Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Caffeine and nicotine are among the most widely used mood-altering drugs in the world. Caffeine is commonly consumed orally in various caffeine-containing foods and beverages such as coffee, tea, soft drinks, and chocolate (Gilbert, 1984). Nicotine, the primary active constituent of tobacco that leads to addiction (U.S. Department of Health and Human Services, 1988; Benowitz, 1996), is commonly self-administered by smoking or chewing tobacco products.

Both caffeine and nicotine are psychomotor stimulants, which may share a common dopaminergic component of central action with classic psychomotor stimulants such as cocaine. It is well documented that cocaine primarily produces its reinforcing and psychostimulant actions through inhibition of dopaminergic reuptake (Ritz et al., 1987; Bergman et al., 1989). A substantial line of evidence also suggests that the reinforcing and psychostimulant effects of caffeine are also mediated by dopamine (Garrett and Griffiths, 1997). Caffeine, via antagonism of adenosine receptors, is proposed to enhance dopaminergic activity by removing a negative modulatory influence of adenosine from dopamine receptors, which are colocalized with adenosine receptors on striatal neurons (Ferré et al., 1992). Similarly, nicotine is proposed to produce its reinforcing and psychostimulant effects by enhancing dopaminergic activity through blockade of dopamine reuptake (Izenwasser et al., 1991) and increasing synaptic dopamine release (Courtney et al., 1991; Benowitz, 1996; Nisell et al., 1997).

Consistent with a shared dopaminergic mechanism of action, human laboratory studies examining intravenous administration of caffeine and nicotine suggest that they produce central effects similar to prototypically abused stimulants such as amphetamine and cocaine. Like intravenous cocaine administration, intravenous nicotine produces dose-related increases in various ratings of positive subjective effects (e.g., ratings of drug liking and high) and stimulant effect, as well as dose-related increases in heart rate and blood pressure in subjects with histories of drug abuse and cigarette smoking (Henningfield et al., 1985; Jones et al., 1999). Similarly, intravenous caffeine administration has been shown to produce dose-related increases in ratings of positive subjective effects, including drug liking and high, as well as increases in blood pressure in subjects with histories of drug abuse and caffeine consumption (Rush et al., 1995).

Although both preclinical and clinical evidence support the idea that nicotine and caffeine have similar pharmacologic stimulant profiles, these drugs have not been directly compared in humans under double-blind procedures and using the same route of drug administration. The aim of the present study was to extend our knowledge about the comparative pharmacology of nicotine and caffeine as stimulant drugs by directly comparing the subjective and physiological effects of intravenous caffeine and nicotine on measures previously shown sensitive to the effects of these drugs as well as cocaine (Preston et al., 1993; Jones et al., 1999) in subjects with histories of using tobacco cigarettes, caffeine, and cocaine. The intravenous route of administration was used because it permits blind administration while producing similar rapid onset of effects.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. Nine volunteers (two women and seven men) between the ages of 28 and 39 years, were recruited through newspaper advertisements and word of mouth. For inclusion in the study, subjects had to report a minimum of 6 months of cocaine abuse and report use of either smoked or intravenous cocaine at least 2 days a week for the previous 6 weeks. All subjects had a diagnosis of cocaine dependence (American Psychiatric Association, 1994) based on a structured interview (Structured Clinical Interview for DSM-IV Axis I Disorders; First et al., 1996). All subjects were regular caffeine consumers. Subjects also smoked at least 20 tobacco cigarettes per day for at least 1 year before participation and had a Fagerstrom questionnaire score (a measure of nicotine dependence and tolerance) of at least 6 and/or an afternoon carbon monoxide level of at least 10 parts/million at screening. All subjects were in good health, without significant medical or psychiatric illness except for drug abuse and nicotine dependence. Before participation, subjects were screened for medical problems and drug use via medical history, physical examination, laboratory tests of blood chemistry, electrocardiography (ECG) results, blood pressure measurement, and urinalysis. A battery of psychiatric instruments was used to screen for psychiatric disorders. Subjects were not enrolled in the study if they had histories of seizure disorders, hypertension, abnormal ECG, significant risk factors for heart disease, or poor venous access. Women were excluded from the study if they were pregnant.

