0145-6008/02/2606-0827$03.00/0
ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH
Vol. 26, No. 6
June 2002
Biphasic Alcohol Response Differs in
Heavy Versus Light Drinkers
Andrea C. King, Tim Houle, Harriet de Wit, Louis Holdstock, and Alyson Schuster
Background: Most studies of risk factors for alcohol-related problems have focused on biological family
history as a primary risk factor. However, other factors, such as early-age heavy drinking, are also risk
factors for sustained or progressive heavy consumption. Little is currently known about the mechanisms
underlying binge or heavy drinking.
Methods: This study examined the acute subjective and objective effects of ethanol in heavy drinkers
versus light drinkers. Thirty-four subjects participated in this within-subjects study consisting of three
early-evening testing sessions in which subjects consumed a beverage containing either 0.8 or 0.4 g/kg
ethanol or placebo.
Results: Compared with lighter drinkers, heavy drinkers were more sensitive to the positive stimulantlike effects of ethanol (p ⬍ 0.05), especially during the increasing limb of the blood alcohol curve. Heavy
drinkers also showed less sedation and cortisol response after alcohol than the light drinkers (p ⬍ 0.05).
Conclusions: The results indicate that young adult binge drinkers show a biphasic alcohol response, with
heightened sensitivity to stimulant-like alcohol effects and greater tolerance to sedative alcohol effects
compared with their light-drinking counterparts.
Key Words: Alcohol Response, Heavy Drinker, Risk for Alcoholism, Biphasic Alcohol Effects Scale,
Cortisol.
H
EAVY ETHANOL DRINKING among young adults
is a serious problem in this country (Chou and Pickering, 1992), and the consequences of this excessive use,
both financial and personal, are widespread. The cost of
ethanol-related health care, loss of productivity, crime, and
accidents totals more than $148 million annually in the
United States alone (Harwood et al., 1999). Approximately
30% of all accidents are connected to ethanol use, and
excessive ethanol use contributes to many chronic and
life-threatening illnesses, including gastrointestinal, liver,
and cardiovascular disease (“Tenth Special Report to Congress on Alcohol and Health,” 2000). Data indicate that
those who start drinking before the age of 14 years are 12
times more likely to be injured in accidents while under the
influence of ethanol than those who start drinking after age
21 (Hingson et al., 2000). Early-onset alcohol drinking is
also strongly associated with lifetime alcohol problems
(Chou and Pickering, 1992). Understanding the factors that
contribute to the escalation and maintenance of excessive
ethanol drinking is crucial to improve prevention, public
From The University of Chicago, Pritzker School of Medicine, Department
of Psychiatry, Chicago, Illinois.
Received for publication September 13, 2001; accepted March 27, 2002.
Supported by the Alcoholic Beverage Medical Research Foundation (AK)
and NIH/NIAAA Grants AA-11133 (AK) and M01-RR00055.
Reprint requests: Andrea King, PhD, University of Chicago, Department of
Psychiatry, 5841 S. Maryland Ave., Chicago, IL 60637; Fax: 773-702-6454;
E-mail: aking@yoda.bsd.uchicago.edu
Copyright © 2002 by the Research Society on Alcoholism.
Alcohol Clin Exp Res, Vol 26, No 6, 2002: pp 827–835
education, and early intervention strategies. At this time, it
is still unclear why some individuals abuse ethanol and
others do not.
One potential source of vulnerability to developing alcohol use problems is the quality and magnitude of acute
subjective responses to alcohol (Fischman and Foltin,
1991). For instance, individuals who experience greater
stimulant-like effects from an acute dose of ethanol also
report greater drug liking and euphoria and have greater
behavioral preference for ethanol (over placebo) compared
with those individuals who experience mostly sedative-like
effects (Chutuape and de Wit, 1994; de Wit et al., 1987,
1989; Duka et al., 1998). Whether these differences in the
quality of the acute effects of alcohol also occur in individuals who are at risk for developing alcohol-related problems (e.g., because of their relatively high habitual consumption of alcohol is unclear). It is not known whether
these patterns of subjective responses to alcohol predispose
certain individuals to consume larger quantities of alcohol
or, viewed another way, whether individuals who consume
larger amounts of alcohol (i.e., heavy drinkers; HD) also
exhibit this pattern of enhanced stimulant-like responses to
alcohol. Heavy consumption of alcohol is in itself a potentially hazardous behavior, and it is also a risk factor for
developing more serious alcohol-related problems, such as
alcohol dependence.
Aside from the differences in stimulant effects of alcohol, vulnerability to alcoholism has also been linked to
other variations in acute responses to alcohol. Most of
827
828
these studies have been conducted by Schuckit and colleagues (Schuckit, 1984, 1985; Schuckit and Gold, 1988;
Schuckit et al., 1987, 1996) and have examined persons at
risk by virtue of a family history (FH) of alcoholism. Their
results have indicated that sons of alcoholics have diminished subjective ratings of sedation, less body sway, and
decreased cortisol responses to ethanol compared with sons
of nonalcoholics. Thus, they proposed that diminished responses to acute alcohol administration are associated with
risk for abuse. However, several other studies have failed to
support this idea. In fact, other researchers (Conrod et al.,
1997, 2001; Finn and Pihl, 1987; Finn et al., 1990; Gianoulakis et al., 1989, 1996; Peterson et al., 1996; Stewart et al.,
1992) have reported that sons of alcoholics exhibit enhanced responses to ethanol, including greater stress response dampening, greater heart rate increases, and enhanced ␤-endorphin–like responses after ethanol
compared with sons of nonalcoholics.
A differentiator model has been proposed to resolve
these apparently discrepant data (Newlin and Thomson,
1990, 1999). This model posits that high-risk subjects (i.e.,
FH positive; FH⫹) experience greater stimulant-like effects from ethanol on the ascending limb of the blood
alcohol curve [when blood alcohol concentrations (BACs)
are increasing] and lower sedative-like effects on the descending limb (when BACs are decreasing). Thus, high-risk
FH⫹ subjects would experience greater stimulant-like effects (which are typically reported to be positive) during
increasing BACs and would experience fewer sedative and
impairing effects (which are typically reported to be negative) during decreasing BACs. It remains to be seen
whether this model can be extended to other at-risk populations, such as young HDs.
