 1997 John Wiley & Sons, Ltd.     

Anti-genotoxicity and glutathione S-transferase activity in mice pretreated with caffeinated and decaffeinated coffee.

In vivo anti-genotoxic effects of caffeinated and decaffeinated instant coffee were compared in mice after pretreatment either by gavage for 10 consecutive days or in the drinking water for 2 weeks. Changes in hepatic sulfhydryl (-SH) content and glutathione S-transferase (GST) activity were evaluated in pretreated animals. Both caffeinated and decaffeinated instant coffee induced a moderate increase in -SH content and GST activity following pretreatment (with 70, 140 and 280 mg/kg body weight) by gavage for 10 days. This enhancement was not always dose dependent. The maximum effect on GST activity was observed at a dose of 140 mg/kg body weight/day. However, such an effect was not observed after administration of drinking water containing 2% caffeinated/decaffeinated instant coffee for 2 weeks. Results of the bone marrow micronucleus test for evaluating genotoxic effects revealed that both caffeinated and decaffeinated instant coffee (140 mg/kg body weight/day) could exert significant anti-genotoxic effects against ip injected benzo[a]pyrene (BP), cyclophosphamide (CPH), 7,12-dimethylbenz[a]anthracene (DMBA), mitomycin C (MMC) and procarbazine (PCB) in animals pretreated by gavage. Anti-genotoxic effects against BP, DMBA and urethane (URE) were evaluated in animals that received drinking water containing 2% caffeinated/decaffeinated instant coffee for 2 weeks. With the exception of the anti-genotoxic effect of decaffeinated coffee against DMBA, there was no significant change in genotoxicity after the above pretreatment. From this work, there is no evidence for any significant difference in the in vivo anti-genotoxicity of caffeinated and decaffeinated instant coffee.     

Instant and brewed coffees in the in vitro human lymphocyte mutagenicity test

Incubation of instant and 'home brew' coffees (caffeinated and decaffeinated) and of coffee aroma with cultured human lymphocytes in the presence and absence of S-9 increased the number of total aberrations. However, the increase was smaller in the presence of S-9 than in its absence. Pure caffeine tested with or without S-9 at doses equivalent to levels in caffeine-containing coffee did not give statistically significant increases of any type of aberration when compared with controls. In all in vitro test systems used to date, coffee and coffee aroma or their reactive compounds were metabolically deactivated in the presence of S-9. This could explain the negative results obtained in mutagenicity assays in vivo.     

Coffee contains cholinomimetic compound distinct from caffeine. I: Purification and chromatographic analysis

Both regular and decaffeinated coffees were found to have cholinomimetic actions when tested in urethane-anesthetized rats. These actions were distinct from those of caffeine and reversible by atropine. The bioactive fraction was purified from alcoholic extracts of instant decaffeinated coffee by liquid column chromatography and preparative TLC. The purified compound showed similar pharmacological actions as the starting material. Chromatographic behavior was further characterized by analytical TLC and HPLC. Chromatographic analyses of extracts of green coffee beans and roasted ground coffees showed that the cardioactive compound was only present in roasted coffees. Similar analyses of other commonly consumed beverages, including teas and cocoa, showed that this compound was not present in beverages besides coffee.     

The acute physiological and mood effects of tea and coffee: the role of caffeine level

The objective of this study was to determine the effect of caffeine level in tea and coffee on acute physiological responses and mood. Randomised full crossover design in subjects after overnight caffeine abstention was studied. In study 1 (n = 17) the caffeine level was manipulated naturalistically by preparing tea and coffee at different strengths (1 or 2 cups equivalent). Caffeine levels were 37.5 and 75 mg in tea, 75 and 150 mg in coffee, with water and no-drink controls. In study 2 (n = 15) caffeine level alone was manipulated (water, decaffeinated tea, plus 0, 25, 50, 100, and 200 mg caffeine). Beverage volume and temperature (55 degrees C) were constant. SBP, DBP, heart rate, skin temperature, skin conductance, and mood were monitored over each 3-h study session. In study 1, tea and coffee produced mild autonomic stimulation and an elevation in mood. There were no effects of tea vs. coffee or caffeine dose, despite a fourfold variation in the latter. Increasing beverage strength was associated with greater increases in DBP and energetic arousal. In study 2, caffeinated beverages increased SBP, DBP, and skin conductance and lowered heart rate and skin temperature compared to water. Significant dose-response relationships to caffeine were seen only for SBP, heart rate, and skin temperature. There were significant effects of caffeine on energetic arousal but no consistent dose-response effects. Caffeinated beverages acutely stimulate the autonomic nervous system and increase alertness. Although caffeine can exert dose-dependent effects on a number of acute autonomic responses, caffeine level is not an important factor. Factors besides caffeine may contribute to these acute effects.     