Drug Preparation and Administration. Doses of caffeine sodium benzoate (Pasadena Research Laboratories, San Clemente, CA) and nicotine hydrogen tartrate (Gallard-Schlesinger Chemical Co., Carle Place, NY) were aseptically prepared individually for injection by diluting the drugs in sterile saline (0.9% sodium chloride). All caffeine and nicotine doses were calculated as milligrams of caffeine and nicotine base. The placebo dose was sterile saline. Placebo, caffeine (100, 200, and 400 mg/70 kg) and nicotine (0.75, 1.5, and 3.0 mg/70 kg) were infused through an indwelling intravenous catheter in a total volume of 5 ml at a 10-s infusion rate. All drug doses were infused manually by a physician.

Study Design and General Procedures. This double-blind study was conducted while subjects resided in a residential research facility at the Behavioral Pharmacology Research Unit of the Johns Hopkins University School of Medicine, for approximately 4 weeks. Before admission and obtaining informed consent, subjects were informed that the objective of the study was to learn more about the behavioral effects of certain drugs. Subjects were allowed to acclimate to the residential unit for a few days during which time written consent for research participation was obtained. Although cigarette smoking was permitted for the duration of the study, subjects were restricted from smoking for at least 8 h before each session. Carbon monoxide levels were assessed at baseline (at least 8 h before each session) and immediately before each session to verify compliance with the smoking restrictions. For the duration of the study, all dietary sources of caffeine were eliminated in keeping with a policy that completely restricts dietary caffeine intake for all subjects on the residential unit. For consistency with the nicotine intake, subjects were administered caffeine (150 mg/70 kg b.i.d.) in capsules, with the last dose administered 15 to 18 h before each session.

Dose Run-up Sessions. The purpose of the dose run-up sessions was to determine the safety and tolerability of the drug doses. Before the start of each of these sessions, an intravenous catheter was inserted into the dominant arm. A slow drip intravenous line was maintained during each session. During dose run-up sessions, subjects received two or three injections of placebo, caffeine (100, 200, and 400 mg/70 kg) or nicotine (0.75, 1.5, and 3.0 mg/70 kg) in ascending dose order (Table 1). The order of exposure to caffeine and nicotine was counterbalanced across subjects (half the subjects received the caffeine dose run-up first, and the other half received the nicotine dose run-up first). All doses of each drug were administered at a 10-s infusion rate with a 60-min interval between infusions. Each dose run-up session was separated by at least 48 h.

Experimental Sessions. Following the dose run-up sessions, seven experimental sessions were conducted up to 5 days per week (Monday through Friday). Before the start of each experimental session, an intravenous catheter was inserted into the dominant arm. During each experimental session, a single dose of placebo, caffeine (100, 200, or 400 mg/70 kg) or nicotine (0.75, 1.5, or 3.0 mg/70 kg) was administered in mixed sequence (Table 1). All doses were administered at a 10-s infusion rate. Each experimental session was separated by at least 24 h. Data collection in the experimental sessions was the same as that in the dose run-up sessions.

Visual Analog Scales. During sessions subjects responded to questions asking, "Do you feel a drug effect?", "Does the drug have any good effects?", "Does the drug have any bad effects?", "Do you like the drug?", "How high are you?", "How drowsy/sleepy are you?", "How alert/energetic are you?", "Do you feel jittery?", "Do you feel relaxed?", "Do you feel stimulated?", and "Do you feel a rush?". Subjects responded by positioning an arrow along a 100-mm line marked from 0 to 100 with 0 being "Not at All" and 100 being "Extremely". These visual analog scales were completed once before the drug injection and every 2 min for 30 min after the injection.

