Satanand Sharma, Kenneth Ziedman and Herbert Moskowitz
Southern California Research Institute and University of California Los Angeles, Los Angeles, California.

INTRODUCTION

Alcohol is the most widely used and abused drug in the United States. It is estimated that more than two-thirds of American adults consume alcohol, and thus adult drinking is normal behavior in most circles (1). Such widespread use of alcohol produces a considerable impact on society especially in alcohol related disease-causation, and alcohol effects on functioning as a member of the community. Thus the effects of alcohol on behavioral performance warrants study. A direct concomitant of the drinking behavior is the implication of drinking in the performance of many complex tasks such as driving, flying, and operation of machinery. It is estimated that approximately 50% of the drivers involved in fatal accidents are legally drunk or impaired (2).

ALCOHOL EFFECTS ON BEHAVIORAL PERFORMANCE

In humans alcohol is rapidly absorbed and distributed to all parts of the body. It exerts its influence on many sites in the body and appears to have multiple mechanisms of action. Therefore, it is difficult to describe the behavioral actions of alcohol in a specific manner based on a single mechanism, and thus few dependable generalizations regarding alcohol action on performance can be advanced. In this review alcohol effects are categorized according to type of tasks.

Since the alcohol literature is voluminous and a great deal of data exists on alcohol effects on performance in humans, no attempt has been made to cite animal data. In the following, human studies have been selected which are representative of alcohol effects on performance. No attempt has been made to present an exhaustive review.
 

Simple Visual Functions

Moskowitz, Sharma and Schapero (3) studied the effects of 0.69 g alcohol per kg body weight on a variety of visual tests using twelve male subjects. They found that dark adaptation, static acuity, and binocular vision were not affected. In tests of lateral and vertical phoria (position the eyes assume relative to each other and to a fixation target when fusion of the fixation target is rendered impossible), alcohol was found to have no effect on vertical phoria. There was an effect on lateral phoria. The mean prism readings increased by 1.7 diopters. In duction tests, which measure the range that the eyes can turn toward each other, away from each other, and vertically apart from each other in order to maintain single binocular vision of a fixated target, alcohol had no significant effect on supraduction, infraduction and abduction. However, abduction was affected slightly.

Glare recovery, where changes in ambient light levels require the eye to readjust to achieve the previous sensitivity to target contrast at the new level, has been reported to both improve and deteriorate under alcohol (4,5). It is not clear why these contradictory results are obtained but differences in methodology probably are involved. Critical flicker fusion has been reported to show an inconsistent but general trend towards decreased fusion thresholds under alcohol (4). Color preception also changes consistently under alcohol (5). Brightness thresholds have also been reported to be impaired by alcohol (6). Peripheral vision has been reported not to be impaired by alcohol (4,7).

Auditory System

There is considerable evidence that the auditory system is insensitive to alcohol. Schwab and Ey (8) found auditory acuity not to be affected by alcohol. Signal detectability also is not affected by alcohol (9,10).

Psychomotor Functions

Several studies have shown that muscular strength is not affected by alcohol (4) but that muscular steadiness is affected at fairly low levels of alcohol (4,11,12). Both auditory and visual reaction times have been found to be consistently affected by blood alcohol levels of 0.1% (11). The Romberg test of body sway has been found to be very sensitive even at low doses of alcohol (4,11). This test requires a person to stand as steadily as possible and body sway is measured. Similarly, hand steadiness is consistently impaired by low doses of alcohol (4). Hand steadiness is most frequently measured as the subject's ability to hold a metal stylus in a small hole in a metal plate without contacting the edges of the plate.

Division of Attention

Several studies now support the notion that alcohol affects division of attention. Situations which require division of attention are those where input data is received from two or more sources. An example of this is automobile driving where the driver has to concentrate on keeping the car on the road (tracking) and to detect environmental signals such as oncoming traffic.

One study (9) examined attention processes in the auditory modality using 0.69 g alcohol/kg body weight. The experimental tasks included conditions of concentrated attention (subjects attended to input only to one ear) and divided attention (subjects attended to simultaneous input to both ears). Alcohol impaired performance only under the demands for division of attention.

