2. Cannabis and Its Effects
PRELIMINARY SUMMARY OF COMMISSION CANNABIS AND ALCOHOL EXPERIMENTS
The Commission has undertaken four experimental projects concerning the acute effects of cannabis and, in some cases, of alcohol on humans. These experiments were designed to fill gaps in the literature in certain critical areas. We were especially interested in the acute effects of cannabis on the functioning of 'normal' users at doses that are socially relevant for Canada today. The purpose of this summary is to supplement the presentation of the findings in the text. The reader is referred to the previous sections of this chapter for discussion, conclusions, general perspective with respect to past literature, current projects, and areas of future attention related to these research areas. Of particular relevance is the previous discussion of dose. This preliminary summary deals with the major aspects of our experimental program. Although the primary results of these experiments have been established. further comprehensive analysis is underway, and a final detailed description of the experiments (with complete statistical procedures"') will appear in a supplement to this report.
One experiment was designed to provide a quantitative comparison of the effects of synthetics, Δ9 THC and natural marijuana in humans, and to establish dose- and time-response relationships with these substances on a number of physiological and psychological measures.' A second experiment sought to determine the effects of cannabis and alcohol on some automobile driving tasks.'5' The purpose of the third study was to examine the effects of cannabis, alcohol, and their combination on psychornotor tracking performance."' The fourth project was concerned with the effects of cannabis on visual signal detection (attention and vigilance) and, secondarily, the recovery of visual acuity after exposure to glare.
"" In all four experiments, subjects were paid volunteers, mostly university students. The subjects were psychiatrically screened, and those with detectable pathology were excluded. Subjects were all experienced with alcohol and cannabis, but not heavy users of either, and had had minimal experience with other psychoactive drugs. No heavy tobacco smokers were included.
A common supply of marijuana, kept frozen under nitrogen. was used for all experiments. Each cigarette was packaged in a nitrogen-filled container and kept frozen until being administered to the subject. Many other conditions of our experiments were standardized and were common to all these proj ects, thereby allowing considerable between-study comparison and a common data base for certain variables.
EXPERIMENT 1: A COMPARISON OF Δ9 THC AND MARIJUANA EFFECTS IN HUMANS"
A variety of chemical and pharmacological data support the contention that A I THC is the principal active constituent in cannabis (at least in part via metabolites). There has been an almost complete reliance on the Δ9 THC content of cannabis to provide the basis for standardizing various samples and comparing the results of different experimental and social studies, even though the pharmacological equivalence of different cannabis preparations accomplished by this process has not been adequately tested. In fact, no direct quantitative comparison of the effects of relatively pure isolated or synthetic THC and crude cannabis preparations or other cannabinoids has previously been conducted in humans. Furthermore, there are increasing suggestions in the literature that a single isomer of THC may not account for all of the major effects of cannabis.
This project was designed to compare the effects of high purity synthetic AI THC with natural marijuana as a further step in determining the active constituents of marijuana in humans. Of comparable interest, acute dose- and time-response functions were obtained on a number of subjective, behavioural and physiological measures.
Fourteen male subjects each attended eight weekly experimental sessions which were six hours in duration. Each subject was assigned a specific time and day of the week for all sessions. After a no-drug practice session, seven experimental conditions were given to all subjects in a double-blind Latin square design as follows: Placebo (extracted alfalfa); three levels of marijuana (9,21, and 88 mcgA' THC per kg body weight, giving average doses of 0.7, 1.5, and 6.2 mg A 9 THC); three equivalent doses of high-purity synthetic A9 THC on extracted alfalfa. Subjects always received a standard 0.4 gram cigarette containing a varying ratio of cannabis to extracted alfalfa material, as appropriate to the condition. The highest marijuana dose was intended to approximate the effects typically sought by regular cannabis users in North America. The lowest quantity was given in an attempt to establish a threshold dose, as suggested by pilot studies and evidence in the literature.
Considerable problems were encountered in obtaining adequate quantitative cannabinoid content figures on the two drug samples. Delta-9 THC values obtained from a variety of authorized laboratories in Canada and the United States differed in the extreme by several hundred per cent. Discrepancies in THC content estimates were finally resolved in an apparently satisfactory manner and the experimental work was begun.IPJ Delta-9 THC made up more than 90% of the cannabinoids in the marijuana sample. Consequently, the study cannot provide information regarding cannabinoids other than THC or possible cannabinoid interactions.