Addiction Research Center Inventory (ARCI). Subjects completed the short form of the ARCI, which is a 49-item questionnaire comprised of five subscales: A (amphetamine)an amphetamine scale that provides an assessment of amphetamine-like effects; BG (benzedrine group)an amphetamine-sensitive scale that provides a measure of benzedrine-like effects, intellectual efficiency, and energy; LSD (lysergic acid diethylamide)a scale that provides a measure of dysphoria and somatic complaints; MBG (morphine-benzedrine group)a scale that provides a measure of euphoria; and PCAG (pentobarbital-chlorpromazine alcohol group)a scale that provides a measure of sedation. Subjects completed the ARCI once before the drug injection and at approximately 35 min after the injection. Because it was anticipated that most drug effects would have dissipated by about 20 min after drug administration, subjects were instructed to answer the questions on the ARCI retrospectively for how they felt since receiving the drug injection.

Pharmacological Class Identification Questionnaire. Approximately 40 min after each drug injection, subjects completed the Pharmacological Class Identification Questionnaire on which they were asked to select the drug class that best described which drug they had received that day. After participants selected the drug class option, the computer screen displayed the names of specific drugs of that drug class. Subjects then chose, from the list of specific drugs, which compound was most similar to the drug they had received. The drug class options, with specific drugs in brackets, were sedatives or muscle relaxants [diazepam (Valium), alprazolam (Xanax), lorazepam (Ativan), Triazolam (Halcion), methocarbamol (Robaxin), barbiturates, alcohol, or other], antihistamines [diphenhydramine (Benadryl), Promethazine (Phenergan) or other], stimulants or weight loss medications [cocaine, amphetamine, nicotine, caffeine, methylphenidate (Ritalin), diethylpropion (Tenuate), phenmetrazine (Preludin), phenylpropanolamine (Control), or other], opiates [heroin, morphine, codeine, Percodan, methadone, or other], hallucinogens [phencyclidine (PCP), LSD, mescaline, MDMA (Ecstasy), marijuana, or other], and blank or placebo.

Sensory Measure Questionnaire. At the end of each session, immediately following the completion of the Pharmacological Class Identification Questionnaire, subjects were asked by the research assistant to describe any unusual visual sensations, tastes, or smells experienced during the session. The research assistant wrote, in detail, the subject's response on a sensory assessment questionnaire form.

Physiological Measures. Subjects were monitored continuously (minute by minute) on these physiological measures: heart rate, blood pressure (systolic and diastolic), respiration rate, and skin temperature. Heart rate and blood pressure were measured automatically by a Sentron Automatic Blood Pressure Monitor (Bard Biomedical Division, Lombard, IL). The blood pressure cuff was placed on the nondominant arm. Respiration rate (breaths/min) was measured with a bellows (Pneumo Chest Assembly, Lafayette, IN) placed around the lower chest and connected to a pressure-sensitive switch (Micro Pneumatic Logic, Inc., Fort Lauderdale, FL). Skin temperature was monitored using a skin-surface thermistor (Yellow Springs Instrument Co., Yellow Springs, OH) taped to the index finger of the nondominant hand. Data for each of these measures were collected and stored using the previously described microcomputer.