Similar findings have been reported for concentrated and divided attention in the visual modality. Recently, a study by Moskowitz and Sharma (7) examined peripheral vision while the subject was occupied with a central fixation light. There were three central visual conditions: the fixation light was either unblinking, blinked at a slow rate, or blinked at a fast rate. Signal detection was examined at thirty-two points in the horizontal peripheral visual field; at sixteen angles from 12° to 102° on both sides of the fixation light. Alcohol treatments of 0.41 and 0.83 g alcohol/kg body weight were compared with a placebo treatment.

This study specifically tested the hypothesis that the appearance of an alcohol-induced deficit in peripheral vision is a function of the attention or information processing demands placed upon central vision or, for that matter, the demands from any source of information occupying the central processing mechanisms. The condition where the central fixation light was unblinking and thus required no major part of the information processing capacity of the brain duplicates the manner in which earlier studies of peripheral vision under alcohol executed the experiment and failed to find impairment.

The study failed to find any impairment in peripheral vision at either alcohol dose when central vision was occupied with an unblinking fixation light. However, when central vision was occupied with counting the light blinks, there were deficits in peripheral light detections. Under the slow blink central light condition, the two alcohol doses produced 14% and 25% drops in signal detections, and under the fast blink condition 18% and 36% drops in detections. Similar results have been obtained by others (13,14).

Psychological Refractory Period (PRP)

The PRP has been widely used as an experimental paradigm. It involves the presentation of two stimuli to the subject in close succession with each requiring a response. When the interval between the two stimuli is less than about 300 msec the response to the second stimulus is delayed because it arrives while central capacity is occupied with processing the first stimulus. Thus a delay of the second reaction time reflects an increase in central processing time.

Using this technique a study by Moskowitz and Burns (15) found that 0.69 g alcohol/kg body weight slows information processing.

Information Processing

Moskowitz and Murray (16) have examined the rate of information processing under alcohol with the visual backward masking procedure. Four letters were presented tachistoscopically for 15 msec duration. After a dark interval (30, 45, 60, 75 msec) a "mask" of letter fragments was presented. This effectively interferes with further processing of the display stimuli from the sensory register into short term memory. When the subject is required to recall the stimulus letters after mask presentation he can recall only those that he had processed from the sensory store into short term memory prior to the onset of the mask. By varying the interval between stimulus and mask and plotting the number of letters recalled against the interval, a processing rate is calculated. Alcohol doses of 0.414 and 0.828 g/kg body weight both slowed the rate.

Tracking Tasks

Tracking tasks are more typical of perceptual motor performance. Alcohol has been found to affect tracking consistently. Wallgren and Barry (4) found that alcohol produces a decrement in tracking performance and that such a deficit is more likely to appear when the tracking task is performed simultaneously with another task which serves to divide attention.

Sturgis and Mortimer (17) tested subjects on a stylus tracking task and found performance to deteriorate under a blood alcohol level of 0.1%.

Ziedman, Sharma and Moskowitz (18) tested subjects under 0.075% and 0.15% blood alcohol levels with a critical tracking task. This task requires subjects to maintain an illuminated line display on the center of an oscilloscope. This display is made to move in a random manner by a random forcing function. The task becomes more difficult over time so that the subject eventually loses control. This is analagous to balancing a long rod on one's finger. As the rod becomes shorter the task becomes more difficult. The tracking ability of the subjects deteriorated under both doses.

Aksnes (20) examined performance in a link trainer. Subjects were flying blind and were required to monitor seven instruments as well as a map of the course they were required to maintain. The course imposed limits in regard to altitude, airspeed, vertical speed, turning speed, and time. Subjects were administered either 0.2 or 0.5 g alcohol/kg body weight. The larger dose produced about 0.05% blood alcohol level and appeared to cause an impairment although no statistical analysis was reported.

Hughes and Forney (21) tested performance on a pursuit tracking task with four levels of complexity of the function to be pursued. They administered 0.52 g alcohol/kg body weight resulting in about 0.05% blood alcohol level, and reported that all functions showed large degrees of impairment at this dose level.

Another study (13) combined pursuit tracking with signal detection. They included a condition where additional stress was introduced by noise. Under the quiet condition, the lower alcohol dose of 0.21 g alcohol/kg body weight did not affect tracking scores, but the higher dose of 0.63 g alcohol/kg body weight did impair tracking performance. With the additional stress of noise both alcohol doses produced impairment.