Recently, we (Holdstock et al., 2000) reported findings
that provided a potential extension of the differentiator
model. We found that heavier social drinkers (range, 12–24
drinks weekly) reported greater stimulant-like subjective
effects to ethanol than lighter drinkers (fewer than five
drinks weekly). In addition, moderate/HDs tended to report fewer sedative-like and aversive effects from ethanol
than light drinkers (LD). Thus, there may be both qualitative and quantitative differences between light and moderate/HDs in the nature of their subjective responses to
ethanol, similar to other at-risk populations. These differences in subjective responses in our previous study occurred in the absence of differences in BACs, and they
were specific; i.e., they were not associated with alterations
in other subjective effects, adrenocortical hormone levels,
or general psychomotor performance. However, the results
were based on post hoc analyses of existing datasets with a
relatively small number of subjects.
The goal of this investigation was to extend the results of
our previous study by using a more robust methodology, a
larger sample size, and more carefully delineated groups of
drinkers. We examined the subjective and objective (cortisol and heart rate response) effects of ethanol on the rising
KING ET AL.
and falling portions of the BAC curve in young adults with
established HD patterns [a minimum of 10 drinks per week,
with at least one weekly binge occasion of 5 or more drinks
(4 drinks for women)] compared with LDs (1–5 drinks per
week with no binge occasions). In line with the differentiator model, we hypothesized that HDs would exhibit greater
stimulant-like effects on the ascending limb of the blood
alcohol curve and reduced sedative-like effects on the descending limb of the curve.
METHODS
Subjects
The subjects were 34 healthy male and female volunteers from the ages
of 24 to 38 years recruited through newspaper advertisements, word-ofmouth referrals, and local flyers. They were initially screened by telephone
to determine whether they met criteria for an LD or HD group. The
criteria for the LD group (n ⫽ 14; 10 men and 4 women) was consumption
of 5 or fewer drinks weekly (1–3 drinks per occasion, 1–2 times weekly)
and never exceeding 5 drinks per occasion (4 for women). To qualify for
the HD group (n ⫽ 20; 16 men and 4 women), respondents had to report
habitual, regular consumption (i.e., predominant pattern since age 21 or at
least the past 2 years) of a minimum of 10 drinks per week and consuming
at least 5 drinks on 1 occasion, 1 to 4 times each week (4 drinks for
women). Candidates who consumed an intermediate amount of alcohol or
whose consumption varied considerably over time were not accepted. The
drinking inclusion criteria for the HD group were based on laboratory,
epidemiological, and clinical studies defining a binge as five or more
drinks (four for women) consumed on one occasion. This level of drinking
departs from normative social drinking, may indicate aspects of loss of
control (Dawson, 2000; Dufour, 1999), and is frequently associated with
adverse consequences (Dawson, 1999; Single, 1996).
Eligible participants from the phone interview were invited to an
in-person screening session that consisted of questionnaires, an interview
by a trained master’s- or doctoral-level clinician, and a routine physical
examination by a resident physician. The screening questionnaires consisted of the Symptom Checklist (Derogatis, 1983), a quantity-frequency
index scale (Cahalan et al., 1969), the Brief Michigan Alcohol Screening
Test (Pokorny et al., 1972), and a family tree for both primary and
secondary biological relatives to assess two-generational FH for alcohol
dependence (American Psychiatric Association, 1994). Exclusion criteria
included a history of alcohol or substance dependence or regular heavy
use of drugs other than alcohol, moderate or variable drinking patterns,
psychiatric or medical disorders, clinically significant abnormalities on
screening laboratory tests (aspartate aminotransferase, alanine aminotransferase, bilirubin, and so on), or a positive urine drug screen (amphetamines, barbiturates, opiates, or cocaine). Female candidates were tested
for pregnancy before participating in the study and before each experimental session.
The protocol was approved by the University of Chicago Institutional
Review Board. Before participating in the study, subjects read and signed
the consent form. The consent form stated that the purpose of the study
was to assess mood responses to commonly used substances and that the
category of substances may include a stimulant, sedative, alcohol, or a
placebo. Each subject was instructed to abstain from any medication or
substance use for 12 hr before and after sessions and to abstain from
alcohol for 48 hr before participating. Also, the subject was instructed to
refrain from eating for at least 2 hr before the session. At the end of each
session, transportation was provided, and the subject was instructed not to
drive or operate heavy machinery for at least 12 hr after the session. After
completing the entire study, all subjects were debriefed and paid for their
participation.
829
BIPHASIC ALCOHOL RESPONSE DIFFERS IN HEAVY/LIGHT DRINKERS
Procedure
The study used a within-subject design. Each subject participated individually in three experimental sessions in which he or she received either
of two doses of ethanol (0.4 or 0.8 g/kg) or placebo administered in
random order and counterbalanced within the group. The ethanol drinks
consisted of 8% volume 180-proof alcohol (0.4 g/kg) and 16% volume
alcohol (0.8 g/kg) mixed with water, Nutrasweet, and sugar-free grape
Kool-Aid® (Kraft, Northfield, IL). The placebo beverage contained the
mix with 1% ethanol as a taste mask to reduce expectancy effects. The
study was conducted double-blind so that neither the subject nor the
experimenter knew the contents of each beverage. Sessions were conducted in a comfortable living room–like laboratory environment, with at
least 48 hr separating each session.
During each session, the subject arrived at 4:30 PM, was given a light
snack (noncaffeine, low-fat meal with 15% of calories from fat on the basis
of body weight), and completed several demographic and personality
questionnaires. At 5:15 PM, the subject submitted to a breathalyzer test,
saliva sample, and cardiovascular recordings and also completed baseline
questionnaires. Other noninvasive cognitive and performance testing was
obtained at various intervals as part of another investigation.
At 5:30 PM, the subject consumed the beverages in the presence of the
experimenter. The beverage was divided equally into two portions, each
consumed within a 5-min interval, with a 5-min rest in between (15 min
total consumption interval). The experimenter was present during the
entire beverage consumption and engaged in high conversation with the
subject. After finishing the beverages, the subject was allowed to read or
watch videotapes and completed subjective (questionnaires) and objective
(cardiovascular, saliva cortisol, BAC) measures 15, 45, 105, and 165 min
after the completion of drinking. In addition to the forms that subjects
completed, the experimenter returned to the testing room frequently, to
observe the activities of the subject and note any signs of intoxication. At
the end of the session, around 9:00 PM, subjects were transported home.
