Effect of sucrose consumption on alcohol-induced impairment in male social drinkers. Camillo Zacchia 1, Robert O. Pihl 1'2, Simon N. Young 2, and Frank R.
Psychopharmacology(1991) 105:49-56 003331589100 t68N
Psychopharmacology © Springer-Verlag 199t
Effect of sucrose consumption on alcohol-induced impairment in male social drinkers Camillo Zacchia 1, Robert O. Pihl 1'2, Simon N. Young 2, and Frank R. Ervin 2 1 Department of Psychology,McGill University, 1205 Docteur PenfieldAvenue, Montr6al, Qu6bec, Canada, H3A 1B1 2 Department of Psychiatry,McGill University, 1033 Pine Avenue West, Montr6al, Qubbec, Canada, H3A 1A1 Received April 25, 1990 / Final version January' 1t, 1991
Abstract. Two studies were conducted to examine the interaction between sucrose and ethanol in normal young fasting adult males. The first experiment employed a 3 (100 g sugar, 35 g sugar, 0 g sugar) x 3 (alcohol, placebo and sober) factorial design, which was carried out double-blind using aspartame to ensure that all the drinks were equally sweet. Subjects were tested for mood, memory, subjective intoxication and psychomotor performance at baseline and at times up to 3.5 h after ingestion of the drinks. An alcohol by sugar interaction was seen at 0.5 after drinking. Sugar attenuated alcohol intoxication at this time without influencing blood alcohol levels. Contrary to previous reports, the combination of alcohol and sugar failed to produce significant hypoglycemia, or any of the adverse behavioral effects associated with hypoglycemia, at later times after drink ingestion. The second experiment involved a simpler design, carried out single-blind in which the subjects receiving no sugar did not get aspartame. This was to rule out the possibility that aspartame was exacerbating alcohol intoxication instead of sugar attenuating it. The second experiment also showed that sugar can attenuate alcohol intoxication in fasting humans without altering blood alcohol levels significantly. Key words: Ethanol - Sucrose - Intoxication
Acute alcohol intoxication causes a variety of effects including increased aggression (Zeichner and Pihl 1979), behavioural impairment (Vuchinich and Sobell 1978) and an increased probability of being involved in a traffic accident (Hyman 1968). However, one consistent finding concerning the behavioral effects of alcohol is the very high level of inter-individual variability. Expectancy (Marlatt and Rohsenow 1980), personality variables (Smart and Schmidt 1970; Signori and Bowman 1974), drinking history (Moskowitz et al. 1974; Lipscomb and Offprint requests to: S.N. Young
Nathan 1980; Lipscomb et al. 1980), self-directed attention (Ross and Pihl 1988) and other personality factors have all been examined as possibly contributing to this variability. Recently investigators have begun examining the effects of the individual's nutritional state on degree of intoxication produced by ethanol (e.g., Zacchia et al. 1987). Several studies have looked at the possible interactions between alcohol and various carbohydrates, including sugar. Sugar is consumed in abundance in our society and alcohol is often consumed with mixers that contain large quantities of sugars such as sucrose, fructose or glucose. Of interest in the present studies is how these two substances can interact to affect performance after intoxication. Alcohol and sugar are both carbohydrates and both can produce an initial rise in blood glucose followed, in some cases, by rebound hypoglycemia after several hours (Wallgren and Barry 1970; Danowski 1978; Marks and Rose 1981). When alcohol and sugar are combined, these effects on glycemia are exaggerated (O'Keefe and Marks 1977). Thus, when examining the effects of sugar and alcohol combinations, tests should be conducted both at high and low blood glucose levels. The behavioral effects of sugar alone tend to be relatively minor (reviewed in Milich et al. 1986 and in Spring 1986). While most experimental studies have examined effects in children, studies with adults have yielded inconsistent results. There is some suggestion that sugar can enhance endurance in men (Ivy et al. 1983), but no effects have been found on psychomotor tasks (Ivy et al. 1983) or on mood (Brody and Wolitzky 1983). Behavioral impaitanent was reported in some subjects at peak blood glucose levels in one study (N/irvfinen 1983) and at hypoglycemic levels in another (Bruce et al. 1974). A small lowering of mental alertness has also been reported after a carbohydrate lunch relative to a protein lunch (Spring et al. 1983). When alcohol and sugar are combined, researchers often find that blood alcohol peaks are lowered, and that this results in lessened behavioral impairment. This has been shown in both rats (Mueller et al. 1971 ; Jones et al.