Influence of genetic polymorphisms and habitual caffeine intake on the changes in blood pressure,pulse rate,and calculation speed after caffeine intake: A prospective,double blind,randomized trial in healthy volunteers

The aim of this study was to investigate the contribution of gene polymorphisms, in combination with habitual caffeine consumption, to the effect of caffeine intake on hemodynamic and psychoactive parameters. A double-blind, prospective study was conducted with 201 healthy volunteers randomly allocated 2:1 to the caffeinated group (150 mL decaffeinated coffee with additional 200 mg caffeine) or decaffeinated group (150 mL decaffeinated coffee). We measured the changes in blood pressure (BP) and calculation speed upon coffee intake, stratifying with gene polymorphisms, e.g., those in adenosine A_2A receptor (ADORA2A) and cytochrome P450 (CYP) 1A2, and daily caffeine consumption (≤90 mg/day and >90 mg/day). Overall, caffeine intake independently increased BP and calculation speed (p-values < 0.05), irrespective of the polymorphisms. In stratified analysis, a statistical significance within the caffeinated group was observed for the change in systolic BP in the stratum of CYP1A2 polymorphism with daily caffeine consumption ≤90 mg/day: change in systolic BP in the CYP1A2 rs762551 CC group (mean ± SD = 11.8 ± 5.9) was higher than that in the AA/CA group (4.1 ± 5.5). Gene polymorphisms may limitedly modify the effect of caffeine intake on hemodynamic parameters in combination with habitual caffeine consumption.     

The effects of black tea and other beverages on aspects of cognition and psychomotor performance

Nineteen healthy volunteers ingested 400 ml black tea, coffee, caffeinated water, decaffeinated tea or plain water on three occasions through the day (0900, 1400 and 1900 hours). A 2 × 2 factorial design with caffeine (0, 100 mg) and beverage type (water, tea) was employed, with coffee (100 mg caffeine) as a positive internal control, based on a five-way crossover. A psychometric test battery comprising critical flicker fusion (CFF), choice reaction time (CRT), short-term memory (STM) and subjective sedation (LARS) was performed at regular intervals throughout the day, and intensively so immediately following each beverage. Consumption of tea compared to water was associated with transient improvements in performance (CFF) within 10 min of ingestion and was not affected by the time of day. Caffeine ingestion was associated with a rapid (10 min) and persistent reduction in subjective sedation values (LARS), again independent of time of day, but did not acutely alter CFF threshold. Over the whole day, consumption of tea rather than water, and of caffeinated compared to decaffeinated beverages, largely prevented the steady decline in alertness (LARS) and cognitive capacity observed with water ingestion. The effects of tea and coffee were similar on all measures, except that tea consumption was associated with less variation in CFF over the whole day. No significant treatment effects were apparent in the data for the STM. Tea ingestion is associated with rapid increases in alertness and information processing capacity and tea drinking throughout the day largely prevents the diurnal pattern of performance decrements found with the placebo (no caffeine) condition. It appears that the effects of tea and coffee were not entirely due to caffeine per se; other factors either intrinsic to the beverage (e.g. sensory attributes or the presence of other biologically active substances) or of a psychological nature (e.g. expectancy) are likely to play a significant role in mediating the responses observed in this study. Received: 18 September 1997/Final version: 16 February 1998     

Caffeine as an intensifier of stress-induced hormonal and pathophysiologic changes in mice

Psychosocially stressed male mice competing in a Henry-Stephens complex population cage develop hypertension, cardiovascular damage, and chronic interstitial nephritis. Their plasma renin, noradrenaline, corticosterone, and adrenal-catecholamine synthetic enzymes are increased and they die prematurely. Adding 3.3 mg of caffeine a day per kilogram of mouse body weight (the equivalent of 20 μg/ml decaffeinated coffee) to their drinking water significantly intensifies most of these changes. A dose of 90 mg/kg of caffeine (the equivalent of 560 μg/ml, i.e., brewed tea or coffee) further increases the effects. The drug-induced enhancement of competitive social stimulation of the neuroendocrine system resulted in a further increase of plasma renin and corticosterone levels as well as blood pressure and adrenal weight. These effects together with accelerated mortality and increased pathology indicate that chronic consumption of caffeinated liquids adds to the risks of psychosocial stress.     