Dependent Measures—Subjective
Biphasic Alcohol Effects Scale (BAES). The BAES (Martin et al., 1993)
is a 14-item adjective-rating scale that is sensitive to ethanol-induced
stimulant- and sedative-like effects. The BAES has been validated to yield
higher stimulation scores during ascending BACs and higher sedation
during descending BACs (Martin et al., 1993). The subject indicates the
extent to which they are feeling each adjective on an 11-point scale from
“not at all” (0) to “extremely” (10). The Stimulation scale is measured by
summing the scores for the adjectives elated, energized, excited, stimulated,
talkative, up, and vigorous. The Sedation scale score is measured by
summing the descriptors down, heavy head, difficulty concentrating, inactive, sedated, slow thoughts, and sluggish.
Addiction Research Center Inventory (ARCI). The ARCI (Martin et al.,
1971) is a standardized questionnaire that consists of 49 true/false statements that are summed to five empirically derived scale scores. The ARCI
is a sensitive, reliable measure of the effects of specific drugs or classes of
drugs, including alcohol (e.g., Fischman and Foltin, 1991). The stimulantlike effects used from this scale were the Morphine-Benzedrine group
(MBG), which reflects drug-induced euphoria, and the Amphetamine (A)
scale, which provides a measure of stimulation. Sedative-like effects were
derived from the Pentobarbital-Chlorpromazine-Alcohol Group (PCAG)
scale.
Drug Effects Questionnaire (DEQ). The DEQ consists of four visual
analog scales for which subjects make a vertical mark on a 100-mm line for
each item. The scale measures positive drug effects in items such as: (a)
Like: “Do you like the effects you are feeling now?” [labeled dislike (0),
neutral (50), and like extremely (100)]; and (b) Want More: “Would you
like more of what you consumed, right now?” [labeled not at all (0), some
(50), and a lot (100)]. The DEQ Scale also measures general drug effects
in items such as (c) Feel: “Do you feel any drug effects?” and (d) High:
“Do you feel high?” [both items labeled not at all (0) and a lot (100)]. In
past studies, “high” has been shown to relate to some drug effects,
although not associated with stimulant like or alcohol effects (Martin et
al., 1993).
Dependent Measures—Objective
BAC. Instrumentation for estimating BACs from breath samples included the Alco-Sensor IV™ (Intoximeters Inc., St. Louis, MO). This
breathalyzer read 0.000 for all measures during testing, with actual levels
downloaded to a computer after the subject completed the study, thereby
ensuring that the experimenter was blinded to alcohol dose condition.
Heart Rate. An automated monitor (Laptron®, Kappa Medical Incorporated, Hauppauge, NY) measured heart rate (in beats per minute) at
each time point and was recorded by the research assistant.
Cortisol. Saliva samples were collected with plain cotton Salivettes™
(Sarstedt, Newton, NC) and stored in a ⫺20°C freezer until assayed by
enzyme immunoassay (catalog No. 0101™, Salimetrics, State College, PA).
Samples were thawed, centrifuged, and assayed in duplicate. The assay
method was standardized and validated in the University of Chicago
Clinical Research Center core laboratory.
Data Analysis
Because this study was primarily concerned with the differential stimulant and sedative effects of alcohol, the main dependent variables were
the Stimulation and Sedation scales on the BAES. These data were
analyzed by 2 ⫻ 3 ⫻ 5 repeated-measures analyses of variance (ANOVAs), with group (HD and LD) as a between-subjects factor and dose
(placebo, 0.4 g/kg, and 0.8 g/kg) and time (baseline and 15, 45, 105, and
165 min) as within-subjects factors. To further examine the subjective
stimulant- and sedative-like effects of alcohol, subscales from the ARCI
and DEQ, as well as the objective measures of salivary cortisol and heart
rate, were also analyzed in repeated-measures ANOVAs.
Because of the rising and falling nature of BAC in an acute administration paradigm, it was expected that subjective effects might also vary in
accordance with these levels. Therefore, a priori analyses examined the
groups on curvilinear functions of different degrees (linear, quadratic,
cubic, etc.) that represent an increase, peak, and decrease for the main
dependent variables (BAES Stimulation and Sedation). These preliminary
examinations with polynomial functions represent unique information
compared with that of the traditional ANOVA and should not be interpreted as indicating a significant interaction unless otherwise stated.
For all analyses, F values were considered significant at p ⬍ 0.05. To
protect against violations of sphericity, Greenhouse-Geisser corrections
were used to adjust the degrees of freedom of the F tests. Finally, in
accordance with the suggestions made by Schmidt (1996), estimates of
effect size will also be reported (␩2) as warranted.
RESULTS
Background Characteristics
Table 1 displays the general demographic and alcoholdrinking characteristics for the HD and LD groups. The
groups were similar on most background characteristics,
except that LDs had more years of education than HDs
(17.2 vs. 15.2 years). By definition, the HDs consumed
more alcoholic drinks per typical drinking occasion, had
more drinking occasions per week, had a greater frequency
of binge drinking [i.e., five or more drinks per occasion
(four for females)] over the last 6 months, and had a
greater lifetime prevalence of Alcohol Abuse (58 vs. 0% in
LDs). Ten subjects in the HD group reported smoking
cigarettes daily, although 60% of these reported smoking
10 or fewer cigarettes per day. In contrast, only one of the
LDs reported current smoking. There were no significant
830
KING ET AL.