Data Analyses. Data from the dose run-up sessions were used for safety assessment purposes only and were not used for statistical analyses. Time course data from the experimental sessions for visual analog scales and physiological data were analyzed using univariate two-factor repeated measures analysis of variance (ANOVA). The factors in the analysis were drug condition (placebo; 100, 200, 400 mg/70 kg of caffeine; and 0.75, 1.5, 3.0 mg/70 kg of nicotine) and time (predrug and 2, 4, 6, 8 ... 26, 28, and 30 min postdrug). Data from the ARCI were expressed as change scores (postdrug minus predrug values) and analyzed by ANOVA with drug condition as the within-subject factor. Tukey's postdrug hoc tests were used to conduct pairwise comparisons. Results were considered significant when p  0.05. For repeated-measure ANOVAs, Huynh-Feldt corrected p values are reported.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Visual Analog Scales. Intravenous caffeine and nicotine produced orderly dose- and time-dependent changes on several of the visual analog scales. As shown in Figs. 1 and 2, caffeine produced dose-related increases in ratings of drug effect, good effect, like drug, high, stimulated, and bad effect. Maximum ratings were generally observed 4 min after drug injection. The low dose of caffeine (100 mg/70 kg) did not produce significant increases over placebo on any ratings. However, the intermediate (200 mg/70 kg) and the high (400 mg/70 kg) caffeine doses produced significant increases over placebo ratings, with significant effects usually lasting 6 to 8 min after injection.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Time-action functions for intravenous caffeine and nicotine on subject ratings of "Do you feel a drug effect?", "Does the drug have any good effects?", "Do you like the drug?", and "How high are you?". x-axes: time after drug injection in minutes; 0 indicates predrug. y-axes: subject ratings in millimeters. Data points show mean values for nine subjects. Filled symbols indicate values that are significantly different from the corresponding placebo value at the same time point; p  0.05, Tukey's post hoc test.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Time-action functions for intravenous caffeine and nicotine on subject ratings of "Do you feel a rush?", "Do you feel stimulated?", "How alert/energetic are you?", and "Does the drug have any bad effects?". x-axes: time after drug injection in minutes; 0 indicates predrug. y-axes: subject ratings in millimeters. Data points show mean values for nine subjects. Filled symbols indicate values that are significantly different from the corresponding placebo value at the same time point; p  0.05, Tukey's post hoc test.

Addiction Research Center Inventory (ARCI). On the LSD scale (dysphoria and somatic complaints) of the ARCI, the high dose of nicotine (3.0 mg/70 kg) produced significant increases over placebo and the low doses of caffeine (100 mg/70 kg) and nicotine (0.75 mg/70 kg) (data not shown). These data are consistent with the occasional reports of unpleasant effects such as slurred speech, blurred vision, and reports of burning, tingling, and numbness at the injection site after administration of the high dose of nicotine. No other significant effects were produced by caffeine or nicotine on the ARCI.

Pharmacological Class Identification Questionnaire. Table 2 shows results from the Pharmacological Class Identification Questionnaire. Placebo administration was identified as a placebo on 56% of occasions, whereas the low and intermediate doses of caffeine were identified as placebo on only 11% of occasions. The high dose of caffeine was never identified as a placebo. In comparison, the low dose of nicotine was identified as placebo on 33% of occasions, whereas the intermediate and high nicotine doses were never identified as placebo. Nicotine produced a dose-dependent increase in the frequency of stimulant identifications. When subjects identified a dose of nicotine as a stimulant, they usually (72%) further identified it as being cocaine and they never identified it as being nicotine. The high dose of nicotine was identified as a stimulant by all nine subjects. This dose was further identified as cocaine by seven subjects, as amphetamine by one subject, and as methylphenidate by the remaining subject. Although intravenous caffeine did not produce an increase in the frequency of stimulant identifications, caffeine was almost always identified as an active drug (e.g., stimulant, opiate, sedative, antihistamine). When subjects identified a dose of caffeine as a stimulant, they equally often (29%) further identified it as being cocaine or nicotine and never identified it as being caffeine.