The smoking technique employed was standardized and closely controlled. A five second smoke inhalation was followed by a brief air inhalation period, and the smoke was retained in the lungs for a total of 25 seconds. A 20second rest period followed exhalation. This cycle was repeated until the entire cigarette was consumed including the butt. On the average, this took about ten minutes. On the basis of available evidence, it would appear that with our standard administration technique nearly all of the THC delivered in the smoke was absorbed. Subsequent to the pharmacological studies. 24 cigarettes, comparable to those used in the main experiment, were smoked by machine using the same timing schedule."" This experiment suggested that the actual THC delivered to the subjects in the smoke was 53% and 48%, of' that originally in the marijuana and THC-alfalfa cigarettes respectively. Consequently. in spite of the fact that the cigarettes in the THC and marijuana conditions contained matched quantities of Δ9 THC, there was apparently a 10% difference in the dose delivered. This discrepancy has been taken into'account in the following discussion.
On the basis of a review of the literature and preliminary laboratory work. a condensed test battery was constructed to assess most of the major acute cannabis effects previously indicated and to attempt to tap some of the prominent features of the 'high' as reported by users. The time required to complete the test battery wasjust over two hours, and with the exception of a few measures, the battery was repeated four hours after smoking. The various tests were given in the same order in all conditions. While there are certain practical and statistical advantages to this fixed procedure, it means that the differential sensitivity of the various measures to cannabis effects is confounded with time-response effects. Certain tests, given near the end of the battery, in which little or no change was detected may have shown changes if they had been given immediately after the drug administration. In most cases, however, the major part of the 'high' resulting from the highest dose lasted throughout the first testing period of the test battery.
The following measures were included in the test battery: heart pulse rate; salivation; conjunctival injection (reddening of the eyes); finger temperature; tonic skin conductance of the fingers (a measure of sweat gland activity); visual imagery (visual impressions with eyes closed in total darkness); autokinetic movement (apparent movement of a stationary point of light in an otherwise dark room); two-flash fusion threshold (the shortest time interval between two brief light flashes at which they are still perceived as two),- spiral after-effect (the duration of visual distortion which follows viewing a rotating spiral); time production (the ability to specify 15-second intervals during a distracting task) and time estimation (the perceived duration of the visual imagery, painting and speech tasks); one-minute sustained finger grip; maximum momentary strength of hand grip; maximum tapping speed (the total number of taps made with a hand-held pencil-like stylus in a one-minute period); short-term serial position memory (after a series of nine different digits were presented, a tenth digit was given and the subject was asked to note the position in which that number appeared in the original series); digit symbol substitution (the number of correct code and symbol substitutions written during a 90-second interval); finger painting (scored with respect to a variety of psychiatric, graphic, and aesthetic variables); speech sample (a three-minute description of TAT pictures data not yet fully analysed); Clyde Mood Sea JeI15 (a questionnaire which measures six different mood dimensions)@ Royal Highness Inventory (RHI) (a multiple-choice questionnaire des igned to assess some subjective effects of cannabis); How High Scale (a rating form on which the subject rates his usual "high" and how "high" he currently is on a scale between "not high" and the "highest you have ever been", successive points are recorded by the subject on a single graph at several times throughout the experimental session so that a time-response curve of the intensity of the subjective "high" is plotted); post-Session Questionnaire and Morning-After Questionnaire (a variety of items relating to possible cannabis effects, filled out by the subject at the end of each session and on the morning following each session, respectively).
No consistent qualitative or quantitative differences were found between the marijuana and the pure 6,9 THC on the various measures described above. Tests that showed clear dose- and time-response effects with the natural marijuana, produced very similar results with the synthetic THC. Likewise, measures that showed little effect due to the marijuana revealed minimal response to the THC alone. In the following discussions, "cannabis effects" represent both the marijuana and the pure THC drug conditions.
Cannabis resulted in an obvious increase in pulse rate and conjunctival injection, and a decrease in salivation and finger temperature. For the highest dose, pulse rate was an average of 33% higher than in the placebo condition at five minutes after smoking, 19% higher at one hour, 15% higher at 11/2 hours, 6% higher at four hours, and not clearly different at five hours after smoking. Likewise, salivation decreased markedly after smoking the highest dose, with the effect being less pronounced at 11/2 hours, and not clearly detectable more than four hours after smoking. The highest dose resulted in an increase in conjunctival injection and a decrease in finger temperature within one hour after smoking, but not at 41/2 hours when a second measure was taken. The medium and lowest cannabis doses resulted in effects on these physiological variables that were progressively smaller in magnitude and shorter in duration than occurred with the highest dose.