Table 1. Sample/Drinking Characteristics by Group
Variable
Demographics
Age (years)
Education (years)
Race (black/white)
Sex (M/F)
Liver function
Bilirubin (total)
Alkaline phosphate
SGOT
SGPT
Drinking behaviora
Drinks/occasion
Occasions/week
5⫹ drink occasions
(last 6 months)
Family historyb ⫹
Alcohol Abuse
(lifetime, DSM-IV)
Smoking behavior
% Current smokerc
Cigarettes/day
Light drinkers
Heavy drinkers
28.7 (0.72)
17.2 (0.56)*
5/8
10/4
28.5 (0.82)
15.2 (0.34)
5/15
16/4
.72 (0.10)
61.2 (4.08)
28.2 (2.59)
24.9 (2.9)
.72 (0.06)
70.0 (4.15)
24.1 (0.92)
24.8 (2.19)
1.9 (0.21)*
1.2 (0.14)*
.54 (0.22)*
31%
0%*
7%*
.07 (0.27)*
4.8 (0.43)
3.4 (0.18)
45 (6.1)
50%
58%
50%
4.6 (6.72)
Data presented are mean (SEM) or % where indicated.
SGOT, serum glutamic oxaloacetic transaminase; SGPT, alanine transaminase.
a
Drinking behavior derived from 6-month quality-frequency index.
b
FH⫹, one or both parents or three or more secondary relatives with alcohol
abuse or dependence.
c
Current smoker, reporting smoking one or more cigarettes weekly.
* p ⬍ 0.05.
differences in subjective responses to alcohol between the
HD smokers and nonsmokers during the sessions (smoking
was not permitted during the sessions). Demographic and
alcohol consumption data were nearly identical for the HD
smokers and nonsmokers, except for a greater frequency of
binge occasions in the last six months in the former versus
the latter group (mean, 63 vs. 35 occasions, respectively; p
⬍ 0.01).
BACs
As expected, BAC levels significantly increased over time
and as a function of increasing alcohol dose [dose, F(2,58);
and time, F(4,116) ⬎ 296.66, p ⬍ 0.0001; dose ⫻ time,
F(8,232) ⫽ 158.45, p ⬍ 0.0001; see Fig. 1]. The LD and HD
groups exhibited nearly identical BAC curves (F ⬍ 1.02; ps
⫽ ns).
Alcohol Effects
Alcohol increased BAES Stimulation scores during the
initial rising BACs and increased Sedation scores over
declining BACs [dose ⫻ time, Fs (8,248) ⱖ 2.29; ps ⬍0.05].
During the rising BAC, alcohol also increased other
stimulant-like and general drug effects derived from the
ARCI and DEQ scales (ARCI: A, MBG; DEQ: Like, Feel,
and High scales; dose ⫻ time, ps ⬍0.05) and increased the
DEQ rating of Want More [dose, F(1,30) ⫽ 4.67; p ⬍ 0.05].
Additionally, alcohol produced increases in other sedativelike effects during declining BACs, such as the ARCIPCAG scale (dose ⫻ time, p ⬍ 0.0001). For the objective
measures, alcohol increased heart rate on both the rising
and declining limbs of the BAC curve [dose, F(2,58) ⫽ 5.14,
p ⬍ 0.01]. After covarying for baseline differences in cortisol between sessions, alcohol tended to increase cortisol
levels, although this was nonsignificant [dose ⫻ time,
F(6,174) ⫽ 1.92; p ⫽ 0.08].
Post hoc comparisons for most of these dependent variables indicated that the effects of ethanol were mainly
apparent at the high dose compared with placebo (ps ⬍
0.05, simple effects), whereas the low dose produced minimal or subtle mood changes. Therefore, the ensuing results on alcohol response refer primarily to the effects of
the high alcohol dose, with the low dose producing a minimal or intermediate response, unless otherwise indicated.
Group Comparisons on Alcohol-Related Effects
Alcohol increased BAES Stimulation scores in the HD
group but not in the LD group [group ⫻ dose, F(2,62) ⫽
3.13; p ⫽ 0.05]. A priori analyses examining effects over the
course of the BAC curve showed that alcohol produced a
rapid, short-lived increase in Stimulation, with a slowing
rate of decline [dose (linear) ⫻ time (cubic): F(1,18) ⫽
Fig. 1. Comparison of blood alcohol curve
estimates from breathalyzer tests in the HD and
LD groups for placebo, 0.4 g/kg (low dose), and
0.8 g/kg (high dose). One HD subject was removed from analyses as an outlier. The x axis
time points indicate the following: base, measurement taken immediately before beverage
consumption; and time points, time after the
beverages were consumed. Dose ⫻ time interaction, p ⬍ 0.0001. 䡲, High dose; Œ, Low dose,
䡩, Placebo.
BIPHASIC ALCOHOL RESPONSE DIFFERS IN HEAVY/LIGHT DRINKERS
831
Fig. 2. Mean stimulation subscale scores
from the BAES for the HD and LD groups.
Data correspond to the time points before
beverage consumption (base) and after the
beverages were consumed. For the HD group,
data fit a curvilinear function (dose ⫻ cubic
time, p ⫽ 0.001). 䡲, High dose; Œ, Low dose,
䡩, Placebo.
16.52, p ⫽ 0.001; ␩2 ⫽ 0.48; see Fig. 2]. Inspection of the
means indicated that alcohol increased Stimulation in HDs
at the 15 min postconsumption time point. In comparison,
after drinking the placebo beverage, the HD group exhibited decreased Stimulation during this acute, rising BAC
time point (p ⬍ 0.05). In the LD group, neither alcohol nor
placebo beverage produced significant changes in selfreported Stimulation [F(1,13) ⫽ 0.41; p ⫽ not significant].
Alcohol also produced similar sharp and rapid increases
in the HDs, but not LDs, on the other measures of
stimulant-like effects during rising BACs, including the
ARCI A and MBG scales [dose (linear) ⫻ time (cubic):
F(1,19) ⱖ 13.94, p ⱕ0.001; ␩2 ⱖ 0.42]. Enhanced alcohol
sensitivity on the DEQ Like scale [group ⫻ dose, F(1,30) ⫽
6.14; p ⬍ 0.05] was also apparent in the HDs, as was a trend
for increased ratings during rising BACs of DEQ Want
More (group ⫻ dose, p ⫽ 0.07). Alcohol increased ratings
(above neutral) for Like and Want more at rising BACs in
90 and 56% of HDs, respectively, compared with only 38
and 31% of LDs, and the relationship between these two
Fig. 3. Mean sedation subscale scores
from the BAES for the HD and LD groups.