50 1979) and in h u m a n s (Franks et al. 1977; G o l d b e r g et al. 1979a, b; Perl and Starmer 1983). These findings have been seen with fructose or with syrups that c o m b i n e fructose with other sugars. While some a u t h o r s attribute the attenuation in i m p a i r m e n t to differences in alcohol a b s o r p t i o n (Franks et al. 1977), reviewers have concluded that fructose, but n o t sucrose, has the ability to enhance ethanol elimination (Mezey 1981), despite giving the a p p e a r a n c e o f affecting a b s o r p t i o n ( C r o w n o v e r et al. 1986). A n u m b e r o f studies have reported on the effects o f sugars on the response to alcohol. F o r example, M c L e o d et al. (1984) f o u n d that intravenous injections o f glucose did n o t prevent ethanol-induced a n t a g o n i s m o f fetal breathing movements. Glucose given simultaneously with alcohol was f o u n d to have no effect on rat perform a n c e on a tilting plane (Sammalisto 1962). In addition, fructose did n o t reduce the severity o f h a n g o v e r in h u m a n s on the m o r n i n g after intake o f a 1.5 g/kg dose o f ethanol (Sepp/il/i et al. 1976), while b o t h fructose and glucose were f o u n d to have no effect o n a subjective measure o f i m p a i r m e n t when given simultaneously with ethanol over a three h o u r period in the evening (Ylikahri et al. 1976). Studies examining the effects o f b l o o d glucose values have f o u n d that conditions o f high b l o o d sugar increase the lethal dose o f alcohol in rats ( H a g g a r d and Greenberg 1937), a n d in mice (Heistand et al. 1953). However, the m e c h a n i s m for this effect m a y be different f r o m that modulating intoxication. A n experimental drug, R o 15-4513, which reverses alcohol-induced impairment, has been f o u n d to have n o effect o n lethal alcohol levels (Suzdak et al. 1986). L o w b l o o d sugar m a y have different effects. A l c o h o l and sugar can c o m b i n e to p r o d u c e a greater hypoglycemic r e b o u n d than either substance alone and this can result in greater behavioral impairment. F o r example, O ' K e e f e a n d M a r k s (1977) f o u n d t h a t subjects w h o consumed gin and tonic s h o w e d greater hypoglycemic reb o u n d and greater n e u r o g l y c o p a e n i a several hours after drinking than subjects w h o c o n s u m e d either tonic alone or gin a n d sugar-free tonic. Thus, when examining the effects o f sugar on alcohol-induced impairment, results m a y depend on whether measures are taken at a time o f peak or nadir b l o o d glucose levels. The present experiment was designed to examine the effects o f alcohol with or w i t h o u t sugar o n behavioral i m p a i r m e n t at a series o f points after ingestion, including at p e a k b l o o d glucose and b l o o d alcohol levels, at a point w h e n b l o o d glucose had returned to baseline levels and at a p o i n t when hypoglycemic r e b o u n d w o u l d be at its m a x i m u m . Pilot testing revealed that 100 g sucrose dissolved directly in alcoholic beverages h a d no effect on b l o o d alcohol levels. It was therefore felt that this mixture could provide a fair test o f alcohol/sugar interactions. It was hypothesized that sugar effects alone w o u l d be minimal with the possible exception o f increased imp a i r m e n t in some subjects. W h e n mixed with alcohol, sugar was expected possibly to enhance p e r f o r m a n c e on strength measures and to have no or only marginal effects on p e r f o r m a n c e measures during testing at p e a k b l o o d
glucose levels, since B A L s were expected to be equivalent. A t the point o f m a x i m u m hypoglycemic rebound, it was hypothesized that alcohol and sugar would p r o d u c e greater behavioral i m p a i r m e n t t h a n either substance alone.
