Download pdf - The effect of (-) Δ9-tetrahydrocannabinol, alone and in combination with ethanol, on human performance

Psychopharmacology 62, 53-60 (1979) Psychopharmacology �9 by Springer-Verlag 1979

The Effect of (--) Trans-A 9-Tetrahydrocannabinol, Alone and in Combination with Ethanol, on Human Performance

B. E. Belgrave 1, K. D. Bird 1, G. B. Chesher 2, D. M. Jackson 2. , K. E. Lubbe 3, G. A. Starmer 2, and R. K. C. Teo 4

1 Department of Psychology, University of New South Wales, N.S.W. 2033, Australia 2 Department of Pharmacology, University of Sydney, N.S.W. 2006, Australia 3 Psychiatrist, Brisbane Street Drug Dependence Service, Health Commission of N.S.W., Australia 4 Traffic Accident Research Unit, Department of Motor Transport, Rosebery, N.S.W. 2018, Australia

Abstract. Twenty five volunteers received (-) trans-A 9- te t rahydrocannabinol (THC) (320 gg/kg) or placebo (both orally, To), and, 60 rain later, they consumed an ethanolic beverage (0.54 g/k~) or placebo. The effects o f this medicat ion were measured at T1 (100 rain after T H C ingestion), T 2 (160 min), T 3 (220rain) a n d T~ (280 rain) using a bat tery o f cognitive, perceptual and mo to r funct ion tests. Factorial analysis indicated that the test procedures could be adequately expressed by four rota ted factors: a reaction speed factor (I'), a cognitive factor (II ') , a standing steadiness factor ( I I I ' ) and a p sychomoto r coordina t ion factor (IV'). The first principal componen t (I) was used as a measure o f general per formance across the whole test battery.

Both T H C and ethanol produced significant decrements in the general per formance factor. Ethanol produced significant decrements in standing steadiness and psychomoto r coordinat ion, while T H C caused a significant deteriorat ion in per formance on all the four rota ted factors. In all cases the peak effect o f ethanol occurred at T 1 and by T 4 the effect had worn off. The per formance decrements induced by T H C were slower in onset and lasted longer than those induced by ethanol. In general, the peak effect o f T H C occurred at T 1 and T 2. There was no evidence of any interaction between T H C and ethanol, and the effects of a combinat ion o f T H C and ethanol were no more than additive. T H C (but no t ethanol) p roduced a significant rise in pulse rate. Prior administrat ion o f T H C did not significantly affect the b lood ethanol levels obtained. The subjects were able to identify correctly which of the treatments they had received.

Key words: ( - ) Trans-A9-te t rahydrocannabinol - Ethanol - H u m a n p e r f o r m a n c e - Cognitive - Perceptual - M o t o r

* To whom offprint requests should be sent

In two previous experiments, we examined the effects o f orally administered A9-tetrahydrocannabinol (THC) (137 and 214 ~tg/kg), alone and in combina t ion with ethanol (0.54 g/kg), on the performance of a series of perceptual, cognitive and m o t o r function tests in h u m a n volunteers (Chesher et al., 1976, 1977). In the first experiment, neither T H C (137 lag/kg) nor ethanol produced significant per formance decrements when given alone, but deteriorat ion did occur after the administrat ion o f the combinat ion. In the second experiment, T H C (214 ~tg/kg) did produce a significant effect, which was accentuated in the presence o f ethanol. In addition, Chesher et al. (1977) reported some evidence o f an an tagonism between T H C and ethanol which occurred approximately 2 - 3 h after the ingestion o f THC.

This paper reports the findings o f a study in which the dose o f T H C was increased to 320 ~tg/kg and the time of measurement was extended to 5 h.