Placebo response to caffeine improves reaction time performance in sleepy people

Caffeine is the most widely used stimulant to counteract sleepiness. However, little is known about any placebo effect of caffeine in sleepy people and the effect of suggestibility. Over a 95 min test period, and in a counterbalanced design, 16 young healthy adults underwent 3 x 30 min sessions at the psychomotor vigilance test (PVT), during an early afternoon 'dip' enhanced by a prior night's sleep restriction (5 h). On both occasions they were given a cup of a decaffeinated coffee; once when the participant was verbally primed to suggest the coffee was caffeinated (Placebo) and on the other under neutral priming (Control). There were significantly fewer lapses and shorter reaction times following Placebo, for the initial two 30 min sessions, indicating that suggestion about consuming caffeine was effective in improving performance in moderately sleepy people.     

An examination of consumer exposure to caffeine from retail coffee outlets

Objective

To analyse the distribution of caffeine doses obtainable from espresso coffee sold by a sample of commercial coffee vendors located on the Gold Coast, Qld, Australia.

Design

A cross section of “Espresso/short black” coffee samples were purchased and analysed for their caffeine content using micellar electrokinetic capillary chromatography (MEKC). Coffees were collected using systematic cluster sampling across five major retail centres.

Results

Ninety-seven espresso samples were analysed. The mean (±SD) quantity of caffeine was 106 ± 38 mg/serve with a concentration of 2473 ± 1092 mg/l. There was considerable variation in caffeine content. The range per serve was 25–214 mg whilst the concentration range was 580–7000 mg/l. Twenty-four samples (24.7%) contained 120 mg of caffeine or higher and 12 samples (12.3%) exceeded 167 mg per serve.

Conclusions and implications

The number of heavily caffeinated samples differentiates these findings from frequently cited caffeine values and supports similar data recently collected throughout the United Kingdom. As a result, the accuracy of any previous intake modelling regarding caffeine use in the Australian population is in doubt. The present data suggests that the probability of consumer exposure to high caffeine doses is greater than previously anticipated. Greater sample numbers from a broader selection of venues is required to confirm the extent of caffeine content variance within retail ground coffees.     

The effect of caffeine on human performance,alone and in combination with ethanol

The effect of caffeine (300 mg/70 kg) on cognitive, perceptual and motor functions was investigated both alone and in combination with ethanol (0.75 g/kg) in 68 healthy student volunteers of both sexes. A test battery consisting of standing steadiness, simple and complex reaction time, manual dexterity, numerical reasoning, perceptual speed and verbal fluency was used. Placebos for both drugs were included. Caffeine was administered in decaffeinated coffee immediately after finishing drinking the alcoholic beverage. A peak plasma ethanol concentration of 92 ± 4 mg/100 ml occurred at 40 min which was not modified by caffeine. Caffeine did not antagonise the ethanol-induced decrement in performance except in the reaction time tests. Caffeine alone caused a significant increase in body sway at 40 min.     

Sex differences in estimation of time intervals and in reaction time are removed by moderate but not high doses of caffeine in coffee

Estimation of the passage of time in the seconds-to-minutes range and reaction time are strongly dependent on a hypothetical internal clock. Dopamine is the neurotransmitter most closely related to the rate of this clock. Caffeine, probably the most consumed drug in the world, leads to an augmentation of dopamine neurotransmission. In this study coffee, which reproduces the conditions under which caffeine is normally ingested, containing 3, 75, 150 or 300 mg of caffeine, was given to healthy male and female volunteers. A computerized time estimation and reaction time test was carried out 50 min after ingestion. Sex differences in placebo control subjects (who took decaffeinated coffee with 3 mg of caffeine), with females making more accurate estimates of time intervals than males and males showing shorter reaction times than females, were removed in subject taking doses of 75 and 150 mg of caffeine in the case of time estimation and 150 mg in the case of reaction time. The 300 mg dose induced overestimation of time in females and shortened the reaction time in males. There were no sex differences in the pharmacokinetics of caffeine, as measured in salivary concentration of caffeine using high-performance liquid chromatography. Results indicating sex differences in time estimation or reaction time should not be generalized from the laboratory to real life without considering the fact that everyday coffee consumption may eliminate these differences. Copyright 2001 John Wiley & Sons, Ltd.     

Oral caffeine consumption by rats: the role of flavor history, concentration, concurrent food, and an adenosine agonist.