Sensory Measure Questionnaire. The percentage of subjects reporting unusual smells (22, 56, and 78%) and tastes (22, 44, and 67%) increased as a function of caffeine dose (100, 200, and 400 mg/70 kg), respectively. In contrast, reports of unusual smells and/or tastes were not appreciably affected by placebo or nicotine (Fig. 3). The unusual smells reported after caffeine were most often described as having a chemical or medicinal quality (i.e., like furniture polish, ammonia, urine, socks, medicine, chemical, onion). Onset of these experiences generally occurred within 30 s of injection and lasted from several seconds to 10 min. Figure 3 also shows that reports of blurry vision tended to increase in a dose-related fashion with nicotine, but less so with caffeine and placebo.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Effects of intravenous caffeine and nicotine on subject reports of any unusual smells, tastes, or visions. Bars show the percentage of subjects who indicated they experienced unusual smells and/or tastes (top panel) or blurry vision (bottom panel). Pl indicates placebo.

Physiological Measures. Physiological effects of intravenous caffeine and nicotine are shown in Fig. 4. The intermediate dose (200 mg/70 kg) of caffeine produced a decrease in heart rate 20 to 26 min after injection and a decrease in skin temperature 10 to 28 min after injection that were significantly different from placebo. This dose of caffeine also significantly increased diastolic blood pressure over placebo 10 min after injection. The physiological effects of the other caffeine doses were not significantly different from placebo. In contrast to caffeine, all of the nicotine doses (0.75, 1.5, and 3.0 mg/70 kg) produced significant increases in heart rate; the magnitude and duration of these effects were dose-related, with the high nicotine dose producing significant differences from placebo up to 30 min after injection. Figure 4 also shows that, like the intermediate dose of caffeine, the high dose of nicotine (3.0 mg/70 kg) produced a significant decrease in skin temperature (10-20 min after injection) and a transient increase in diastolic blood pressure (4 and 10 min after injection).

View larger version (39K): [in this window] [in a new window]   Fig. 4.   Time-action functions for intravenous caffeine and nicotine on selected physiological measures. x-axes: time after drug injection in 2-min blocks; P indicates predrug injection. y-axes: heart rate (bpm), skin temperature (°C), systolic blood pressure (mm Hg), diastolic blood pressure (mm Hg). Data points show mean value for nine subjects. Filled symbols indicate values that are significantly different from the corresponding placebo value at the same time point; p  0.05, Tukey's post hoc test.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study documents orderly dose- and time-related subjective and physiological effects after intravenous administration of caffeine and nicotine. By directly comparing these compounds when administered via the same route of administration under double-blind procedures, this study provides the first rigorous characterization of their similarities and differences.

Effects of Caffeine. The subjective effects of caffeine characterized in the present study had many similarities with, but also some differences from, a previous study from this laboratory, which examined the effects of intravenous caffeine (37.5-300 mg/70 kg) in a similar subject population (Rush et al., 1995). In both studies, caffeine produced orderly dose-dependent increases in ratings of positive mood effects (i.e., good effect, like drug, high, and stimulated) and reports of unusual smells and/or tastes. One difference between the two studies is that Rush et al. (1995) showed more significant effects and effects of greater magnitude despite the fact that Rush et al. tested a somewhat lower maximal dose of caffeine (300 versus 400 mg/70 kg). Another difference is that the frequency with which subjects identified caffeine, as a stimulant on the Pharmacological Class Identification Questionnaire, was lower in the present study (22% at 400 mg/70 kg) than in the Rush et al. study (95% at 300 mg/70 kg). One major methodological difference between the present study and the Rush et al. study that might account for these differences is that subjects in the Rush study were maintained on a completely caffeine-free diet, whereas subjects in the present study received 300 mg/70 kg caffeine/day (150 mg/70 kg b.i.d.) for the duration of the study. It is possible that caffeine tolerance (i.e., reduced responsiveness to the drug after repeated administration) might play a role in some of these differences (Evans and Griffiths, 1992; cf. Griffiths and Mumford, 1996). An alternative hypothesis to the development of tolerance is a possible behavioral contrast between conditions. The Rush study evaluated only placebo and various doses of caffeine, whereas the present study evaluated caffeine, as well as doses of nicotine that produced pronounced ratings in subjective effects. It is possible that the subjective effects of a drug that produces intermediate ratings when evaluated alone will be decreased when evaluated in the context of other drug conditions that produce much greater effects. Another possibly relevant methodological difference is that cigarette smoking was restricted before sessions for at least 8 h in the present study, but for only 2 h in the Rush et al. study. Because the half-life of nicotine is approximately 2 h (Jacob et al., 1988), it is likely that caffeine was tested in the presence of significant plasma nicotine levels in the Rush et al. study, but not in the present study. This may be important because, as discussed in more detail below, preclinical, human experimental, and human epidemiological studies have demonstrated interactions between caffeine and nicotine.