Materials and Methods

Subjects. The subjects were healthy, paid (mainly University stu- dents) volunteers of both sexes (12 male, 13 female), aged 18- 35 years (median 22.3 years), with body weights of 51 - 81 kg (median 62.5 kg). All were 'social' drinkers of ethanol and were non-naive as regards cannabis. The drug experience of the subjects was as follows: a) ethanol; a median consumption of 1.4 'standard' drinks (285ml beer, 30 ml whisky, or 230 ml table wine) per day (range 1 - 20) for a median duration of 9.6 years (range 1-23); b) cannabis; a median rate of use of two occasions per day, (range of < 1 to > 5) for a median duration of 6 years (range i - 14). The detailed drug-use history of these volunteers and of others in subsequent experiments is being compiled. Before admission to the experiment, all subjects were medically examined by one of us (K. E. L.) to ensure that no past or present disease precluded their participation. Six subjects of an initial 31 were excluded; two because of hypertension, two because they had measurable blood alcohol when they first attended, one because of obvious psychiatric disturbance and one because of recent regular use of bronchodilators. The purpose and design of the experiment was fully explained to the subjects and their informed consent obtained.

0033-3158/79/0062/0053/$ 01.60

54 Psychopharmacology 62 (1979)

Drugs. Both THC and ethanol were administered orally. THC was dissolved in sesame oil and sealed into capsules containing 2.5, 5.0 or 10.0 mg. Each subject was given four capsules and the dosage was adjusted to deliver approximately 320 gg/kg according to the following schedule: subjects weighing 51 - 58 kg received 17.5 mg; 59 - 66 kg, 20.0 rag; 6 7 - 74 kg, 22.5 mg; and 75 - 82 kg, 25.0 mg. The median dose was 320 Ixg/kg (range 302-343 gg/kg). Placebo capsules contained only sesame oil. Ethanol (0.54 g/kg) was presented as a beverage containing 20 ~o v/v ethanol in lemon squash. Ethanol was omitted from the placebo beverage. The dose of ethanol was the same as that used in our previous studies (Chesher et aI., 1976, 1977).

Test Battery

Standing Steadiness. The apparatus consisted of a platform beneath which a displacement transducer was mounted. The actual movement of the platform in the vertical plane was less than 1 gm. The subject stepped on to the platform and was instructed to relax and to stand as still as possible without talking or moving his head. Any shift in position created an electrical impulse that was amplified and recorded on a Grass Polygraph. The impulses were integrated to give an overall measure of body sway, based on frequency and amplitude, under two conditions: eyes open and eyes closed.

Simple and Complex Reaction Time. The timer used was made by Schfihfried, Stuttgart, West Germany. The subject sat with his finger poised over a 1-cm 2 response button and reacted to signals by pressing as quickly as possible. The signals consisted of red and white lights (2.5 cm in diameter, separated by a distance of 8 cm and positioned 6.5 cm from the response button) and a sound (acoustic power 0.1 W, 1250Hz); these were presented in programmed sequence. A timing device measured the interval between the ap- pearance o f the stimulus and the subject's response in milliseconds. For simple visual and auditory reaction time, the subjects were required to respond to a presentation of the white light or the sound. For complex reaction time, they were to respond only when the white light and auditory stimulus occurred simultaneously although the other stimuli, alone and in combination, were also presented. In the experimental sessions, each subject responded to five visual, five auditory and five complex stimuli.

The Vienna Determination Apparatus (VDA). This apparatus (Schfihfried, Stuttgart, West Germany) generates a sequence of visual and auditory stimuli and records button and foot pedal responses. The correct and incorrect responses to the signals were recorded. A correct response was registered when the appropriate response was made during the presentation of the signal. In this experiment, the subjects were required to respond to a series of 100 randomized signals of 1.22-s duration.

Pursuit Rotor. The Motorische Leistungsserie (Schfihfried, Stuttgart, West Germany) apparatus was used and the test was basically one of hand-arm coordination. The task required the subject to track, with a photocell stylus, a 15-mm square lighted area on a horizontal work plate, which rotated at 15 revolutions per minute in a clockwise direction. The number of times the stylus went off the target and the total time it was off target were automatically recorded. The test time was 32 s.

Arithmetic. This is a concentration and attention test. The apparatus used was the Arbeit und Konzentration Testger/ite (Zak, Simbach am Inn, West Germany). The subject was presented with a series of random single digit addition and subtraction displays. The subject was required to key in the answers by pressing one of the ten buttons situated just below the displays. Each response generated another display and the number of correct responses was automatically recorded. The test time was 2 min.