An increase in visual imagery and autokinetic movement, and a decrease in two-flash fusion threshold resulted from the cannabis, as compared to the placebo condition. Estimates of the time spent in the visual imagery, speech, and painting tasks were longer with cannabis than with no drug; the cannabis resulted in over-estimates of clock time, while under-estimates of clock time were usually observed in the placebo condition. These effects, which were dose-dependent, were observed within a two-hour period following smoking. The drugs did not result in a change in 15-second time interval production (measured 15 minutes after smoking) or in the spiral after-effect (measured 30 minutes after smoking). In the second presentation of the test battery (which did not include speech and painting time estimations), no substantial cannabis effects on any of these variables were observed.
The cannabis resulted in a less forceful finger grip over a one-minute interval than occurred in the placebo condition. A lower level of sustained force was also observed at both 11/2 and 41/2 hours after smoking. No drug effect was found on the momentary maximum strength of hand grip, or the maximum speed of stylus tapping.
An impairment in short-term serial position memory resulted from the highest dose of cannabis. The effect was observed within 15 minutes after smoking, but not at the time of the second measure, four hours after drug administration. Cannabis did not affect performance on the digit symbol substitution or finger painting measures, both of which were obtained rela-tively late in the experimental session (11/2-2 hours after smoking).
The subjective measures were generally found to be sensitive to cannabis, although subjects displayed a rather wide range of response to the doses administered. How High Scale ratings that were made immediately after smoking indicated that the highest dose made the subjects feel, on the average, about as "high" as they typically feel when they use cannabis. The lower doses produced less effect. The second rating, taken about one hour after smoking, was somewhat higher than the initial score for the two upper doses, but lower for the lowest cannabis dose and the placebo. Subsequent ratings declined for all doses during the remainder of the session, with the effects of the highest dose lasting longer than the lower two. The Royal Highness Inventory (RHI) scores, measured at five minutes and at two hours after smoking, also increased as a function of dose. Five hours after smoking, the How High Scale ratings and the RHI scores obtained in the highest dose conditions were slightly (yet still significantly) elevated compared to the placebo condition. The Clyde Mood Scale (CMS) scores, measured 11/2 hours after smoking, indicated that the highest cannabis dose resulted in a decrease in "clearthinking". This finding was reflected in both the subjects' and the experimenters' ratings on such items as "efficient", "alert", "able to concen-trate", and "business-like". None of the other five CMS dimensions reflected drug effects. Five hours after smoking, there was an increase in the CMS factor "sleepy" following the highest dose, but no difference in any of the other dimensions. In the Post-Session Questionnaire, the subjects typically rated the amount of marijuana received as "quite a bit" for the highest dose condition, as "a moderate amount" for the middle dose, as between "a moderate amount" and "very little" for the low dose, and as "very little" in the placebo condition. On the Morning-After Questionnaire, after the pla-cebo condition the subjects usually reported feeling "about average" between the time they left the experiment and the time they went to bed that evening, while after cannabis sessions there were more reports of feeling "very well" during the same period. No other indication of prolonged cannabis effect occurred. No signs of hangover were noted the next morning.
Cannabis threshold. Our lowest dose was intended to approximate a mini-mally effective quantity of cannabis. On the conjunctiva, saliva and Post-Questionnaire measures there was a significant difference between the low dose and the placebo condition. Moreover, all of the other variables that showed substantial effects from both the medium and high doses (pulse rate, How High Scale, RHI and visual imagery) also showed average scores in the low dose condition that were different from the placebo in the same direction. Overall, the data indicate that the low dose did have sorne pharmacological effects even though they were very slight and close to the threshold of measurable subjective and objective response. If we are correct in assuming that approximately 50% of the THC dose in the cigarettes was actually absorbed by the subject, these data suggest the human threshold for THC effects is lower than previously thought (apparently less than 4.5 mcg/kg THC absorbed).
Our data show cannabis dose- and time-response effects on a variety of physiological, perceptual, cognitive, psychomotor, and subjective functions. Cannabis effects on most of these measures were detected early following smoking and, for the most part, were minimal three to five hours later.