Effects of Nicotine. In the present study, the profile of positive subjective effects produced by nicotine (i.e., increases in ratings of good effect, like drug, high, rush, stimulated, and alert/energetic), along with some increases in bad effects on the LSD scale, is very similar to that from a number of previous studies (Henningfield et al., 1985; Soria et al., 1996) as well as a study from this laboratory, which compared the effects of intravenous nicotine (same doses) and cocaine in a similar subject population (Jones et al., 1999). Also similar to the Jones et al. study, the high dose of nicotine produced elevations in reports of blurry vision. The most notable difference between the present study and the Jones et al. study was the dose-effect relationship in the Pharmacological Class Identification Questionnaire. In the present study, nicotine produced dose-related increases in stimulant identifications, with all subjects identifying the highest dose as a stimulant. In the Jones et al. study, in contrast, although 80% of subjects identified the intermediate dose of nicotine as a stimulant, only 50% identified the highest dose as a stimulant (with the remaining subjects identifying it as an opiate or sedative). A possibly relevant methodological difference between the present study and the Jones et al. study is that subjects in the Jones et al. study were maintained on a caffeine-free diet, whereas subjects in the present study received 300 mg/70 kg/caffeine/day for the duration of the study. Preclinical studies have shown that caffeine increased the self-injection of nicotine in rats and squirrel monkeys (Shoaib et al., 1996, 1999), and human studies have shown that caffeine increased tobacco smoking under some conditions (Marshall et al., 1980a,b; Rose, 1986; Brown and Benowitz, 1989; cf. Lane and Rose, 1995). Epidemiologically, there is a strong, significant positive relationship between caffeine consumption and tobacco smoking (Hopp, 1994; cf. Swanson et al., 1994). Further research investigating mechanisms of interactions between nicotine and caffeine may be of both basic science and clinical relevance.

Comparison of Caffeine and Nicotine. Inspection of Figs. 1 and 2 and post hoc comparisons show that nicotine generally produced greater subjective effects than caffeine. To the extent that these differences reflect real differences in maximal efficacy, this study suggests that nicotine produces much more prominent mood-altering effects than caffeine. However, it is also possible that relatively higher doses of nicotine than caffeine were studied. Post hoc comparisons between the intermediate dose of nicotine and the high dose of caffeine showed no significant differences on most subjective measures. Thus, for purposes of evaluating possible qualitative similarities and differences between caffeine and nicotine, it is most appropriate to compare the intermediate dose of nicotine (1.5 mg/70 kg) with the high dose of caffeine (400 mg/70 kg). Inspection of Figs. 1 and 2 shows that both of these doses increased ratings of drug effect, good effect, like drug, high, stimulated, and bad effect. The only scale to show differing effects was rating of rush, which was significantly increased by nicotine, but not caffeine. Figures 1 and 2 also show that nicotine had a somewhat faster onset time and time to peak effect than caffeine. Nicotine but not caffeine produced dose-related increases in stimulant identifications, with subjects usually identifying it as being cocaine. As discussed above, this difference in stimulant identification between caffeine and nicotine might be due to the daily administration of oral caffeine in the present study. Finally, caffeine produced dose-related increases in reports of an unusual smell and/or taste, whereas nicotine produced dose-related increases in reports of blurry vision.