The 'Boggles' Word Construction Test. In this test subjects were presented with a 4 x 4 matrix of 16 letters, and asked to construct as many English words as possible in 3 rain by rearranging letters from unbroken groups of adjacent letters (adjacent by row, column or diagonal). A different form of this test was given on each of the twenty testing occasions. The raw scores on each form were corrected for differences in difficulty level.

Blood Ethanol Concentrations. These were measured by breath analysis using the Alcotmeter (Model AE-D, Lion Laboratories, Cardiff, U.K.). A sample of deep alveolar air is collected when the exhalation pressure is falling and any ethanol present is oxidised by an electro chemical sensor (fuel cell) at 40 ~ The resultant signal is amplified and presented as a peak-reading digital display. The apparatus is calibrated by means of a standard aerosol of ethanol in argon (Nalco, Lion Laboratories, Cardiff, U.K.). In a preliminary experiment, correlation coefficients for the relationship between capillary blood ethanol concentrations determined by gas-liquid chromatography (Franks et al., 1976) and as shown by the Alcolmeter were calculated for four samples of ten pairs of readings. .The correlation coefficients were between 0.93 and 0.97 (P < 0.001).

Procedure

The experiment was conducted on four successive week-ends. Each subject was randomly assigned to one of four groups and each group received each of the following treatments according to a 4 x 4 Latin square: THC + ethanol; THC placebo § ethanol; THC placebo + ethanol placebo; ethanol placebo + THC. The experiment was conducted double blind in that neither the subjects nor the observers were aware of which treatment had been administered until the series was complete. The subjects arrived at the laboratory approximately 2 h after consuming a light breakfast. The test battery was administered to each subject before any drug treatment was given. After the control run (To), the subjects received THC or placebo. One hour later, the subjects were given their ethanolic or placebo beverages and they consumed them under supervision over a 20-min period at a constant rate. Twenty minutes after drinking finished (i.e. 100 min after THC administration), the subjects went through the test battery again (T1) and at hourly intervals thereafter (T2, T3, T4). Pulse rate and blood ethanol were determined at the mid-point of each test sequence and the subjects were asked to estimate the nature and degree of their intoxication. The subjects were allowed to mingle freely, but the tests were conducted in separate cubicles to reduce subjec t -subjec t interaction. A light lunch of sandwiches and de- caffeinated coffee was consumed after the second (i.e. T z, 100 min) post-ethanol trial.

In summary, a three factor within-subjects design was used (Table 1).

Analysis o f Data

In view of the relatively large number of outcome measures employed in this study, it was decided to base the statistical analysis of the data on linear combinations of measures defined by factor analysis (Butler et al., 1972). A principal components analysis carried out across total experimental variation (Horst, 1963) suggested that systematic variance in the test battery could be summarised adequately in terms of scores on four rotated factors, which jointly accounted for 76.8 ~ of total test battery variance. Scores on the first unrotated principal component were also analysed, since it was expected that this particular linear combination would be more reliable than either the rotated factors or the individual measures (Cattell, 1966). Interpretations of the factors was based on the pattern o f f actor loadings given in Table 2. All variables in this table are scaled (after reflection if necessary) in such a way that a high score indicates superior performance. The first

B. E. Belgrave et al.: (-) Trans-A9-Tetrahydrocannabinol 55

principal component (I), which accounted for 36.0 % of total test battery variance, is simply an efficient measure of general level of performance on the test battery as a whole. The following interpretations of rotated factors are based on common features of variables with substantial loadings (italicised in Table 2):

Factor I' - reaction speed factor incorporates visual reaction time, auditory reaction time and complex reaction time data.

Factor II ' - cognitive factor incorporates word construction ('Boggles') and arithmetic.

Factor III' - standing steadiness factor incorporates standing steadiness (eyes open and eyes closed) data.

Factor IV' - psychomotor coordination factor incorporates pursuit rotor and Vienna determination data.

F rom the correlations between the rotated factors (Table 3), it is clear that the factors represent relatively independent dimensions of test battery performance.

Results

Blood Ethanol Levels. There were no significant differences in ethanol levels at any time between the THC

Table 1. Experimental design

Factor Factor levels

T H C (A)

Ethanol (B)

Time of testing (C)

al, 320 gg/kg a2, Placebo

bl, 0.54 g/kg b2, Placebo

c o - First reading before drug administration, To c 1 - First reading after drug administration, T 1 c2 - Second reading after drug administration, T 2 % - Third reading after drug administration, T 3 c4 - Four th reading after drug administration, T 4

placebo-ethanol group and the THC-ethanol group (Table 4).