Data correspond to the time points before
beverage consumption (base) and after the
beverages were consumed. For both groups,
data fit a curvilinear function (dose ⫻ cubic
time, p ⬍ 0.05). 䡲, High dose; Œ, Low dose,
䡩, Placebo.
psychological constructs was significant [r(31) ⫽ ⫹0.64; p ⬍
0.001]. In contrast, on the more general scales of drug
effects, including the DEQ Feel Drug and High, alcohol
produced greater rising-BAC increases in LDs compared
with HDs, and this augmented response was apparent even
at the lower alcohol dose (group ⫻ dose ⫻ time, ps ⱕ0.01).
Alcohol increased BAES Sedation scores to a greater
extent in the LD group than the HD group [group ⫻ dose,
F(2,62) ⫽ 5.33, p ⬍ 0.01; ␩2 ⫽ 0.15]. This was supported in
our a priori examination of dynamic effects over time for
the groups [group ⫻ dose (linear) ⫻ time (cubic), F(1,31)
⫽ 5.67, p ⫽ 0.024; ␩2 ⫽ 0.16]. For the LDs, increases in
Sedation emerged during the early portion of the BAC
curve and were sustained throughout testing [dose (linear)
⫻ time (cubic), F(1,13) ⫽ 4.79, p ⬍ 0.05; see Fig. 3].
However, in the HD group, alcohol increased Sedation
ratings, but this was primarily during declining BACs [dose
(linear) ⫻ time (cubic), F(1,18) ⫽ 5.13; p ⬍ 0.05]. Similarly,
alcohol increased ratings on the other measure of sedativelike effects, the ARCI-PCAG scale, to a greater extent in
832
KING ET AL.
Fig. 4. Mean saliva cortisol response for
the HD and LD groups. Data correspond to
the time points before beverage consumption
(base) and after the beverages were consumed. 䡲, High dose; Œ, Low dose, 䡩,
Placebo.
the LD group and did so earlier in the BAC curve than in
the HD group [group, p ⬍ 0.05; dose, p ⬍ 0.001; dose
(linear) ⫻ time (quadratic), F(8,248) ⫽ 7.34, p ⬍ 0.001].
Alcohol increased cortisol levels in the LD group, but not
in the HD group [group ⫻ dose, F(2,58) ⫽ 3.78; p ⬍ 0.05].
This cortisol increase in LDs occurred during the later
portions of the BAC curve [i.e., 165 min; p ⬍ 0.05, simple
effects; see Fig. 4], following the time course for LD/HD
differences in Sedative-like responses. Finally, the HD
group exhibited a significantly higher heart rate than the
LDs at baseline and throughout the study, regardless of
whether they received alcohol or placebo [group, F(1,29) ⫽
4.41; p ⬍ 0.05], but the groups did not differ in their heart
rate increases to alcohol (ANCOVA: group ⫻ time, group
⫻ dose, ps ⫽ not significant).
Relationships Between Main Measures
Although the BAES and ARCI scales were developed by
using different methodologies and subject populations
(normal drinkers and experienced addicts, respectively),
there were significant correlations between these measures:
BAES Stimulation was positively associated with the ARCI
A and MBG scales during rising BACs [r(33) ⬎ 0.42; ps
⬍0.05], with ARCI A and MBG scales highly associated
with each other [r(33) ⬎ 0.71; ps ⬍0.001]. Results for the
BAES Sedation scale were positively associated with the
ARCI PCAG scale during declining BACs [r(33) ⬎ 0.55; ps
⬍0.001]. Sedation, but not Stimulation, was also correlated
with the cortisol response to alcohol [r(31) ⫽ 0.37; p ⬍
0.05].
Examination of FH Effects
Although this study was not designed to specifically address the effects of subjects’ differential FH of alcoholism,
information about FH was obtained during screening as
part of an attempt to equate groups on this variable. Chisquare analysis revealed that the HD and LD groups did
not differ in incidence of FH⫹, defined as having at least
one parent or three or more secondary relatives with alcoholism (50 vs. 31%; Table 1). Furthermore, for the main
variables of interest—the BAES Stimulation and Sedation
scales, as well as the objective measures of cortisol and
heart rate—FH as a covariate did not alter the significant
effects of drinking group. The only exception noted was a
trend for the subgroup of HDs with FH⫹ to show the most
blunted cortisol response to alcohol (FH ⫻ group ⫻ dose ⫻
time, p ⫽ 0.08).
DISCUSSION
In this study, we observed that HDs, compared with LDs,
were more sensitive to the positive or euphorigenic subjective effects of alcohol during the early portion of the BAC
and less sensitive to the sedative-like effects of alcohol
during both rising and declining BACs. These results were
obtained on a range of measures, including both alcoholspecific scales (BAES Stimulation and Sedation) and
stimulant- and sedative-like drug effect scales (ARCI and
DEQ). The HD subjects, compared with LD subjects, also
had lower salivary cortisol levels during declining BACs.
Although some studies have reported greater alcoholinduced heart rate changes in subjects at risk for alcoholism
(Conrod et al., 1997, 2001; Peterson et al., 1996) and a
positive relationship between heart rate increase and selfreported stimulation during rising BACs (Conrod et al.,
2001), we did not observe differential effects of alcohol on
heart rate changes between the HD and LD groups. This
discrepancy between our study and other studies may be
due to differences in the risk groups examined (FH⫹ versus
HD) and in frequency of measurement and other instrumentation factors.
BIPHASIC ALCOHOL RESPONSE DIFFERS IN HEAVY/LIGHT DRINKERS
The collective pattern of response to alcohol observed on
a variety of dimensions in this study provides partial support
for the differentiator model (Newlin and Thomson, 1990,
1999). Newlin and Thomson proposed in their model that
individuals with an FH of alcoholism would experience
more stimulant effects on the ascending limb and less
sedation on the descending limb of the BAC. Our study
extends these findings to another at-risk group, i.e., young
people who habitually engage in heavy social drinking beyond the college years. The most striking subjective differences between our groups occurred during the rising portion of the BAC (Figs. 2 and 3). During the ascending limb
of the BAC, the HDs showed greater subjective stimulation
and less sedation than LDs, but the differentiator model
proposes that the sedation effect would occur only during
declining BACs. Thus, this initial pattern after consumption of a moderate-to-high dose of alcohol, if supported in
future studies or in other risk groups, may support an
alternative “rising limb theory” that may be more applicable than the differentiator model to account for enhanced
risk for the development and maintenance of alcohol use
disorders.