The fact that our crude marijuana and synthetic THC had very similar effects on the various measures does not completely rule out the possibility that other psychoactive substances might be present in this or other strains of marijuana. It may be that the relatively high THC content of our marijuana obscured minor effects due to substances that were present only in small amounts. It is also possible that our tests were not sensitive enough to show slight differences in drug effects, or that we did not give the appropriate tests to detect such differences. However, the data do strongly support the hypothe-ses that THC is the major active constituent of marijuana in humans. It also appears that synthetic THC is functionally equivalent to the natural plant product, and that THC content is probably an adequate basis for standard-ization of cannabis preparations. But possible interaction between THC and other cannabinoids (for example, CBD and CBN) in the areas of metabolism, protein-binding, etc., as well as possible direct effects of other cannabinoids in high doses, remains to be explored. Such follow-up work had been anticipated as part of this project, but had to be abandoned due to time limitations. Until further comparative analysis is done with other cannabis samples (for exam-ple, hashish and other marijuana strains), it would be prudent to specify CBD and CBN as well as the THC content of cannabis in experimental situations. In addition, if oral administration is used, the carboxylic acid forms of these cannabinoids should also be specified.
In this experiment, extracted alfalfa was used as placebo. In the other three experiments which follow, hexane-extracted marijuana (of a different strain than the active material) served this control function. In all four studies, cigarettes were smoked under identical conditions using similar subjects, and no evidence of a differential response to these two different placebo materials was obtained. There was, consequently, no suggestion of cannabis-like activ-ity in non-cannabinoid substances remaining in the marijuana after hexane extraction.
It should be stressed that our research was focussed on acute effects and can provide only suggestive evidence regarding the equivalence of THC and marijuana in chronic use. Other substances, which elicit little short-term response, might contribute, over time, to possible effects of long-term use.
EXPERIMENT 2: EFFECTS OF MARIJUANA AND ALCOHOL ON SOME AUTOMOBILE DRIVING TASKS25I
Existing literature indicates that under certain conditions cannabis can produce significant alteration in attention and certain other perceptual and cognitive processes. The results of psychomotor tests and driving simulator studies have been somewhat inconsistent, but tend to show some decrement due to cannabis. Available epidemiological data in North America do not indicate that cannabis has contributed significantly to traffic hazards to date, however. This may, in part, be a function of restricted intensity and fre-quency of use. The following study was intended to explore the possibility of cannabis effects on certain driving tasks. It was not intended, however, to provide a direct basis for generalizing to cannabis-related traffic hazards as a whole. Many other relevant variables remain to be measured.
This project is divided into two separate but related studies. In the main experiment, 16 licensed drivers (4 females and 12 males) each attended four weekly experimental sessions in addition to a preliminary no-drug practice session. The four experimental conditions, given to all subjects in a double-blind Latin square design, were: Placebo (extracted marijuana, and a non-alcoholic drink); two levels of marijuana (21 and 88 mcg THC per kg body weight, producing average doses of 1.4 and 5.9 mg A9 THC); and one dose of ethanol (producing an average blood alcohol level of 0.07%—the equivalent of about three cocktails). Subjects always received a standard drink followed by a 0.4 gram cigarette, each with or without drug, as appropriate to the condition. The cigarettes were smoked following the same schedule described for Experiment 1. In these experiments blood alcohol level (b.a.1.) was measured with a Breathalyzer.
Subjects began the first driving trial, consisting of six course laps (about 6 minutes each), directly after smoking. Each lap involved driving through a 1.1 mile course which included both slow forward and backward manoeu-vring and higher speed (about 25 mph) straight and curved sections, marked out with wooden poles and plastic cones. In one section of the course the subjects were asked to maintain their speed at 25 mph. For the other sections the subjects were instructed to drive as quickly as possible without making too many hits or awkward movements, and without exceeding 30 mph at any time. A second trial (three laps) started three hours after smoking. The subject's driving was scored on: hits of cones and poles, rough handling (superfluous or awkward movements) and speed. Supplementary physiolog-ical and psychological measures were also obtained during the session.
A separate sample of 12 subjects (three females, nine males), experienced with alcohol but not cannabis, were given only the alcohol and placebo drink conditions, and were tested on only one trial consisting of six laps. Course and rating conditions and practice trials were the same as described for the main study.