Performance Data. The mean factor scores obtained in the various experimental conditions are shown graphi- cally in Figs. 1 - 5. Scores on each factor were subjected to a 4 x 2 x 2 x 5 analysis of variance with repeated measures on the last three factors (Winer, 1962), where the first factor is concerned with the order in which the drug treatments were presented. The resulting F ratios for tests of interest (excluding effects related to order) are presented in Table 5. Since all tests were carried out at the 0.01 significance level, the Bonferroni inequality (Harris, 1975) ensures that the overall type I error rate for each effect cannot exceed 0.05. F rom Table 5 and Figs. 1 - 5, the following conclusions can be drawn:

1. The first principal component provides a good summary of the trends across the four rotated factors (Table 5). According to this component, both ethanol

Table3. Correlations between rotated factors

I' II' II ' IV'

Reaction speed I' - 0.149 0.171 0.381 Cognitive performance II' 0.149 - 0.037 0.399 Standing steadiness III' 0.171 0.037 - 0.192 Psychomotor

coordination IV' 0.381 0.399 0.192 -

Table4. Blood ethanol levels (rag per 100mi _+ SEM) of subjects administered ethanol plus T H C or ethanol alone

Alcohol + T H C Alcohol + placebo

T1 69.9 _+ 3.6 74.6 _+ 4.5 T 2 52.8 _+ 3.6 52.9 _+ 2.9 T 3 34.0 • 3.2 35.1 _ 3.0 T 4 15.4 _+ 2.4 15.6 _+ 2.1

Table 2. Factor loadings a

Variables b Factors

I I' II' III' IV'

Visual reaction time (-) 0.684 0.908 0.019 0.038 - 0 . 0 3 6 Auditory reaction time (-) 0.680 0.937 0.017 0.011 - 0 . 0 5 9 Complex reaction time (-) 0.735 0.759 -0 .053 0.028 0.155 Word construction 0.427 - 0.038 0.929 0.008 - 0.056 Arithmetic 0.532 0.043 0.838 - 0.020 0.070 Standing steadiness (eyes open) 0.379 -0 .031 - 0 . 0 2 7 0.897 0.098 Standing steadiness (eyes closed) 0.327 0.045 0.015 0.923 -0 .081 Pursuit rotor (errors) (-) 0.650 0.131 0.037 0.052 0.887 Pursuit rotor (time off target) (-) 0.707 - 0 . 1 2 5 - 0 . 0 9 6 - 0 . 0 3 7 0.839 VDA (no. correct) 0.699 0.i36 0.224 0.039 0.571

Loadings for rotated factors (I' to IV') are factor pattern coefficients produced by a direct oblimin rotation (Harman, 1967) The symbol @) refers to variables on which scores were reflected (multiplied by --1) in order to ensure that a high score indicates superior performance

56

Table 5. F ratios from analyses of variance

Psychopharmacology 62 (1979)

Source of df variation

F ratios from factor scores*

I I' II" III ' IV'

A (THC) 1.21 14.98 2.31 8.74 18.59 9.27 B (Alcohol) 1.21 11.86 4.55 0.04 2.64 9.61 C (Time of testing) 4.84 20.35 6.87 28.36 16.37 11.70 AB 1.21 1.82 0.23 1.82 1.79 3.31 AC 4.84 10.42 5.84 2.73 7.49 2.06 BC 4.84 12.19 2_04 0.32 8.81 7.17 ABC 4.84 0.19 0.55 1.19 0.46 1.42

* Significant F values in italics (F.01 ; l, 21 = 8.02, F.ol ; 4, 84 = 3.57)

i . O U

U m

e-

+1

-1

-2

I General Performance Factor

Time (mi n) . . ~ . z > . ~

H THC + Ethanol - - Ethanol �9 - THC c>-~o Placebo

+0-2

It

o

m -0.4

r

=E

-0.-e

TPReaction Speed , , < 3 "