Some determinants of caffeine consumption by rats were examined using the two-bottle choice test. To describe the role of flavor history, groups of eight rats each received one of three fluids as their only source of fluid beginning at 29 days of age and continuing throughout the experiments. One group ("water") received tapwater, a second group ("caffeine") received 0.5 mg/ml caffeine in tapwater, and a third group ("quinine") received 0.01 mg/ml quinine in tapwater. Two-bottle choice tests began when rats were 40 days old. In the initial tests, caffeine rats drank more caffeinated water than water rats. Quinine rats were midway between these two groups. On a second block of tests, quinine and water rats' caffeine consumption increased so that the three groups were indistinguishable. When 0.5 mg/ml caffeine was available for 24 h, about one third of the total fluid consumption was of caffeinated water for all three groups. The presence of food greatly increased both caffeine and water consumption across a range of caffeine concentrations spanning 0.125-4.0 mg/ml. Increasing caffeine concentration generally increased consumption of plain water and decreased that of caffeinated water (but not total caffeine consumed) for water rats. Caffeine rats generally drank more caffeine than water rats, largely due to a tendency toward increased consumption of the 0.5-mg/ml concentration. Consumption of caffeinated water peaked at 0.5 mg/ml and showed graded decreases at higher and lower concentrations. Caffeine consumption showed dose-related increases with presession administration of l-phenylisopropyl adenosine. The serines of experiments characterize some of the determinants of caffeine consumption in rats.(ABSTRACT TRUNCATED AT 250 WORDS)     

Investigation of coffee in Drosophila genotoxicity tests

Two preparations of coffee (instant coffee and freeze-dried home-brew coffee) were tested in different mutagenicity assays in germ cells as well as in somatic cells of Drosophila melanogaster. The three end-points assayed in germ cells were sex-linked recessive lethals (mainly gene mutations and small chromosome aberrations), dominant lethals (cytotoxic effects as well as genotoxic effects) and sex-chromosome losses (chromosome breakage and non-disjunction). The aqueous coffee solutions were fed either to adult male flies for 3 days or to growing larvae during the whole larval development. Treated males were crossed with appropriately marked females, and the different genetic end-points were analysed in the F1 or F2 generation. The test concentrations (instant coffee 4% (w/v), home-brew coffee 3%) were acutely toxic in adult males (killing approximately 75 and 90% of the exposed flies, respectively). No increase in deaths was caused in larvae by the same concentrations. Only cytotoxic effects were observed in the test for dominant lethals. No conclusive genotoxic effects could be detected in any of the three germ cell assays. The coffee preparations were also tested for induction of mutation and mitotic recombination in somatic cells of the wing imaginal disc. Larvae trans-heterozygous for two recessive wing hair markers were fed high concentrations of the coffees for varying periods of time. Wings of surviving adult flies were analysed for mosaic spots. Twin spots exhibiting both mutant phenotypes are produced by mitotic recombination; single spots showing one or the other phenotype are the result of somatic mutation, such as gene mutation or deletion, or of mitotic recombination. Both coffees had weak effects on normal (repair-proficient) cells as well as on excision repair-defective cells in this assay. Additional experiments with pure caffeine and decaffeinated coffee show that these weak effects in somatic cells were most probably caused by the caffeine present in the two coffees.     

Variability in caffeine consumption from coffee and tea: possible significance for epidemiological studies

Five surveys, using a previously developed high-performance liquid chromatography procedure to measure caffeine concentrations, indicated great variations in the concentrations of caffeine in tea and coffee. In the study of beverages prepared at home, data on caffeine concentrations in 58 samples of tea and coffee, volumes of cups, and numbers of cups consumed/day, indicated that the range of caffeine intakes for the women participating was 49-1022 mg/day. There were considerable day-to-day variations in caffeine contents in coffee samples from some commercial coffee shops. When 17 samples of five national brands of instant coffee were made into beverages in the laboratory, variations in caffeine concentrations between lots were small but between brands were significant. A considerable range of caffeine concentrations was also found when 12 samples of coffee prepared at work by different individuals using the same jar of instant coffee were analysed. Analysis of tea samples prepared in the laboratory indicated that steeping time had an important influence on resulting caffeine and theobromine concentrations. People preparing their own beverages were found to drink more liquid than the volume offered commerically. The mean caffeine 'contents' of home-made coffee and of coffee prepared by individuals at work were 79.4 and 81.7 mg/cup respectively, indicating a mean intake of approximately 80 mg caffeine/cup. When this amount (80 mg/cup) was used to estimate daily intakes of caffeine from coffee, on the basis of the number of reported cups/day, and the values obtained were compared with the amounts actually consumed by individuals, the potential for misrepresentation of individual consumption became obvious. For example, for subjects consuming three cups of coffee, only 25% would have been correctly categorized in the expected range for the daily intake of caffeine, 39% would have been overestimated and 36% underestimated for the amount of caffeine consumed. These variations in caffeine concentrations and in the volume of coffee consumed have frequently been ignored in examinations of the possible relationship between coffee consumption and various health problems, and this could perhaps partly explain some conflicting results seen in epidemiological studies.     