One important question that arises is whether the heightened alcohol stimulation and reduced sedation in HDs
represent an acquired response to alcohol, evident only
after multiple repeated exposures, or an inherent predisposition, or premorbid individual difference. In terms of
the latter, because the data are cross-sectional in nature, we
can only speculate that HDs might have been more likely to
have early initial alcohol sensitivity to stimulation, tolerance to sedation, or both, even during their first exposure.
This initial pattern of positive reinforcement without concomitant negative effects could be an important factor
rendering them at risk for the acquisition of binge-drinking
patterns. Unfortunately, due to ethical constraints in testing this model in human subjects, such as performing alcohol challenge studies on children or adolescents, and methodological difficulties, such as examining retrospective reports
of initial exposures in adults (Schuckit et al., 2001), the premorbid sensitivity hypothesis will remain speculative.
With respect to the former hypothesis (acquired alcohol
response), it is possible that after repeated exposures, HDs
become sensitized to the euphoric effects (or liking) of
alcohol and then are more likely to continue to engage in
HD. Alternatively, a recent incentive-sensitization theory
postulates that the repeated use of addictive substances,
including alcohol, sensitizes brain systems that mediate
incentive motivation or reward (or wanting), resulting in
further use (Robinson and Berridge, 1993, 2001). Because
our results showed that HDs exhibited heightened alcohol
liking and wanting, especially during the rising limb, and
that these two factors were positively correlated with each
other, these effects do not seem to be dissociated in this
sample. Finally, it is possible that acquired tolerance to the
sedative-like effects of alcohol, especially during rising
BACs, may be the key component that contributes to in-
833
creased heavy drinking levels (Kalant, 1998; Schuckit,
1994). Schuckit and colleagues have shown that low level of
alcohol response, as measured by subjective (mainly
sedative-like) and physiologic measures, was evident in
people with an FH of alcoholism and also predicted future
drinking (Schuckit, 1994). However, as Conrod et al. (2001)
point out, this theory is somewhat inconsistent with most
current theories of the mechanisms of motivational states
and drug-seeking behavior (i.e., addictive or rewarding
properties).
Several neurobiological systems may be involved in the
stimulant effects of alcohol, including dopaminergic, opioidergic, and serotonergic neurotransmitter systems. The mesolimbic dopamine system, through pathways connected to
the nucleus accumbens, has been postulated as the final
common pathway for all drugs of abuse, including alcohol
(Leshner, 1997; Wise and Bozarth, 1987). Two recent findings provide some support for dopamine involvement.
First, acute administration of the dopamine antagonist haloperidol blocks alcohol craving in alcoholic subjects (Modell et al., 1993), and second, in healthy volunteers, haloperidol treatment reduced alcohol-related increases in BAES
Stimulation (Enggasser and de Wit, 2001). Evidence for the
involvement of the endogenous opioid system in alcohol
euphoria comes from the finding that naltrexone, an opioid
antagonist, reduces ethanol intoxication in social drinkers
(Davidson et al., 1996; Swift et al., 1994), especially those
with an FH of alcoholism (King et al., 1997). Finally, there
is also limited evidence that serotonin is involved in alcohol
reward (Meert et al., 1991; Myers et al., 1992), but these
findings are mixed in human clinical trials (Johnson et al.,
1996, 2000). Future investigations examining acute alcohol
response paradigms with pharmacological manipulations or
brain imaging techniques may help elucidate the mechanisms and role of several potential neurobiological systems
in alcohol sensitivity and tolerance.
There are two other important findings in this study that
are worth discussion. First, the HD subjects demonstrated
a rapid decrease in subjective stimulation in the HDs during the placebo beverage session, in contrast to their rapid
increases in stimulation after consuming alcohol (Fig. 2).
This differential effect may represent an opponent process
(Siegel, 1983) or an antagonistic placebo response (Newlin,
1985), such that the high-risk HD, when exposed to
alcohol-related cues (i.e., the placebo contained 1% alcohol as a taste mask), exhibits conditioned responses in the
direction opposite to the drug effect. Paradigms involving
more extensive, repetitive doses of alcohol or alcoholrelated cues may help elucidate the role of this differential
placebo response to the etiology of alcohol-related disorders—a link that has yet to be established (Newlin, 1985).
Second, the LDs’ absence of alcohol stimulation, in combination with increased sedation and stress hormone levels
after alcohol consumption, may be an important factor
rendering them at very low risk for the development of
future alcohol disorders. Understanding the mechanisms
834
KING ET AL.
underlying LDs’ alcohol response, a group putatively at low
risk for harmful drinking, may provide clues as to the
protective factors associated with LD, beyond racial or
ethnic characteristics. The majority of research on people
at low risk for future alcoholism has largely been confined
to studies of individuals with unusual genetic variants
(Agarwal et al., 1981; Wall et al., 1992; Yoshida et al.,
1991).
This study had several limitations. First, our sample size
was modest; therefore, analyses on the basis of sex or
individual difference factors could not be fully assessed.
For example, we were unable to compare men and women.
Although the prevalence of heavy alcohol use is higher in
men, there has been a steady increase in binge drinking in
young women (Dawson, 1996; Johnston et al., 2000), and
the factors controlling drinking in women may be different
than in men (King et al., 2002). Second, although the
groups were similar on most demographic variables measured, the HDs had fewer years of education and higher
smoking rates than LDs. Although these factors were not
significant covariates in our analyses, it remains to be determined whether they are a factor in terms of etiology
and/or maintenance of binge drinking. Third, the data do
not provide longitudinal information, and therefore any
conclusions remain speculative until prospective or largescale studies are conducted.
In summary, although early onset of drinking (Chou and
Pickering, 1992) and engagement in HD both strongly predict progression to future alcohol misuse, the factors or
mechanisms underlying earlier-stage HD are not well understood. This study demonstrates biphasic alcohol responses, including rising BAC–associated increases in stimulation and declining BAC increases in sedation, in young
HD/binge drinkers compared with light social drinkers. In
contrast, LDs did not show biphasic alcohol response, with
no increases in stimulation, and heightened sedation
throughout both limbs and cortisol increases during the
declining limb. The collective pattern of results may highlight the potential relative importance of changes during
the ascending limb of the BAC in risk for alcohol use
disorders. Finally, continued examination of factors involved in various risk groups beyond FH of alcoholism is
warranted given the multiple etiologies for the complex
disorders of alcohol abuse and dependence. Acute subjective and objective responses to alcohol remain an important
potential determinant along with the myriad of environmental, psychological, and genetic influences on the degree
of risk for alcohol-related problems.