Time (min) f

Ethanol ~ . / -- = Ethanol

/ _- _- THC o ~ o Placebo

Fig. 1

The effect of THC (320 pg/kg) plus ethanol placebo, THC plus ethanol (0.54 g/kg ), THC placebo plus ethanol or THC placebo plus ethanol placebo on the performance of the general performance factor (I). The data are expressed as the mean factor score and the y axis represents the time after administration of THC (To). Each subject was tested five times: at T O (immediately before THC or THC placebo), and at T t (100rain), T2 (160min), T 3 (220min) and T 4 (280rain). The ethanol or ethanol placebo was administered 60 min after To

Fig. 2 The effect of combining THC and ethanol on the performance of a reaction speed factor (I'). For details see Fig. 1 and text

+0'8

+0.4

Fig.3 The effect of combining THC and ethanol on the performance of a cognitive performance factor (II'). For details see Fig. 1 and text

100

O t.>

0

e-

1T I Cognitive Performance

/•ime (min)

/ 260

B. E. Belgrave et al. : (-) Trans-Ag-Tetrahydrocannabinol 57

3~o

r r T H C -i- Ethanol - - Ethanol �9 - T H C

o - - - o Placebo

-0.4

Fig. 4 The effect of combining THC and ethanol on the performance of a standing steadiness factor (III'). For details see Fig. 1 and text

+1

O u u~

e-

- 2

]1I I Standing Steadiness

T ime (min) q

~- �9 T H C + E t h a n o l

�9 " " Ethanol - - T H C

o - - o P l a c e b o

360

and THC produced highly significant decrements in general performance.

2. Ethanol significantly affected standing steadiness (III') and psychomotor coordination (IV'), although non-significant decrements in performance also occurred on reaction speed (I') (Figs. 1 - 5).

3. THC produced significant decrements, which were generalisable across all performance dimensions. It is also evident that the effect of THC administered orally is very long lasting with decrements in performance being still evident at the T 4 test (i.e. 280 min after THC administration).

58


0 u ul

0

Z v/.

t -

n~

:E

~" Psychomotor Coordination

< Time (rain) ~ ~ ~

~ h a n o l _- _- THC o---o Placebo

Fig.5 The effect of combining THC and ethanol on the performance of a psychomotor coordination factor (IV'). For details see Fig. 1 and text

Table6. Intoxication ratings

Means and SE

T1 T2 Ta T4

THC + ethanol 5.92 + 0.63 6.80 _+ 0.44 6.20 _+ 0.45 3.80 _+ 0.55 THC 4.28 + 0.57 4.88 _+ 0.60 3.76 _+ 0.46 2.00 _+ 0.36 Ethanol 4.40 _+ 0.44 3.40 + 0.41 1.88 _+ 0.38 0.72 + 0.22 Nothing 0.28 _+ 0.15 0.04 _+ 0.04 0.04 _+ 0.04 0.16 _+ 0.13

4. There was no evidence (Table5) of any interaction between THC and ethanol; i.e. the effects of THC and ethanol were no more than additive.

5. It is noteworthy that none of the AB or ABC ratios (Table 5) on the ten individual measures or on the principal component plus the four rotated factors are significant at P = 0.05, despite the fact that the expected number of significant results from 30 tests at this level would be 1.5 by chance alone. Thus there is no suggestion of THC-ethanol interaction in the data.

Subjective Assessment of Intoxication. A three-way analysis of variance of the intoxication ratings showed substantial and highly significant ( P < 0.001) main effects for all factors (THC, ethanol and time), together with small, but significant (P < 0.05) interactions in- volving time (Table6). The pattern of interaction effects shows that ethanol ratings reached a maximum at T1 and substantially declined throughout the course of the experiment, while THC ratings reached a slightly

higher maximum at T 2. Although the THC-ethanol interaction (ignoring time) was not significant, there was evidence of a THC-ethanol-time interaction ( P < 0.05); ratings in the THC-ethanol condition were lower than would be expected from additive effects at T 1, but the reverse was the case at T 4.

Pulse Rates. There were no significant differences between any of the four groups at T o (i.e. before drug administration, Table 7). At T 1, Y2, T 3 and T 4 the pulse rates in the THC-treated subjects were higher than those in the ethanol treatment or double placebo treatment groups.