A naturalistic investigation of the effects of day-long consumption of tea, coffee and water on alertness, sleep onset and sleep quality

Rationale: The effects of caffeine, especially caffeinated coffee, on human performance have been extensively studied. However, few studies have been naturalistic representations of how tea/coffee is normally consumed in terms of dose and time of consumption. Objectives: This study investigated the effects of day-long consumption of tea, coffee and water on cognitive and psychomotor performance, and sleep quality at night. Methods: Thirty healthy volunteers received equal volume drinks equivalent to either 1 or 2 cups of tea (containing 37.5 mg or 75 mg caffeine), or coffee (75 mg or 150 mg caffeine), or water, in a randomised five-way crossover design. Drinks were administered on four occasions during the day (0900, 1300, 1700 and 2300 hours). A psychometric battery consisting of critical flicker fusion (CFF), choice reaction time (CRT) and subjective sedation (LARS) tests, was administered pre-dose and at frequent time points post-dose. The Leeds Sleep Evaluation Questionnaire (LSEQ) was completed each morning and a wrist actigraph was worn for the duration of the study. Results: Caffeinated beverages maintained CFF threshold over the whole day (P<0.05), independent of caffeine dose or beverage type. During the acute phase of beverage ingestion, caffeine significantly sustained performance compared to water after the first beverage for CFF and subjective sedation (P<0.05), and after the second beverage for the Recognition component of the CRT task (P<0.05). Additionally, there were significant differences between tea and coffee at 75 mg caffeine after the first drink. Compared to coffee, tea produced a significant increase in CFF threshold between 30 and 90 min post-consumption (P<0.01). However, following the second beverage caffeinated coffee at 75 mg significantly improved reaction time (P<0.05), compared to tea at the same dose, for the Recognition component of the CRT task. Caffeinated beverages had a dose dependent negative effect on sleep onset (P<0.001), sleep time (P<0.001) and sleep quality (P<0.001). Conclusions: These results indicate that ingestion of caffeinated beverages may maintain aspects of cognitive and psychomotor performance throughout the day and evening when caffeinated beverages are administered repeatedly. This study also demonstrates that day-long tea consumption produces similar alerting effects to coffee, despite lower caffeine levels, but is less likely to disrupt sleep. Other differences between tea and coffee were more subtle, and require further investigation. Received: 16 February 1999 / Final version: 20 December 1999     

Caffeine expectancies influence the subjective and behavioral effects of caffeine

Objectives

This study investigated the independent and interactive effects of caffeine pharmacology and expected effects of caffeine on performance and subjective outcomes.

Methods

Abstinent coffee drinkers (n?=?60) consumed decaffeinated coffee with either 280 mg or 0 mg added caffeine. Caffeine dose was crossed with varying instructions that the coffee would either enhance or impair performance in a 2?×?2 factorial design. Performance, mood, caffeine withdrawal, and negative somatic effects were assessed.

Results

Relative to placebo, caffeine improved reaction time and accuracy on the rapid visual information processing task, a measure of vigilance. However, there was a significant dose by expectancy interaction that revealed that among participants given placebo coffee, “impair” instructions produced better performance than “enhance” instructions. Caffeine also improved psychomotor performance as indicated by a finger tapping task with no main effects of expectancy or interactions. Impair instructions produced greater reports of negative somatic effects than enhance instructions, but only when caffeine was administered.

Conclusions

Manipulating the expected effects of caffeine altered the behavioral and subjective effects of caffeine. A significant dose by expectancy interaction revealed a somewhat paradoxical outcome in the placebo conditions whereby those told “impair” performed better than those told “enhance.” This may reflect compensatory responding as has been observed in similar studies using alcohol (Fillmore et al. Psychopharmacology 115:383–388, 1994). Impair instructions led to greater negative somatic effects only when caffeine was administered supporting the active placebo hypothesis.