Laboratory Studies of Human Aggression and Impulsivity: Relationship to Treatment of Substance Abuse

Aggression in particular and impulsivity put people at risk for substance abuse. If you can modify these, you may reduce the risk for substance abuse and other anti-social behaviors. Predictors of drug abuse include the aggressive child at the highest risk for substance abuse, subsequent criminal behavior, and violence. When we bring high-risk adolescents into the lab and measure impulsivity and aggression, the differences are in aggression and not impulsivity.

Aggression precedes drug use. Drug use in these populations can escalate violence. Drug use doesn’t make Dr. Jekell/Mr. Hyde; it makes Mr. Hyde a lot worse.

Three laboratory methods have been used to measure aggression. The first procedure by Buss has gone into disuse. All of these measures have used an aversive stimulant to measure aggression. Buss and Taylor used electric shock. Our study is called the Point Subtraction Aggression Paradigm (PSAP). What we did was change the procedure quite a bit and used subtraction for a negative action that people chose as a response, thus reducing the potential to earn more funds. People are brought in to respond to three buttons on a counter that shows the money being accumulated. Occasionally the counter blinks or goes down instead of up and the subjects are told that a fictitious person is taking the money. The subject can retaliate by pushing one of the buttons, or protect the earnings and do nothing to the other person by pushing the designated button. If no action is taken, we call it an escape response. If they retaliate and push the button to take money away, we call that the aggressive response.

We looked at matched controls: parolees who either have a history of childhood conduct disorder or those who do not. The history of substance abuse was extremely high.

The external validity question was "do subjects with a history of violence display more frequent aggressive responses in the laboratory?" Based on criminal history, the subjects were assigned to a violent or non-violent group. Male parolees with a history of violence were three or four times more aggressive. When males and females were studied, violent females demonstrated higher violence than the non-violent males and females, but were not as violent as the violent males.

The reliability question was whether other laboratories using these procedures would achieve similar levels of experimental control. There have been a number of laboratories that have used this, and they get similar kinds of data. There is an article in a recent issue of Drug and Alcohol Dependence describing this procedure. In one study of adults with a history of childhood conduct disorder, we provide eight sessions in a day. The study was published in 1999 in

Psychopharmacology. We looked at three doses of D, 1-FEN which releases serotonin and dopamine. We found it decreased aggressive

responding, it decreased impulsive responding, and it produced a slight stimulation. The monetary responding function is a good control for motor activity. The acute release of serotonin and dopamine produced substantial decreases in all subjects.

The second study of conduct disorder, which is in press and not yet published, was with paroxetine. Subjects were divided into two groups: one to receive paroxetine and one to receive a placebo throughout. There was a week of baseline data, two weeks of placebo, three weeks of paroxetine (20mg daily), and then two weeks of placebo. Compliance was checked with MEMS bottles. However, more importantly, we obtained plasma levels of paroxetine. We validated to see if people were taking the doses. The significant change we found only occurred after three weeks of paroxetine and the change was on impulsivity rather than on aggression. Aggression is decreased by 5-HT/DA released (d,1-Fen) and less so by chronic 5-HT reuptake inhibition (paroxetine). Impulsivity was decreased by d,1-Fen and by paroxetine treatment.

Recent studies using d-fenfluramine and baclofen were also discussed. The baclofen study is attempting to manipulate various aspects of the GABA system. While a person may come in three days a week, placebos would be administered two of the three days. Aggression was decreased by release of 5-HT and GABA-B agonist in conduct disorder groups at the same time the non-conduct disorder group is increasing its aggression. The different response is due to the baseline level of serotonergic functioning in that particular group. Pharmacotherapy aimed at reducing these behaviors may reduce risk for substance abuse and other antisocial behaviors. Such interventions would likely be most effective if initiated prior to the development of substance abuse.