Physician's Observations. Three adverse reactions were noted. One subject after receiving THC plus ethanol, was pale at T2 and felt nauseated, anorectic and was poor ly coordinated. Intermittent twitches of the arms and legs were also noted. The subject had recovered by T 4. Another subject vomited at T 3 after consuming

B. E. Belgrave et al.:(-)Trans-Ag-Tetrahydrocannabinol 59

Table 7. Pulse rate means and SE

To TI T2 T3 T4

THC + ethanol 82.8 + 2.2 109.4 + 4.7 101.7 _+ 3.8 107.2 _+ 3.0 99.4 +_ 2.9 THC placebo + ethanol 83.7 + 2.2 86.4 + 2.8 82.2 + 2.6 85.7 + 2.6 89.4 + 2.8 THC + ethanol placebo 82.4 +_ 2.1 99.7 + 3.7 93.5 + 3.2 101.8 _+ 3.3 97.6 _+ 2.8 THC placebo + ethanol placebo 86.4 _+ 2.6 83.8 + 2.4 80.4 + 3.1 82.8 +_ 2.6 84.8 + 2.8

THC plus ethanol. His blood pressure at this time was 88 57 and pulse rate 120. He felt nauseated, sleepy,

uninterested and responded only to loud and repeated questions. He considered himself unable to take part in further experimentation on that day. However, at T 4 he had fully recovered. A third subject displayed paranoid ideation 45 min after THC administration and she felt overwhelmed by the room and was frightened of the experimenters and wanted to run out. She complained of being abnormally aware of her body, especially pulse and heart, and intermittent shoulder muscle twitches

110 occurred. Blood pressure was ~ - and pulse rate 1 2 0 -

130 throughout. The subject at this time was amenable to discussion and her paranoid ideation faded. Forty- five minutes later a 25-s episode of thready pulse and intermittent absence of pulse at the wrist occurred. The subject was pale and was twitching at the shoulders during this episode. However, a few minutes later the twitching stopped, the subject felt much better and completed the experiment. This reaction, which we considered to be serious, is similar to that previously described by Melges (1976).

Discuss ion

In our previous studies, it was reported that THC (137 and 214gg/kg, Chesher et al., 1976, 1977) produced what appeared to be a dose-related decrement in performance of a series of cognitive, perceptual and motor function tests. In the present study, we used more subjects and a more powerful method of data analysis. It was confirmed that THC produced a substantial general performance decrement. This decrement was noted in cognition, standing steadiness, psychomotor coordination and reaction speed. The first principal component provides a good summary of the effects of this drug on the general level of performance on the test battery as a whole. Ethanol produced a marked de- terioration in general performance, and specifically in standing steadiress and psychomotor coordination. It is interesting that there was no effect of ethanol (in contrast to THC) on cognitive performance, and only a marginal effect (which was significant at the less

conservative P = 0.05 level) on reaction time. Although it is conceivable that the significant effect of THC on cognitive performance may have been a result of type I error, it is more likely that the effect was real, in which case the poor performance of subjects administered THC might be related to the effects of THC on short- term memory (Darley et al., 1973). The lack of effect of ethanol on cognition was not surprising since it has been reported that cognitive functioning is affected by ethanol to a much lesser extent than is motor performance (Carpenter et al., 1961; Myrsten et al., 1970). The results obtained for the THC--ethanol combination can be accounted for in terms of an additive model. In other words, there was no statistical evidence for synergism and no evidence of antagonism as suggested by Chesher et al. (1977). This conclusion is in general agreement with our earlier report (Chesher et al., 1976).

The absorption of ethanol was confirmed by breath analysis, but owing to the absence of satisfactory analytical techniques, the absorption of THC could not be verified. However, a marked increase in pulse rate in subjects who received THC and subjective reports by the observers suggested that it had been absorbed. Moreover, the subjects themselves estimated the oral dose of THC to be equivalent to smoking 11/2 'joints' of marihuana (data not given), and we also found that the subjects were able to correctly assess which of the four treatments they had been given. Furthermore, the time courses of the subjective intoxication ratings (see Table 6) for the three drug conditions closely followed the corresponding time courses of the first principal components of the performance measures. It is also interesting that the subjects' estimation of their own intoxication at T 4 when given both drugs was greater than that which would be expected from the sum of the individual ratings.