Although most studies of pursuit tracking under alcohol have found impairment, there are a few equivocal studies. In an experiment by Gibbs (22) using a pursuit step-tracking apparatus, which involved steps of unequal probabilities, an alcohol treatment resulting in a peak BAL of 0.10% showed impairment on improbable steps but no impairment on probable steps.

Eye Movements

One area of ocular motor control which has been universally reported to show the effect of alcohol is the threshold for induction of nystagmus (4). There is a unique form of nystagmus which appears under alcohol which is known as positional alcohol nystagmus or PAN. In this situation an individual tilts his head, (preferably with closed eyes) and a nystagmus develops if the blood alcohol level is in the region of at least 0.06 to 0.08%. The nystagmus involved does not occur without an adequate stimulus; the head must be tilted. In animals and humans who have had bilateral interference with input from the semi-circular canals there is no PAN under the presence of alcohol.

A recent study (23) showed that alcohol at dose levels of 0.5 and 1.0 ml 95% ethanol/kg body weight produced significant reductions in dynamic visual acuity (DVA). DVA is a complex task requiring coordination of sensory and motor functions in the resolution of detail in moving targets. It has been suggested that the components involved in DVA are static acuity, ocular pursuit of the target by a combination of saccadic and pursuit eye movements (23). It should be noted that static acuity per se is not affected by alcohol (2,4).
Another study (24) studied human eye movement electrooculographically before and after doses of alcohol producing blood alcohol levels of 0.08% and 0.11%. It was found that saccadic and pursuit eye movements where the eyes move rapidly were slowed under both doses of alcohol. The authors reported that the movement of the eyes under alcohol became jerky and inefficient due to replacement of the smooth movement by a succession of irregular saccades. The effect of alcohol on saccadic and pursuit movements raise questions implicating alcohol effects on skills performance which require rapid perception of stationary and moving visual stimuli. One such task is driving.

In a recent experiment (25) the effects of alcohol on visual scanning patterns in a simulated driving situation was determined.

The authors examined eye movements of subjects as they viewed a movie of driving scenes. They used an experimental apparatus with a relatively high data sampling rate (100 per second), a relatively long viewing period (17 + minutes) and a computer data recording and analysis system permitting the rapid extraction of a wide variety of performance variables.

The subjects sat in a driving simulator consisting of the front half of an actual car body facing a 3.66-m wide rear projection screen. Driving films (35 mm) were rear-projected on the screen and subtended a 70 degree horizontal visual angle.

Twenty-one subjects were tested, divided randomly into three groups represented by three alcohol doses, 0.0%, 0.075% and 0.15% blood alcohol levels.

Table 1 shows the allocation of viewing time to dwells, pursuits, saccades, and blinks during the 17 + minutes (1022 seconds) of actual traffic scenes.

For the placebo group, dwells accounted for 64% of total viewing time, pursuits of 19%, and saccades of 14%. In comparison, the alcohol treatment groups showed a trend toward decreased time in dwells and saccades but increase time in pursuits. In contrast to the small changes in total time allocated to dwells, pursuits, and saccades, there were many large changes in the frequencies and mean durations of dwells and pursuits as can be seen in Table 2.

The authors noted that the changes under alcohol can be summarized as (i) an increase in mean time per dwell, (ii) a concomitant decrease in dwell frequency, combined with (iii) an increase in both the frequency of pursuits and mean duration of pursuits. Thus, a person under the influence of alcohol can examine fewer events or examine the same event fewer times. The authors noted that the subjects tend to pursue moving objects more often and for a longer time, further limiting the opportunity for sharing attention between different events. They conclude by suggesting that the major factor underlying the increased accident potential of alcohol use while driving is the impairment of visual search behavior due to a decreased ability to process information as reflected primarily in increased time necessary for information extraction in dwells and pursuits.

Driving Simulation Studies

One very significant aspect of American living is driving.

Most American adults drive. Thus the effects of alcohol on driving behavior is important. Studies of driving behavior under alcohol have avoided testing driving-related skills by placing subjects in typical traffic situations for obvious safety reasons. One alternate method of studying driving behavior is by use of driving simulators.