Both the alcohol dose and the higher dose of cannabis were found to result in poorer car handling performance. During the first trial, significantly more hits (of cones and poles) resulted when subjects received alcohol or the higher dose of cannabis than when they received only the placebos. The mean number of hits per lap was 13.2 in the placebo condition, 16.8 in the higher cannabis condition, and 17.4 in the alcohol condition. There was no differ-ence between the number of hits made in the low cannabis and placebo conditions. In the second trial, given three hours after smoking, the number of hits in the alcohol and higher marijuana conditions had decreased to a level approaching that of the placebo condition. Rough handling tended to be greater after drug treatment than in the placebo condition, although only the alcohol scores appeared significant.
Driving speed was affected by the higher dose of cannabis. In the first trial, subjects drove slower in that condition than when no drug was given. The difference was small (about 1.5 miles per hour or 7%) but consistent. Driving speeds in the alcohol and low cannabis conditions were not significantly different from placebo. In the second trial, differences in driving speed among the four experimental conditions were only slight.
Thirteen of the 16 subjects reported that they were experienced in driving in normal traffic after smoking cannabis or drinking alcohol, while three had never done so. Of the 13 experienced subjects, all but one reported having driven when feeling at least as 'high' as they felt when getting the lower cannabis dose, while seven reported having driven when feeling at least as 'high' as they did after getting the higher cannabis dose. Eleven of the 13 subjects had driven when feeling as 'high' as they felt after getting the alcohol dose. Ratings made before and after the first driving trial indicated that the higher dose of cannabis made most subjects feel as 'high' or a little 'higher' than they typically get from the drug. The alcohol ratings were somewhat more varied, but they suggest a similar situation for that drug. For both the alcohol and the upper cannabis dose the subjects rated their driving ability as lower than they did in the placebo condition. Moreover, they felt that driving took more effort after either drug, and that normally they would be less likely to drive when feeling as they did.
Pulse rate, visual imagery, Clyde Mood Scale ratings and Royal Highness Inventory measures were also taken during each session. The cannabis re-sulted in an increase in pulse rate that was of a similar magnitude and duration as that observed with the same doses in Experiment 1. The alcohol resulted in a smaller but more prolonged increase in pulse rate. Measurements taken immediately after driving indicated that the driving effort resulted in a slight increase in pulse rate, but that the change was similar under the various treatment conditions. An increase in eyes-closed visual imagery was found with the higher cannabis dose, as in Experiment 1. There was a tendency toward decreased imagery in the alcohol condition compared to placebo. The Clyde Mood Scale ratings indicated a decrease in "clearthinking" due to the alcohol and the higher dose of cannabis. Also, alcohol resulted in a decrease in the factor "unhappy" and an increase in the "friendly" score. The can-nabis resulted in Royal Highness Inventory (RHI) scores that were compara-ble to those observed with the same cannabis doses in Experiment 1. The RHI scores also increased in the alcohol condition, but to a lesser extent than occurred with the higher cannabis dose.
The subjects in the second study, who were given only the alcohol and placebo conditions, showed alcohol effects which were similar to those found in the main experiment.
These results show a decremental effect of both the higher cannabis dose and the alcohol dose on car handling performance. The fact that there were no major drug differences in the rough handling score, rated by the observer in the car, suggests that the drug effects on performance, at the doses used here, are not dramatic. It would be premature to predict from these results whether or not cannabis does or will have serious effects on traffic safety. This initial study only measures car handling in rather artificial circumstances. For example, the difficulty and risk involved in traffic was not represented in the test situation, and there was no explicit penalty for making errors, unlike normal driving conditions. In addition, subjects were always aware that they were performing and under observation. These results do serve to point out the possibility that cannabis may adversely affect traffic safety and to under-line the urgent need for extensive research into this question. The reader is referred to the text of the section on Psychomotor Performance and Driving for further discussion and interpretation.
EXPERIMENT 3: EFFECTS OF MARIJUANA AND ALCOHOL ON PSYCHOMOTOR TRACKING PERFORMANCE5I8
The effects of cannabis on psychomotor performance have been shown to be related to dose, frequency of use, and task complexity and familiarity. Also, there are suggestions from some human and animal data that cannabis and alcohol may have additive effects on certain functions, including psycho-motor performance. Some attempts have been made to predict the influence of cannabis on automobile handling and other common tasks by investigating performance on certain laboratory psychomotor measures, but the predictive validity of such tests has not been demonstrated. This study represents an attempt to obtain basic information about the effects of cannabis and alcohol, alone and in combination, on tracking performance. Secondarily, the effects of these drugs on visual perception and several other physiological and psychological variables were also investigated.