We chose the oral route for THC administration because smoking presents considerable difficulties in the experimental setting for accurate delivery of dosage. Because of this, THC was dissolved in sesame oil and administered orally, a route which has been shown to produce peak levels of THC between 75min and 180 min after administration (Perez-Reyes et al., 1973) with a duration of subjective effect of about 4 h. In the


present study, the first assessment of subjects was made 100 min after THC administration. The peak subjective effects as assessed by the subjects occurred at Tz (i.e. 160 rain after THC administration, Table 1). THC did not affect the blood ethanol levels obtained, although it is possible that ethanol could have influenced blood THC levels. This is unlikely, however, because of the temporal separation of the doses.

From these results it is apparent that a combination of THC (320 ~tg/kg) and ethanol (a dose which produces a peak blood level of less than 0.08 g per 100 ml) can produce performance decrements in tasks related to skills required in driving a motor vehicle. Although there was no evidence of interaction between the drugs on performance, the additive effect de- monstrated is clearly of social significance.

Acknowledgements. We thank the National Institute of Drug Abuse for THC capsules; Mrs. R. Sark and Dr. R. Einstein for help in performing the experiment, and the New South Wales Drug and Alcohol Authority and the Traffic Accident Research Unit for financial help.

References

Butler, D. C., Gocka, E. F., Hartley, J. A., Pinnesu, L. R. : Analysis of factor variance: two cases. Psychol. Rep. 31, 267-279 (1972)

Carpenter, J. A., Moore, O. K., Snyder, C. R., Lisansky, E. J.: Alcohol and higher-order problem solving. Q. J. Stud. Alcohol 22, 183-222 (1961)

Cattell, R. B. : The meaning and strategic use of factor analysis. In: Handbook of multivariate experimental psychology, R. B. Cattell, ed., pp. 174-243. Chicago: Rand McNally 1966

Chesher, G. B., Franks, H. M., Hensley, V. R., Hensley, W. J., Jackson, D. M., Starmer, G. A., Teo, R. K. C. : The interaction of ethanol and Ag-tetrahydrocannabinol in man. Effects on perceptual, cognitive and motor functions. Med. J. Aust. 2, 159- 163 (1976)

Chesher, G. B., Franks, H. M., Jackson, D. M., Starmer, G. A., Teo, R. K. C.: Ethanol and A%tetrahydrocannabinol: interactive effects on human perceptual, cognitive and motor functions. II. Med. J. Aust. 1,478-481 (1977)

Darley, C. F., Tinklenberg, J. R., Hollister, J. E., Atkinson, R. C. : Marihuana and retrieval from short-term memory. Psycho- pharmacologia 29, 231-238 (1973)

Franks, H: M., Hensley, V. R., Hensley, W. T., Starmer. G. A., Teo, R. K. C. : The relationship between alcohol dosage and performance decrement in humans. J. Stud. Alcohol 37, 284-297 (1976)

Harman, H. H.: Modern factor analysis. 2rid edition. Chicago: University of Chicago 1967

Harris, R. J. : A primer of multivariate statistics. New York: Academic Press 1975

Horst, P. : Multivariate models for evaluating change. In: Problems in measuring change, C. W. Harris, ed., Wisconsin: University of Wisconsin 1963

Melges, F. T. : Tracking difficulties and paranoid ideation during hashish and alcohol intoxication. Am. J. Psychiatry 133, 1024- 1028 (1976)

Myrsten, A. L., Kelly, M., Neri, A., Rydberg, U. : Acute effects and after effects of alcohol on psychological and physiological functions. In: The psychological laboratories. The University of Stockholm. No. 314 (1970)

Perez-Reyes, M., Lipton, M. A., Timmons, M. C., Wall, M. E., Brine, D: R., Davis, K. H. : Pharmacology of orally administered A 9- tetrahydrocannabinol. Clin. Pharmacol. Ther. 14, 4 8 - 55 (1973)

Winer, B. J. : Statistical principles in experimental design. New York: McGraw-Hill 1962

Received July 27, 1978