An example of such a study is one by Moskowitz (26). He placed subjects in a car where they were required to view a large circular screen depicting a traffic movie. The subject's manipulation of the accelerator controlled the speed of the movie projection, and his manipulation of the steering wheel moved the projected image laterally. Braking slowed the speed of the movie. The subject "felt" as if he was in a real driving situation. Twenty-five performance measures of car control and tracking were derived (such as steering movement, brake actuation and speed). None of the parameters showed any impairment under 0.1% blood alcohol level. The study was then replicated with the inclusion of a simple subsidiary task which required the driver to respond appropriately to two colored lights presented at one of two positions on a random basis, with a frequency of one per minute during a thirty-one mile drive. Under the additional requirement for information processing of the subsidiary task, alcohol produced a decrement in signal detectability of the subsidiary task as well as impairment of twelve of the twenty-five car control measures. The author suggested that the alcohol effect was due to alcohol induced impairment of the capacity for information processing, especially as required in time sharing of several concurrent tasks.

These results are in agreement with those of another study (27) where alcohol effects were assessed in an actual flying situation.

In this study by Billings et al. (27), sixteen subjects took off, instrument-flew and landed a plane under four alcohol treatments, resulting in 0, 0.04, 0.08 and 0.12% blood alcohol levels. Eight of the subjects were highly experienced professional pilots, while the other eight were fairly experienced non-professionals. Flights took place with a safety co-pilot plus a physician located behind the pilot in order to incapacitate him, if necessary. Although the tracking demands of flying are more difficult than those of driving, the experienced pilots suffered no significant decrement in their tracking ability even at the highest dosage. However, beginning at the lowest dosages they committed procedural errors which were a hazard to flight. At the highest dose level, the safety co-pilot had to take command of the plane eleven times to prevent an imminent accident. The inexperienced pilots exhibited impairment in their tracking skills and accumulated far more procedural errors including taking off with full flaps, flying without lights, taking off with carburetor heat on, turning the wrong way in response to instructions, and flying a landing approach tuned to the wrong frequency. Catastrophic procedural errors included loss of control in flight, turns toward oncoming traffic and landing errors that would involve striking the ground.

The authors noted the pilots placed tracking or guiding the aircraft as their primary task and relegated all other operations to secondary tasks. They pointed out that as the pilots were progressively affected by alcohol, they became progressively less able to cope with the various facets of their task, and the secondary tasks were the ones that suffered first and the most.

It is clear from the foregoing that alcohol affects a wide variety of performance tasks, from simple reaction time to complex operations such as flying an aircraft. It is noteworthy that alcohol is widely distributed throughout the body, to all tissues and fluids of the body. Alcohol is present in the plasma, in erythrocytes and in the cerebrospinal fluid. It primarily affects the central nervous system. Thus, alcohol is available at most cells of the body and while the exact nature of its interaction with the cells is still unknown, it influences a wide and disparate range of physiological and behavioral mechanisms.

In summary, it is clear that alcohol impairs performance, especially that requiring division of attention. Psychomotor tasks are only slightly affected (28). The ocular motor effects, the slowing of information processing rate, and the deleterious effects of alcohol on other preceptual tasks make it hazardous to perform complex tasks under the influence of alcohol.