Materials and methods Experiment 1 was a 3 x 3 x 4 split plot design examining alcohol, placebo and sober conditions by high, medium and no sugar conditions with repeated measures on the time factor. One hundred and thirty three male subjects between the ages of t8 and 35 (median= 22) were recruited through newspaper advertisements. They had no history of criminal behavior, psychiatric illness, alcoholism, or diabetes and normally consumed between 2 and 25 alcoholic drinks per week (median = 6). Participants were randomly assigned to the various treatment conditions. There was no significant difference between the groups in age or the number of drinks they consumed per week. Subjects in the alcohol condition received 3 ml/kg of 80 proof vodka mixed with 6 ml/kg club soda, 5 ml pure lime juice and 15 mg quinine to give the taste of a vodka and tonic mix. Subjects in the placebo group received the same mix, with the exception that water was used in the place of vodka. Olfactory and gustatory cues were provided to enhance placebo believability by spraying the terrycloth holder of each glass with vodka, and layering 4 ml of vodka on each drink. Subjects were told they would reach a legal level of intoxication. Sober subjects received the same drinks as placebo subjects, with the exception of the olfactory and gustatory cues, and were told they would not be consuming any alcohol. In addition, the drinks contained 100 g, 35 g or 0 g of sucrose. In order to ensure equivalent taste, 228 mg, and 350 mg of aspartame were added to the 35 g and 0 g sugar drinks, respectively. Fifteen subjects participated in each of the nine conditions with the exception of the sober- 100 g sugar group and the placebo-35 g sugar group, in which 14 subjects participated. The following measures were taken from each subject: pursuit rotor performance, in which subjects were required to keep a stylus in contact with a 2 cm disc located near the periphery of a turntable rotating at 45 rpm while simultaneously being distracted by a reaction time task; reaction time, in which subjects were asked to press one of eight buttons below eight corresponding lights as soon as possible after a light came on. This was done with the non-dominant hand both with and without the simultaneous measurement of pursuit rotor performance. This yielded a measure of simple reaction time and divided attention reaction time; anterior-posterior body sway in which subjects stood with feet together, eyes closed and head tilted back for 1 min; free recall memory, in which the subjects were read a list of 40 common words (Thorndike and Lorge 1944) at the rate of one per second and were then immediately asked to write as many as they could remember in 90 s; subjective intoxication was measured by the Sensation Scale (Maisto et al. 1980) and a Likert-type "How Drunk" self-rating; mood was measured by the "Right Now" version of the Profile of Mood States (McNair et at. 1971), the Multiple Affect Adjective Check List (Zuckerman and Lubin 1965) and the State-Trait Anxiety Inventory (STAIState) (Spielberger et al, 1970); sleepiness sensations were measured by the Stanford Sleepiness Scale (Hoddes et al. 1973); grip strength and deterioration was measured by the use of a dynamometer for five consecutive trials. Average performance was used as a measure of grip strength while change from first to fifth trial was used as a measure of strength deterioration; hand steadiness was calculated by counting the number of contacts each subject made with a metal probe on the inner edges of holes ranging from 25 mm to 3 mm in decrements of 2 mm; blood glucose was measured by means of an Ames Glucometer; blood alcohol levels (BAL) were read on an Intoximeter Inc. Alcosensor III breathalyser. Subjects arrived in the laboratory at 9:00 a.m., having fasted since midnight the previous night and having abstained from alcohol for a period of 24 h. Subjects were tested in groups of three to
51 seven with all subjects on a given day participating in the same experimental condition. The experimenter administering the tests remained blind to the drinks received. During the first hour subjects were given 8 rain of practice on the pursuit rotor reaction time apparatus and were asked to complete a drinking history questionnaire. At 10:00 a.m. baseline testing was conducted on all measures except the "How Drunk" scale and the breathalyzer. Drinks were administered in three equal portions at 11 : 00 a.m. and subjects were required to consume the three drinks in 20 rain. All tests were repeated at 0.5 h, 1.5 h, and 3.5 h following the completion of the drinking period with the addition of the breathalyzer reading and the completion of the "How Drunk" rating of subjective intoxication. The tests were designed to be completed on all subjects in approximately 20-30 min. In between test periods subjects were shown a movie on video cassette (The Right Stuff). Experiment 2 was a simple two group comparison with subjects receiving alcohol and 100 g sugar (n = 17) being compared to subjects receiving alcohol alone without any sweetener (n= 16). This resulted in a single blind procedure since subjects were aware of their experimental condition. The remaining procedures in experiment 2 were identical to those in experiment 1 with the following exceptions: (i) the memory lists were recorded on audio tape to provide a more standardized presentation of stimuli, (ii) the hand steadiness and mood measures were not taken, (iii) the 3.5 h post alcohol testing period was not conducted and (iv) on each day subjects in both experimental conditions were tested.
