THE HFG REVIEW OF RESEARCH (Vol. 3, No. 1, Spring 1999)

Alcohol: The Aggression Elixir?
Peter Giancola

There are those who argue that alcohol is a very strong "elixir" for aggressive behavior—that alcohol causes aggression. We have all known or heard about people who, when sober, are fun-loving, sociable, and well-tempered, but, after a few drinks, undergo a "transformation of personality" characterized by loudness, anger, belligerence, provocation, and sometimes violence. Research evidence tends to support the alcohol-aggression link, indicating that alcohol consumption is strongly linked to verbal aggression, aggressive threats, family violence, marital aggression, violence-related emergency-room visits, child abuse, firearm use, sexual aggression, homicide, and suicide. A well-designed study of alcohol involvement in violence found that in 40-50% of violent incidents, either the victim, the assailant, or both had been drinking (Pernanen, 1991). While there is considerable variability among studies in estimated magnitude of alcohol involvement, some suggest rates of alcohol presence at the time of offending as high as 86% for homicide offenders, 72% for robbery offenders, 60% for sexual offenders, 70% for suicide attempters, 57% for marital violence perpetrators, and 54% for child molesters (Roizen, 1993).

However, there is another side to this coin. There are also the stories of people who behave quite aggressively without the help of alcohol and those who, no matter how much they drink, will never utter an angry word or raise a hand to anyone—alcohol just seems to facilitate joviality, emotional warmth, or sleep in these people. Thus, it may be more accurate to say that alcohol is an aggression elixir only in some persons. Clearly then, the task for researchers is to determine what traits characterize persons who are most at risk for becoming violent while under the effects of alcohol. Identifying those most at risk will help better elucidate the causes of alcohol-related violence and it will also help researchers craft better prevention and treatment interventions.

Measuring the Relationship | Before attempting to delineate a "risk profile" for alcohol-related aggression, it is important to discuss some different approaches to assessing the alcohol-aggression relationship. One very well-known approach involves simply asking people about their past aggressive behavior and whether or not alcohol was involved. A number of widely used self-report and interview-type formats have been designed to record these behaviors.

A second approach, not as well known, is an in vivo assessment of the actual behavior of physical aggression. Such an assessment almost always occurs in a laboratory setting. Most studies that have measured aggression in such a manner have used the Taylor Aggression Paradigm (TAP; Taylor, 1967) the Point Subtraction Aggression Paradigm (PSAP; Cherek, 1981), or a modified version of one of these protocols. The tap involves instructing research subjects that they are competing against another, unseen person on a reaction-time task. Following a winning trial, subjects administer an electric shock to their opponent; following a losing trial, they receive a shock. In actuality, no opponent exists and the experimenter administers the shocks to the subjects according to a predetermined win/lose order. Aggression is operationalized as the average shock intensity subjects select for their opponent during an experiment. Some modified versions of the TAP also include shock duration as a measure of aggression. In the PSAP, subjects are led to believe that they are competing against another person on a task in which they can earn points that are later redeemable for money but that they also may have points taken away by their opponent. In most versions of this paradigm, subjects can either press a particular button 100 times to earn a point or press a different button 10 times to subtract a point from their opponent. Aggression is measured as the number of times the point-subtraction button is pressed. (For a full description and evaluation of these paradigms see Giancola and Chermack 1998.)

Some have argued that responses in these laboratory paradigms do not generalize to aggression in the "real world." An abundance of data shows this criticism to be wholly unfounded. The validity of these procedures has been established in numerous ways. For example, shock-level selection on the TAP correlates significantly with self-report measures of physical assault, behavioral hostility, and outwardly directed anger. That aggression rather than a related disposition or behavior is being measured is shown in the lack of correlation between shock selection and measures of guilt, suspicion, resentment, inwardly directed anger, helping, and competition. Additional data supporting the validity of these paradigms come from studies showing that adolescents whose teachers rate them high on aggressiveness are more aggressive in a version of the TAP than adolescents with low ratings. Violent offenders respond more aggressively on the PSAP than do nonviolent controls. A recent comprehensive review of studies that have used laboratory measures of aggression concluded that the TAP, the PSAP, and their modified versions are safe, effective, and valid measures of aggression for both men and women (Giancola and Chermack, 1998).