REFERENCES

1. First Special Report to the U. S. Congress on Alcohol and Health. U.S. Dept. of Health, Education & Welfare, Maryland (1971).
2. Filkins, L.D., Clark, C.D., Rosenblatt, C.A., et al.: Alcohol Abuse and Traffic Safety: A Study of Fatalities, DWI Offenders, Alcoholics and Court-Related Treatment Approaches. U.S. Dept. of Transp. Contract FH-11-7129, Highway Safety Research Inst., University of Michigan, (1970).
3. Moskowitz, H., Sharma, S. and Shapero, M.: A comparison of the effects of marihuana and alcohol on visual functions, Current Research in Marijuana. Edited by Lewis, M.P., Academic Press, New York, p. 129-150 (1972).
4. Wallgren, H. and Barry, H.: Actions of Alcohol. Vol. I. Biochemical, Physiological and Psychological Aspects. Elsevier, New York (1970).
5. Adams, A.J. and Brown, B.: Alcohol prolongs time course of glare recovery. Nature 257:481-483 (1975).
6. Hicks, J.A. III: An evaluation of the effects of sign brightness on the sign-reading behavior of alcohol-impaired drivers. Human Factors 18:45-52 (1976).
7. Moskowitz, H. and Sharma, S.: Effects of alcohol on peripheral vision as a function of attention. Human Factors 16:174-180 (1974).
8. Schwab, W. and Ey, W.: Experimentelle untersuchungen an hor- and gleichgewichtsapparat unter alkohol. Arch. Obr-, Nas-, uKehlkHeilk. 164:579-628 (1934).
9. Moskowitz, H. and DePry, D.: Differential effect of' alcohol on auditory vigilance and divided-attention tasks. Quart. J. Stud. Alc. 29:54-63 (1968).
10. Schneider, E.W. and Carpenter, J.A.: The influence of ethanol on auditory signal detection. Quart. J. Stud. Alc. 30:357-370 (1969).
11. Franks, H.M., Hensley, V.R., Hensley, W.J. et al.: The relationship between alcohol dosage and performance decrement in humans. J. Stud. Alc. 37:284-297 (1976).
12. King, H.E.: Ethanol induced slowing of human reaction time and speed of voluntary movement. J. Psychol. 90: 203-214 (1975).
13. Hamilton, P. and Copeman, A.: The effect of alcohol and noise on components of a tracking and monitoring task. Br. J. Psych. 61:149-156 (1970).
14. Huntley, M.S., Jr.: Effects of alcohol and fixation-task difficulty on choice reaction time to extrafoveal stimulation. Quart. J. Stud. Alc. 33:1155 (1972).
15. Moskowitz, H. and Burns, M.: The effect of alcohol upon the psychological refractory period. Quart. J. Stud. Alc. 32:782-790 (1971).
16. Moskowtiz, H. and Murray, J.: Alcohol and backward masking of visual information. J. Stud. Alc. 37:40-45 (1976).
17. Sturgis, S.P. and Mortimer, R.G.: Effects of practice and alcohol on selected skills: Implications for an automobile alcohol ignition interlock. Percep. & Motor Skills 37:267-274 (1973).
18. Ziedman, K., Sharma, S. and Moskowitz, H.: Alcohol effects on a critical tracking task. In preparation.
19. Newman, H. and Fletcher, E.: The effect of alcohol on driving. J.A.M.A. 115:1600-1602 (1940).
20. Aksnes, E.G.: Effect of small doses of alcohol upon performance in a link trainer. J. Aviation Med. 25:680688 (1954).
21. Hughes, F.W. and Forney, R.B.: Comparative effect of three antihistamines and ethanol on mental and motor performance. Clin. Pharmacol. Ther. 5:414-421 (1964).
22. Gibbs, C.B.: The effect of minor alcohol stress on decision processes in step-tracking task. IEEE Transactions on human factors in electronics. HFE 7:145-150 (1966).
23. Brown, B., Adams, A.J., Haegerstrom-Portnoy, G. et al.: Effects of alcohol and marihuana on dynamic visual acuity: 1. Threshold measurements. Perception and Psychophysics 18:441-446 (1975).
24. Wilkinson, I.M.S., Kime, R. and Purnell, M.: Alcohol and human eye movement. Brain 97:785-792 (1974).
25. Moskowitz, H., Ziedman, K. and Sharma, S.: Visual search behavior while viewing driving scenes under the influence of alcohol and marihuana. Human Factors 18:417-432 (1976).
26. Moskowitz, H.: The effects of alcohol on performance in a-driving simulator of alcoholics and social drinkers. Reports ENG.-7025, Inst. of Transportation and Traffic Engineering, U.C.L.A., California (1971).
27. Billings, L.E., Wick, R.L., Gerke, R.J. et al.: The effects of alcohol on pilot performance during instrument flight. Report FAA-AM-72-4, Federal Aviation Administration, Washington, D.C. (1972).
28. Levine, J.M., Kramer, G.C. and Levine, E.N.: Effects of alcohol on human performance: An integration of research findings based on an abilities classification. J. Appl. Psychol. 60:285-293 (1975).