ACKNOWLEDGMENTS
We appreciate the support of The University of Chicago Clinical Research Center (CRC). In particular, we thank Sujata Patel,
Core Lab Director, for the cortisol assays; Veronica Wald, CRC
administrative director, for oversight and organizational support;
and the entire CRC nursing and dietary staff for their assistance.
Paul Meyer was also helpful in the early data-collection phase of
this study. Finally, we thank Emil Coccaro, MD, for helpful suggestions on this manuscript.
REFERENCES
Agarwal DP, Harada S, Goedde HW (1981) Racial differences in biological sensitivity to ethanol: the role of alcohol dehydrogenase and aldehyde dehydrogenase isozymes. Alcohol Clin Exp Res 5:12–16.
American Psychiatric Association (1994) Diagnostic and Statistical Manual
of Mental Disorders. 4th ed. American Psychiatric Press, Washington,
DC.
Cahalan V, Cisin I, Crossley HM (1969) American Drinking Practices.
Rutgers Center for Alcohol Studies, New Brunswick, NJ.
Chou SP, Pickering RP (1992) Early onset of drinking as a risk factor for
lifetime alcohol-related problems. Br J Addict 87:1199 –1204.
Chutuape MA, de Wit H (1994) Relationship between subjective effects
and drug preferences: ethanol and diazepam. Drug Alcohol Depend
34:243–251.
Conrod PJ, Peterson JB, Pihl RO (2001) Reliability and validity of
alcohol-induced heart rate increase as a measure of sensitivity to the
stimulant properties of alcohol. Psychopharmacology 157:20 –30.
Conrod PJ, Peterson JB, Pihl RO, Mankowski S (1997) Biphasic effects of
alcohol on heart rate are influenced by alcoholic family history and rate
of alcohol ingestion. Alcohol Clin Exp Res 21:140 –149.
Davidson D, Swift R, Fitz E (1996) Naltrexone increases the latency to
drink alcohol in social drinkers. Alcohol Clin Exp Res 20:732–739.
Dawson D (1996) Gender differences in the risk of alcohol dependence:
United States, 1992. Addiction 91:1831–1842.
Dawson D (1999) Alternative definitions of high risk for impaired driving:
the overlap of high volume, frequent heavy drinking and alcohol dependence. Drug Alcohol Depend 54:219 –228.
Dawson D (2000) Drinking patterns among individuals with and without
DSM-IV alcohol use disorders. J Stud Alcohol 61:111–120.
Derogatis L (1983) SCL-90-R Manual—II. Clinical Psychometric Research, Towson, MD.
de Wit H, Pierri J, Johanson CE (1989) Assessing pentobarbital preference in normal volunteers using a cumulative dosing procedure. Psychopharmacology 99:416 – 421.
de Wit H, Uhlenhuth EH, Pierri J, Johanson CE (1987) Individual differences in behavioral and subjective responses to alcohol. Alcohol Clin
Exp Res 11:52–59.
Dufour MC (1999) What is moderate drinking? Defining “drinks” and
drinking levels. Alcohol Res Health 23:1–14.
Duka T, Stephens DN, Russell C, Tasker R (1998) Discriminative stimulus properties of low doses of ethanol in humans. Psychopharmacology
136:379 –389.
Enggasser J, de Wit H (2001) Haloperidol reduces stimulant and reinforcing effects of ethanol in social drinkers. Alcohol Clin Exp Res
25:1448 –1456.
Finn PR, Pihl RO (1987) Men at risk for alcoholism: the effect of alcohol
on cardiovascular response to unavoidable shock. J Abnorm Psychol
96:230 –236.
Finn PR, Zeitouni NC, Pihl RO (1990) Effects of alcohol on psychophysiological hyperreactivity to nonaversive and aversive stimuli in men at
high risk for alcoholism. J Abnorm Psychol 99:79 – 85.
Fischman MW, Foltin RW (1991) Utility of subjective-effects measurements in assessing abuse liability of drugs in humans. Br J Addict
86:1563–1570.
Gianoulakis C, Beliveau D, Angelogianni P, Meaney M, Thavundayil J,
Tawar V, Dumas M (1989) Different pituitary ␤-endorphin and adrenal
cortisol response to ethanol in individuals with high and low risk for
future development of alcoholism. Life Sci 45:1097–1109.
Gianoulakis C, Krishnan B, Thavundayil J (1996) Enhanced sensitivity of
pituitary ␤-endorphin to ethanol in subjects at high risk of alcoholism.
Arch Gen Psychiatry 53:250 –257.
BIPHASIC ALCOHOL RESPONSE DIFFERS IN HEAVY/LIGHT DRINKERS
Harwood HJ, Fountain D, Fountain G (1999) Economic cost of alcohol
and drug abuse in the United States, 1992: a report. Addiction 94:631–
635.
Hingson RW, Heeren T, Jamanka A, Howland J (2000) Age of drinking
onset and unintentional injury involvement after drinking. JAMA 284:
1527–1533.
Holdstock L, King AC, de Wit H (2000) Subjective and objective responses to ethanol in moderate/heavy and light social drinkers. Alcohol
Clin Exp Res 24:789 –794.
Johnson BA, Jasinski DR, Galloway GP, Kranzler H, Weinreib R, Anton
RF, Mason BJ, Bohn MJ, Pettinati HM, Rawson R, Clyde C (1996)
Ritanserin in the treatment of alcohol dependence—a multi-center
clinical trial. Psychopharmacology 128:206 –215.
Johnson BA, Roache JD, Javors MA, DiClemente CC, Cloninger CR,
Prihoda TJ, Bordnick PS, Ait-Daoud N, Hensler J (2000) Ondansetron
for reduction of drinking among biologically predisposed alcoholic patients. A randomized controlled trial. JAMA 284:963–971.