Alcohol and Aggression | A wealth of studies, in numerous laboratories in North America and in Europe, have examined the alcohol-aggression relation using the paradigms described above. This work has documented a very robust and reliable finding: Persons who are given an alcohol beverage exhibit more aggressive behavior than those who receive a nonalcohol or a placebo beverage (for review see Chermack and Giancola 1997). Placebo groups are used to rule out the possibility that it is not the pharmacological properties of alcohol that facilitate aggression but the mere belief that alcohol has been consumed. Placebo manipulations involve giving a nonalcohol beverage and then informing the drinker in a convincing way (and the credibility of the ruse can be tested) that she or he has ingested alcohol. Then, if subjects who received alcohol are more aggressive than those who received a placebo, and if those who received a placebo are no more aggressive than those who received no alcohol (and were told that they received no alcohol), it can be concluded that the belief that alcohol has been consumed plays little role in the expression of aggression.

Sketching a "Risk Profile" | In summary then, experimental data show that acute alcohol consumption, and not the belief that alcohol has been consumed, significantly increases the probability of aggressive behavior. It is not the case, however, that alcohol invariably causes aggression. Not all persons who commit an aggressive act are intoxicated, and alcohol does not lead to aggression in all persons who ingest it. In other words, alcohol is neither a necessary nor a sufficient agent in the elicitation of aggressive behavior. Rather, alcohol-related aggression is the product of individual characteristics and contextual variables interacting with alcohol pharmacodynamics.

Executive Cognitive Functioning (ECF) | ECF is a higher-order cognitive construct that covers planning, initiation, and regulation of goal-directed behavior. Cognitive abilities included in this construct are control of attention, strategic goal planning, abstract reasoning, cognitive flexibility, hypothesis generation, temporal response sequencing, and the ability to organize and adaptively utilize information contained in working memory. The prefrontal cortex and its subcortical circuits are thought to be the neurological substrates that subserve ECF. People with psychiatric disorders that sometimes entail aggressive behavior, such as antisocial personality disorder, psychopathy, conduct disorder, and attention-deficit/hyperactivity disorder, all show poorer performance on neuropsychological measures of ECF. Low ECF capacity has been linked to increased aggression in laboratory tests in preadolescents and young adult males, increased disruptive, delinquent, and physically aggressive behavior in adolescent females, and increased mother and teacher reports of aggression and conduct problems. It has been hypothesized that low ECF facilitates the expression of aggression by decreasing behavioral inhibition and interfering with the ability to generate alternative, nonaggressive responses in provocative situations. It is well known that acute alcohol consumption detrimentally affects cognitive functioning in general. However, its predominant disruptive effects are on ECF. Neuroimaging studies have corroborated this finding by demonstrating that acute alcohol consumption reduces glucose metabolism predominantly in the prefrontal cortex. Therefore, it can be hypothesized that acute alcohol consumption disrupts ECF, which then facilitates aggression. Recall, however, that alcohol does not facilitate aggression in all people. It is thus reasonable to suggest that the proposed process is more likely to occur in those who already have low ECF.

Alcohol Expectancies | Alcohol expectancies are beliefs about the effects of alcohol on behavior (different from placebo effects). Some research suggests that intoxicated aggression results, in part, from the belief that alcohol increases aggression. It is well known that people vary in their belief that alcohol increases arousal, power, assertiveness, verbal aggression, and physical aggression. Significantly, self-report studies indicate that the association between alcohol consumption and aggression is stronger among individuals who expect alcohol to increase aggression. Only one published laboratory study, using the TAP, related the effects of alcohol on subjects' aggression to their beliefs about the effects of alcohol on aggression. Results indicated that under conditions of high provocation, intoxicated subjects with high expectancies about the effects of alcohol on aggression were more aggressive.

Dispositional Aggressivity | Dispositional aggressivity—a person's typical level of aggressiveness across a range of situations—is strongly related to self-reported husband-to-wife violence and violent behavior in college males. In addition, as noted earlier, dispositionally aggressive individuals, such as those with antisocial personality or conduct disorder, show low levels of ECF. Only one study has assessed the combined effects of acute alcohol consumption and dispositional aggressivity on aggression as measured by the TAP. Acute alcohol consumption increased aggression in men with high levels of dispositional aggressivity but not in those with low or moderate levels.

Drinking History | Quantity of past alcohol consumption is positively related to self-reported aggression in both male and female social drinkers. Theory suggests that increased alcohol consumption and aggressive behavior are both components of an overarching construct of "deviant behavior." However, the underlying mechanisms, or causal dynamics, of that construct are not known. Moreover, one laboratory study found that acute alcohol consumption increased aggression on the TAP only in males with low rather than moderate or high levels of past-year drinking. The authors hypothesized that alcohol's detrimental effects on cognition were greater in those with a low tolerance for alcohol compared with those with a higher tolerance.