Twenty-two male subjects each attended six weekly experimental sessions in addition to two preliminary no-drug practice sessions. The six experimen-tal conditions (given to all subjects in a double-blind Latin square design) were: Placebo (extracted marijuana and a non-alcoholic drink), two levels of marijuana (21 and 88 mcg 6,9 THC/kg, giving average doses of 1.6 and 6.8 mg A9 THC), two levels of alcohol (0.07 and 0.03% blood alcohol level), and the low marijuana and low alcohol doses combined. Again, subjects always received a standard drink followed by a 0.4 gram cigarette, each with or without drug as appropriate to the condition. Subjects followed the same smoking schedule as described in Experiment 1.
Following smoking, subjects began the first trial consisting of six tracking runs (three minutes each). A 20-minute break occurred between the third and fourth runs, during which supplementary measures were taken. In the track-ing task' the subject sat in front of an 8x10-inch screen which displayed a fixed central horizontal target line and a small circle which continuously moved up and down in a random fashion when the tracking control was at rest. Forward and backward movements of a hand-operated "joy stick" changed the rate and direction of the circle movement in proportion to the stick deflection. By such deflections the subject could compensate for the signal fluctuations, and was asked to keep the circle as close to the target line as possible. The distance between the circle and line was the error, and a score was calculated for each three minute run based on the integral of the square of the error divided by the integral of the square of the amplitude of the input signal. The larger the error, the higher the score. In addition, for simple tracking (see below) the subjects' dynamic response was described by simple, linear differential equations, called "describing functions".
In four of the six tracking runs the subject was required only to perform the compensatory tracking (simple tracking). In the other two runs (complex tracking) two additional complications were added to the task: (1) Three times in each run the control dynamics between the "joy stick" and the circle were reversed unexpectedly (that is, it became necessary to push in order to get the circle to move in the direction that had earlier been achieved by pulling, and vice versa). Performance on this task was measured by the speed of reaction and adaptation to the polarity change. (2) Four other times in each run the number 1, 2 or 3 appeared without warning on an electronic tube above the target screen. The subject was required to push a left or right pedal, or continue pushing a middle pedal with his foot depending on which of the three numbers appeared. Performance on this task was measured by the speed of foot choice reaction, as well as the number of times the subject either failed to respond on cue (false negatives) or responded when he should not have (false positives).
A second trial consisting of two simple tracking runs and one complex run was started four hours after drug administration. The secondary physiolog-ical and psychological measures were obtained during periods when the subject was not tracking.
In the first trial, the alcohol and, less consistently, the upper cannabis dose, resulted in an increase in tracking error scores, indicating impairment in both simple and complex tracking. The effect was greater for the higher doses than for the lower doses of each drug. The high alcohol dose resulted in signifi-cantly more tracking error than resulted from cannabis. The combination of low cannabis and low alcohol produced greater error scores in complex tracking than resulted from the corresponding doses of each drug given alone. Interpretation of the subjects' "describing functions" obtained on simple tracking runs indicated that the alcohol resulted in an increase in "effective reaction delay time" in the subjects' continuous tracking perform-ance, and an increase in "random output uncorrelated with input". These effects increased with dose. Examination of the "describing functions" for cannabis revealed a dose-dependent increase in "random output uncorrelated with input", with the effect due to the higher cannabis dose roughly equiv-alent to that of the lower alcohol dose. The cannabis did not result in a change in "effective reaction delay time" or other aspects of the "describing func-tions". The combination of alcohol and marijuana resulted in an increase in "random output uncorrelated with input" that was not substantially greater than that found with the same low doses of each drug separately. "Effective reaction delay time" in the combination condition was not noticeably differ-ent from that observed with the alcohol dose given alone.
Foot choice reaction time was faster in the placebo condition than in any of the drug conditions; however, only the decrement due to the higher alcohol dose was significant. Neither cannabis nor alcohol, nor the combination of the two, resulted in a tendency to miss signals or to respond at the wrong time on this secondary attention and reaction time task. Reaction time to tracking control polarity reversals was reliably longer with the higher dose of alcohol compared to placebo, but no significant effects were seen in the other drug conditions. Speed of adaptation to tracking polarity change has not yet been completely analysed. No major drug effects on psychomotor performance were observed on the later trial which was given four hours after drug administration.