Results
Experiment 1
jects revealed no effect o f sugar across test phases, suggesting that in experiment 1 the sugar m a n i p u l a t i o n did not affect alcohol a b s o r p t i o n or elimination. M e a n b l o o d glucose readings are presented in Fig. 2. R e p e a t e d measures analysis o f variance revealed a significant sugar b y time interaction ( F = 7.62, df= 6, 240; P < 0.001) and a significant alcohol by time interaction ( F = 2.18, df= 6, 240; P < 0 . 0 5 ) . T u k e y ' s m e t h o d revealed t h a t at 0.5 h subjects in the 0 g sugar g r o u p were f o u n d to have significantly lower glucose readings than the 35 g and the 100 g groups ( P < 0.001). A t 1.5 h the 35 g g r o u p was f o u n d to have lower scores t h a n the other sugar conditions (P < 0.00t). A n examination o f individual glucose levels revealed t h a t 4 o f the 133 subjects h a d readings lower than 50 mg/dl after 1.5 h, and 5 subjects h a d such scores after 3.5 h. These occurrences were n o t m o r e frequent in the high sugar conditions. M e a n p e r f o r m a n c e on the f o u r p s y c h o m o t o r measures (pursuit rotor, b o d y sway, reaction time and divided attention reaction time) are presented in Fig. 3. Repeated measures M A N O V A revealed an alcohol by sugar by time interaction ( F = 1 . 7 3 , d f = 4 8 , 1400; P < 0.005). U s i n g change scores at the point o f m a x i m u m change (baseline to 0.5 h) a M A N O V A revealed an alcohol by sugar interaction ( F = 1.83, df = 16, 367; P < 0.05). Univariate analysis revealed significant interaction effects on pursuit r o t o r p e r f o r m a n c e ( F = 2.48, df= 4, 123; P < 0 . 0 5 ) , divided attention reaction time ( F = 2 . 8 6 , df=4, 123; P < 0 . 0 5 ) and b o d y sway ( F = 3 . 3 1 , df=4,
M e a n b l o o d alcohol levels are s h o w n in Fig. 1. A repeated measures analysis o f variance a m o n g alcohol sub-
100
/'~ 90
i "...
0.10
ALCOHOL
\\....\ 80
70
~D
== o =o o
o m
0.05
I
I
o eo
I
"% ~
I
NO ALCOHOL
/~....
"o
o o m
I
100
90
o
"o
80
o
o
m
•
/
"'.
'
0 0
1 Time (h)
2
3
4
after alcohol ingestion
Fig. 1. Blood alcohol values as measured by breathalyzer. Values are the mean of 15 for subjects who received no sugar (....... ), 35 g sugar ( - - ) or 100 g sugar ( ...... )
-...
I
I
I
I
I
0
1
2
3
4
T i m e (h ') a f t e r a l c o h o l
ingestion
Fig. 2. Blood glucose values. Values for subjects who received alcohol are the mean of 15. The values for subjects who did not receive alcohol include subjects from both placebo and sober groups and are the mean of 29 or 30 subjects. -.... No sugar; - - Iow sugar; ...... high sugar
52 DIVIDED ATTENTION REACTION TIME 1.5
1.4 "0 cO
~.'-~
MEMORY
REACTION TIME 1.5
1.3
t,*~ ~,
8
ft.5
6
=~
5
"---. ...,
~
. . . ............................ ",
////
CO ,." !""
1.4
,...,~
O
1.2 f
I
I
I
i ".l
i
PURSUIT ROTOR
",,//
4 I
110
........
50
I
,*, BODY SWAY
I
~ ~oo
I
I
SUBJECTIVE INTOXICATION
5
°"-.
x
..t"
45
.......... ,
~ .~
90
~ "~
80
.--
, i\
...... ....
.."
g
"o~
4
40 g
35
~./.
70
~'~ ~e a
"~co 60
g-
O 0~ CO
0
1
f
I
i
I
2
3
4
0
1
i
i
2
3
4
Time (h) after alcohol ingestion
2 I
Fig. 3. Effect of alcohol and sugar on performance. Because there was no effect of sugar on any performance measure in the no alcohol group, values for no sugar, 35 g sugar and 100 g sugar for both placebo and sober subjects are combined to give a mean of 88 subjects. Values for all three alcohol groups are the mean of 15. Measurements for simple reaction time, divided attention reaction time and pursuit rotor performance were obtained as follows: the pursuit rotor was on for four 20-s periods, with 20-s intervals in between. The reaction time task was carried out for the entire 160 s, with ten tests while the pursuit rotor was off, giving the simple reaction time, and ten while the rotor was on, giving the divided attention reaction time. Body sway is given as total units of measurement in 60 s, where I unit= 6 mm. Symbols: No alcohol .-.-.-; alcohol no sugar .... ; alcohol low sugar - - ; alcohol high sugar ......
123; P < 0 . 0 5 ) . The reaction time univariate was not significant (F = 1.11, d f = 4 , 123; NS). Tukey's tests revealed that subjects in the alcohol-100 g sugar condition showed less deterioration after drinking than alcohol-0 g sugar subjects on the pursuit rotor ( P < 0 . 0 5 ) and body sway ( P < 0.05) measures, and less than the alcohol 35 g subjects on pursuit rotor performance ( P < 0 . 0 5 ) and divided attention reaction time (P