Biochemistry | Both animal and human research have demonstrated a positive relation between testosterone levels and physical aggression. A recent study found that healthy young males with high levels of testosterone, measured from saliva, were more aggressive on the TAP than those with low levels. Another study reported an increase in aggressive responding on the PSAP subsequent to the administration of testosterone cypionate to anabolic steroid users. Heightened impulsive aggressive behavior has also been related to a low level of serotonin, a neurotransmitter in the brain. Interestingly, recent studies using the TAP and the PSAP have demonstrated increased aggression in healthy males who received a tryptophan-depleted dietary mixture. Tryptophan is the biochemical precursor for serotonin; its dietary depletion leads to lowered brain serotonin levels.

Theorists have argued that serotonin is involved, in part, in the inhibition of behavior. As such, it may be that alcohol affects the serotonin system in such a manner as to impair its ability to properly inhibit behavior, including aggression. Pertinently, there is some evidence to show that low serotonin may affect aggression by interfering with ECF. Acute alcohol consumption leads to an initial increase, and then decrease, of serotonin, which may promote aggression. Others have argued that the serotonin-depleting effect of alcohol may be more profound in some people. However, these explanations are highly speculative because laboratory studies show that alcohol's aggression-inducing effects occur quite quickly, that is, when serotonin levels are supposedly still rising. In fact, a recent study indicated that a given blood alcohol level is more likely to cause aggression when the blood alcohol concentration is rising (shortly after consuming a few drinks) as opposed to when it is falling (during the detoxification process). Based on this, it is clear that a great deal of research is still needed to elucidate the complex interactions of alcohol and serotonin in the expression of aggressive behavior. In addition, it has been suggested that alcohol may also facilitate aggression through its effects on brain receptors for another neurotransmitter, gamma-aminobutyric acid (GABA). It is believed that alcohol acts at the same receptor complex as sedative drugs, such as the benzodiazepines (e.g., Valium, Xanax). Laboratory studies and clinical reports indicate that these "sedatives" actually tend to facilitate aggressive reactions by reducing anxiety, including fear of punishment. GABA functions, in part, to suppress fearful responses to punishment and fear-related cues. Alcohol increases chloride transmission at the GABA receptor complex, which augments GABA neurotransmission and thus may facilitate aggression by reducing fear.

Alcohol, the Person, and the Situation | I have neglected to discuss social and contextual factors, although they are clearly important determinants of whether alcohol will, or will not, facilitate aggression. Even a high-risk profile for alcohol-induced aggression will generate aggression only in certain contexts. However, because more research has been conducted on social and situational influences than on person or trait variables, I have stressed the development of risk profiles for alcohol-related aggression that focus on the person. To summarize, people with low ECF, strong beliefs that alcohol increases aggression, high dispositional aggressivity, a heavy (or possibly light) drinking history, high testosterone, or low serotonin levels, or any combination of these traits, will be more likely to exhibit aggression under the influence of alcohol than those without these vulnerabilities.

It should be made clear that the risk factors described here do not comprise an exhaustive list. Other traits that are potentially important include age, perspective-taking, self-awareness, negative affect, temperament, emotionality, sensation seeking, anxiety, irritability, hostility, frustration tolerance, impulsivity, psychopathology, early physical abuse, perceived self-esteem, tolerance and sensitivity to alcohol, and physiological reactivity to stress. There is no single risk profile for intoxicated aggression. Identifying key vulnerabilities is important because it will provide researchers with a better understanding of the causal structure of the alcohol-aggression relation. With such an understanding, scientists will be in a better position to develop effective prevention and treatment interventions.


Cherek, D. 1981. Effects of smoking different doses of nicotine on human aggressive behavior. Psychopharmacology 75: 339-345.

Chermack, S., and P. Giancola. 1997. The relationship between alcohol and aggression: An integrated biopsychosocial approach. Clinical Psychology Review 6: 621-649.

Giancola, P., and S. Chermack. 1998. Construct validity of laboratory aggression paradigms: A response to Tedeschi and Quigley (1996). Aggression and Violent Behavior 3: 237-253.

Pernanen, K. 1991. Alcohol in Human Violence. New York: Guilford Press.

Roizen, J. 1993. Issues in the epidemiology of alcohol and violence. In S. Martin (ed.) Alcohol and Interpersonal Violence: Fostering Multidisciplinary Perspectives (NIAAA research monograph No. 24, NIH Publication No. 93-3496, pp. 3-36). Rockville, MD: U. S. Department of Health and Human Services.

Taylor, S. 1967. Aggressive behavior and physiological arousal as a function of provocation and the tendency to inhibit aggression. Journal of Personality 35: 297-310.

| back |