The subjective ratings of intoxication during the period of the first trial indicated that the higher doses of alcohol or cannabis typically made the subjects slightly "higher" than they usually get when using these drugs. The lower cannabis or alcohol doses made them feel somewhat less "high" than they typically get. The combined alcohol-cannabis treatment resulted in alcohol ratings that were similar to those obtained when the low alcohol dose was given alone, and cannabis ratings that were similar to those obtained when only the low cannabis dose was given. While the subjects were able to discriminate which drug or drugs were given in this study and in the driving study, they were somewhat less efficient in discriminating the different can-nabis doses than in Experiment 1 in which no alcohol was given. Clyde Mood Scale self-ratings indicated a decrease in "clearthinking" and "unhappy" for all drug conditions. The alcohol-cannabis combination did not result in scores on these factors that were different from those of either the low alcohol or low cannabis conditions. The higher cannabis dose resulted in an increase in the factor "friendly", while the other drug conditions were not different from placebo in this respect. Also the higher alcohol dose resulted in an increase in the factor "dizzy" while the other conditions did not. No significant drug differences were found in the factors "aggressive" and "sleepy". The Royal Highness Inventory (RHI) scores increased with cannabis and alcohol as described in Experiment 2. There was no evidence of an additive effect on the RHI in the combined alcohol-cannabis condition. Four and one-half hours after the higher cannabis dose was administered, RHI scores were still significantly elevated, while drug differences on other subjective self-report measures had largely disappeared.
Measures of eyes-closed visual imagery, depth perception, visual acuity and time estimation were also collected. Within 11/2 hours after cannabis adminis-tration, visual imagery was increased as in the other experiments. The alcohol conditions resulted in a slight decrease in imagery, as was the case in the driving study. The combined alcohol-cannabis dose resulted in a degree of imagery that was intermediate to the low alcohol and low cannabis doses.
There was a suggestion that accuracy of depth perception decreased slightly due to the higher cannabis dose, but not with the other drug conditions. Visual acuity measures were not affected by the drugs. Time estimations of the intervals spent tracking were longer for the cannabis conditions compared to the placebo condition, as was the case in Experiment 1. However, the time spent tracking was under-estimated in the placebo condition, and in this study the time estimates in the cannabis conditions were more accurate. Three to 41/2 hours after drug administration, there was a negligible drug effect on these measures.
Physiological measures included pulse rate, conjunctival injection, blood pressure, and palmar tonic skin conductance. Pulse rate and conjunctival injection increased with cannabis, as in Experiment 1. Increases in pulse rate and conjunctival injection also resulted from alcohol, and the combined alcohol-cannabis condition resulted in a greater increase in these measures than was observed with each dose separately. Within 15 minutes after smok-ing the higher cannabis dose, diastolic blood pressure (while sitting) was an average of 14% higher than in the placebo condition. The difference did not appear when measurements were made 45 minutes after drug administration. There were no reliable changes in this measure under the other drug condi-tions. No significant drug differences were found in systolic blood pressure, although there was a tendency for an increase in the higher cannabis condi-tion. No copsistent effects on tonic skin conductance were found in any of the drug conditions.
Blood alcohol level was measured (with a Breathalyzer) at three times over the course of each session. Estimates obtained after the administration of the low alcohol dose alone were not different from the corresponding measures in the low alcohol-cannabis combination condition. This suggests that at the low doses studied, cannabis does not have a significant effect on the rate of alcohol appearance in and disappearance from the blood.
These results indicate that alcohol, cannabis, and the combination of these drugs can result in decreased psychomotor tracking performance. The clearer and more pronounced performance decrement in complex tracking that resulted from the combination of alcohol and cannabis (compared to the same low doses of each drug separately) suggests that the effects of the drugs combine on this measure. Effects of the two drugs also appear to be additive on some of the supplementary variables (for example, pulse rate and conjunc-tival injection), but to not interact on others (for example, the RHI and the Clyde Mood Scale), and were possibly antagonistic on visual imagery. None of the drug conditions caused subjects to miss signals or respond at the wrong time in the choice reaction test, suggesting no drug effects on the level of attention required to perform this simple task. The high alcohol dose did result in an increase in choice reaction time, in reaction time to tracking control polarity reversals, and in "effective reaction delay time" during continuous tracking. No similar consistent changes in reaction time measures were found with the doses of cannabis employed.