Johnston LD, O’Malley PM, Bachman JG (2000) Monitoring the Future
National Survey Results on Drug Use, 1975–1999. Volume 2: College
Students and Young Adults (NIH Publication No. 00-4803). National
Institute on Drug Abuse, Bethesda, MD.
Kalant H (1998) Research on tolerance: what can we learn from history?
Alcohol Clin Exp Res 22:67–76.
King AC, Bernardy NC, Hauner K (2002) Stressful events, personality,
and mood disturbance: gender differences in alcoholics and problem
drinkers. Addict Behav, in press.
King AC, Volpicelli JR, Frazer A, O’Brien CP (1997) Effect of naltrexone
on subjective alcohol response in subjects at high and low risk for future
alcohol dependence. Psychopharmacology 129:15–22.
Leshner AI (1997) Addiction is a brain disease, and it matters. Science
278:45– 47.
Martin CS, Earleywine M, Musty RE, Perrine MW, Swift RM (1993)
Development and validation of the Biphasic Alcohol Effects Scale.
Alcohol Clin Exp Res 17:140 –146.
Martin WR, Sloan JW, Sapira JD, Jasinski DR (1971) Physiologic, subjective and behavioral effects of amphetamine, methamphetamine,
ephedrine, phenmetrazine and methylphenidate in man. Clin Pharmacol Ther 12:245–258.
Meert TF, Awouters F, Niemegeers CJ, Schellekens KH, Janssen PA
(1991) Ritanserin reduces abuse of alcohol, cocaine, and fentanyl in
rats. Pharmacopsychiatry 24:159 –163.
Modell JG, Mountz JM, Glaser FB, Lee JY (1993) Effect of haloperidol
on measures of craving and impaired control in alcoholic subjects.
Alcohol Clin Exp Res 17:234 –240.
Myers RD, Lankford M, Björk A (1992) Selective reduction by the 5-HT
antagonist amperozide of alcohol preference induced in rats by systemic
cyanamide. Pharmacol Biochem Behav 43:661– 667.
Newlin DB (1985) Offspring of alcoholics have enhanced antagonistic
placebo response. J Stud Alcohol 46:490 – 494.
Newlin DB, Thomson JB (1990) Alcohol challenge with sons of alcoholics:
a critical review and analysis. Psychol Bull 108:383– 402.
Newlin DB, Thomson JB (1999) Chronic tolerance and sensitization to
alcohol in sons of alcoholics: II. Replication and reanalysis. Exp Clin
Psychopharmacol 7:234 –243.
835
Peterson JB, Pihl RO, Gianoulakis C, Conrod P, Finn PR, Stewart SH,
LeMarquand DG, Bruce KR (1996) Ethanol-induced change in cardiac
and endogenous opiate function and risk for alcoholism. Alcohol Clin
Exp Res 20:1542–1552.
Pokorny AD, Miller BA, Kaplan HB (1972) The brief MAST: a shortened
version of the Michigan Alcoholism Screening Test. Am J Psychiatry
129:342–345.
Robinson TE, Berridge KC (1993) The neural basis of drug craving: an
incentive-sensitization theory of addiction. Brain Res Brain Res Rev
18:247–291.
Robinson TE, Berridge KC (2001) Incentive-sensitization and addiction.
Addiction 96:103–114.
Schmidt FL (1996) Statistical significance testing and cumulative knowledge in psychology implications for training of researchers. Psychol
Meth 1:115–129.
Schuckit MA (1984) Subjective responses to alcohol in sons of alcoholics
and control subjects. Arch Gen Psychiatry 41:879 – 884.
Schuckit MA (1985) Ethanol-induced changes in body sway in men at high
alcoholism risk. Arch Gen Psychiatry 42:375–379.
Schuckit MA (1994) Low level of response to alcohol as a predictor of
future alcoholism. Am J Psychiatry 151:184 –189.
Schuckit MA, Edenberg HJ, Kalmijn J, Flury L, Smith TL, Reich T, Bierut
L, Goate A, Foroud T (2001) A genome-wide search for genes that
relate to a low level of response to alcohol. Alcohol Clin Exp Res
25:323–329.
Schuckit MA, Gold E, Risch C (1987) Plasma cortisol levels following
ethanol in sons of alcoholics and controls. Arch Gen Psychiatry 44:942–
945.
Schuckit MA, Gold EO (1988) A simultaneous evaluation of multiple
markers of ethanol/placebo challenges in sons of alcoholics and controls. Arch Gen Psychiatry 45:211–216.
Schuckit MA, Tsuang JW, Anthenelli RM, Tipp JE, Nurnberger JI (1996)
Alcohol challenges in young men from alcoholic pedigrees and control
families: a report from the COGA project. J Stud Alcohol 57:368 –377.
Siegel S (1983) Classical conditioning, drug tolerance, and drug dependence, in Research Advances in Alcohol and Drug Problems (Smart RG,
Glaser FB, Isreal Y, Kalant H, Popham RE, Schmidt W eds), vol 7, pp
207–246. Plenum Press, New York.
Single E (1996) Harm reduction as an alcohol-prevention strategy. Alcohol Health Res World 20:239 –243.
Stewart SH, Finn PR, Pihl RO (1992) The effects of alcohol on the
cardiovascular stress response in men at high risk for alcoholism: a dose
response study. J Stud Alcohol 53:499 –506.
Swift RM, Whelihan W, Kuznetsov O, Buongiorno G, Hsuing H (1994)
Naltrexone-induced alterations in human ethanol intoxication. Am J
Psychiatry 151:1463–1467.
Tenth Special Report to Congress on Alcohol and Health (2000) National
Institute on Alcohol Abuse and Alcoholism, Bethesda, MD.
Wall TL, Thomasson HR, Schuckit MA, Ehlers CL (1992) Subjective
feelings of alcohol intoxication in Asians with genetic variations of
ALDH2 alleles. Alcohol Clin Exp Res 16:991–995.
Wise RA, Bozarth MA (1987) A psychomotor stimulant theory of addiction. Psychol Rev 94:469 – 492.
Yoshida A, Hsu LC, Yasunami M (1991) Genetics of human alcoholmetabolizing enzymes. Prog Nucleic Acid Res Mol Biol 40:225–287.