The mechanism for the interaction of alcohol and cannabis effects is not clear. It would appear that cannabis can enhance certain alcohol effects in the absence of discernible alteration of alcohol absorption, metabolism or excre-tion. We were not, of course, able to assess the possible effects of alcohol on blood levels of THC and its metabolites. The differential pattern of effects of the alcohol, cannabis, and combination conditions suggests that these drugs do not interact solely by one simply enhancing the general effects of the other. Other doses and conditions of administration must be explored to elucidate the mechanism manifested here.
EXPERIMENT 4: EFFECTS OF MARIJUANA ON VISUAL SIGNAL DETECTION AND GLARE RECOVERY6°3
Numerous studies indicate that a change in directed attention is a frequent consequence of acute marijuana use. It has been suggested that an alteration in this basic function may underly many of the other major cannabis effects noted in the psychopharmacological literature. However, few studies have focussed directly on cannabis influences on attention and vigilance. In the present study, these functions were assessed using a visual signal detection task."'
Secondarily, the experiment also explored cannabis effects on the recovery of dim-light visual acuity after bright glare. There is little evidence that a significant effect occurs with marijuana, but because of the potential impor-tance of this question to night automobile driving, further investigation was deemed worthwhile.
Five male subjects were trained for six practice runs (over three days) and tested in two placebo (extracted marijuana) and two marijuana sessions each. Two identical sessions, under each of the two conditions, were necessary to obtain sufficient data for the signal detection analysis. The four experimental sessions were given in a balanced alternating order (single blind) with one week between drug treatments. A single dose of marijuana was used (66 mcg
THC/kg, averaging 4.7 mg 6,9 THC per subject). Cigarettes (.4 gm) were administered with the same standard smoking technique used in the other studies. Immediately after smoking, subjects were given a test battery in which the major focus was on the assessment of sustained attention and vigilance (signal detection).
The signal was a brief offset (gap) in a point of light that was otherwise continuously on. The tests were run in a dimly lit room. Five blocks of 100 trials each, separated by 30-60 seconds, were given over a 40-minute period. Subjects were asked to indicate on each trial the presence or absence of a temporal gap in the light, and to rate their confidence in their decision. A warning tone was presented before each trial. A "signal" occurred in only 50% of the trials, in a random sequence.
After the signal detection task (approximately three-quarters of an hour after drug administration) dim-light visual acuity was measured. A forced-choice procedure was employed using Landolt C test figures. Subjects were then exposed to a large bright-glare light (440ft-L) for five seconds, and the time required to recover to the previous visual acuity level was recorded. This procedure was then immediately repeated. Other supplementary measures included pulse rate and palmar skin conductance, the Royal Highness Inven-tory, a subjective "highness" rating, and eyes-closed visual imagery.
Signal detection. The subjects were generally less accurate in identifying the signal after marijuana and there was a reliable decrement in d' (a measure of sensitivity to the signal) in all subjects for the drug condition. Furthermore, subjects were less confident in identifying a true signal after the drug. Subjects did not fail to respond in any more trials when under the influence of marijuana than in the placebo condition, and there were no systematic changes in the response criteria employed. There were no drug-related shifts in performance over blocks of trials within sessions. The significant decrease in d' was interpreted as an attentional decrement. The performance decre-ments correlated significantly with the individuals' subjective ratings of the magnitude of the drug-induced "high".
Visual acuity and glare recovery. There was a suggestion of reduced dim-light acuity under the drug condition in some individuals, but this trend was not consistent over sessions or subjects. No significant marijuana effect on glare recovery time was found. The possible confounding effect of a decrease in attention during this task cannot be separated out from the primary variables of interest.
Other measures. As in the previously described experiments, marijuana significantly increased scores on visual imagery, Royal Highness Inventory, pulse rate and the subjective "highness" rating. Again, no consistent effects were seen on tonic skin conductance.
The consistent decrement in performance on the visual signal detection task can be interpreted as strong evidence for a marijuana-related decrease in directed attention when a boring and tedious task is involved. It is interesting that this effect correlated well, across subjects, with the subjective sense of drug "highness". No drug-related shift in performance occurred over time within the 40-minute session. Other testing conditions (such as lower signal probability and no warning signal) might give different results. The effects of cannabis on attention and vigilance should be further explored in a broader stimulus and reward context, with varying drug doses, and a greater number of subjects.
No consistent and significant changes were found in glare recovery time and dim-light acuity in this experiment. It is unlikely that a major effect (under the dose and time conditions studied) with practical relevance (to night automobile driving, for example) would have escaped detection, al-though more subtle effects may have been missed in this secondary aspect of the study.