1. Cherek DR, Lane SD. Effects of d,1-fenfluramine on aggressive and impulsive responding in adults with a history of conduct disorder. Psychopharmacology. 1999;146:473-481.
2. Cherek DR, Lane SD. Fenfluramine effects on impulsivity in a sample of adults with and without a history of conduct disorder. Psychopharmacology. 2000;152: 149-152.
3. Kellam S, et al. Attentional problems in first -grade and shy and aggressive behaviors as antecedents to later heavy or inhibited substance abuse. NIDA Research Monograph. 1989;95:368-9.

This presentation addresses special population studies that show a relationship between substance use and aggression. Etiologic considerations addressing neurobiology as one of the richest areas describing this relationship and treatment data are examined.

Emergency room studies show that individuals are more likely to have been involved with drug and alcohol use. In one study on violent injury, 37% occurred in a bar/restaurant with 42% of the individuals with blood/alcohol level (BAL) of >.80 with significantly higher of urine drug screens (UDS) than other individuals presenting at the emergency room (ER) that same night. In a study on life-threatening assaults, 84% were alcohol positive and 19% were cocaine positive. Another source of data on this relationship is 1998 Department of Justice data on positive urine drug screens in violent crime arrests. Even this data may be underreported as the window for detection is narrow, so that by the time the urines were collected, many of the individuals in the study may no longer have been positive.

There is also a body of evidence around substance abuse and suicide. The literature suggests that suicide is a form of aggression and violence; substance abuse increases the risk for suicide. In one study, alcohol use was associated with 36% of people hospitalized with a suicide attempt; another study found a positive correlation between the severity of the attempt and alcohol use. Another study showed the use of alcohol within two hours of the attempt by 46% of the cases.

A case I saw this year was Mr. A, 24 years old, a married white male who presented for the treatment of alcoholism. He wasn’t there of his own accord. His wife was going to leave him if he didn’t come in for treatment. He had been recently arrested in a barroom brawl, had lost two jobs because of alcohol and fighting, and lost his driver’s license because of several DUIs. He had been violent with his wife while intoxicated and was apologetic about it afterwards. His parenting style was inconsistent but not violent with the children. The family of origin was alcoholic.

In the model of Intergenerational Transmission of Family Violence and Substance Abuse, you have to pick a place where you might want to intervene. Helping people with their parenting skills, both men and women, is a good place to start. The biopsychosocial interaction results because of the effect of the drugs themselves. Alcohol can be disinhibiting. Cocaine certainly increases people’s level of activity in general; it can make people paranoid. These behaviors can lead to aggressive actions.

Impaired executive cognitive functioning is found in adolescent males/females with substance abuse disorders and aggressivity. There is a neurobiology that may be common to both aggression and substance use disorders. There is a serotonin deficiency hypothesis of alcohol-induced aggression. Studies show low CSF 5HIAA in aggressive and/or alcoholic individuals. Alcohol can decrease serotonin synthesis. Susceptible individuals have enhanced serotonin depletion as a result of alcohol use and that is probably the pathway to aggression.

In one study, in non-alcoholic individuals it did not matter if they were tryptophan-depleted or tryptophan-supplemented, there was a certain amount of shock they were willing to deliver, and no higher level, to an unseen individual in another room. But the combination of alcoholism with tryptophane-depletion seemed to be problematic in increasing this one measure of aggression.

In conducting the case history of Mr. A, we found a family history of depression and possible bi-polar disorder with a possible suicide by a grandfather. There was much alcoholism in the family. He reported feeling sad, frustrated, and irritable but really didn’t have anything that sounded like bi-polar disorder with the exception of angry and irritable mania rather than euphoric mania. The alcoholism makes it hard to differentiate what is happening. He was a sad individual, occasionally suicidal with poor problem solving and abstract thinking skills. His deficits in executive cognitive thinking were classic.

Common comorbidities in individuals who are aggressive substance abusers include bipolar disorder, PTSD, GAD/panic, and depression. There is an intriguing body of evidence around anticonvulsant mood stabilizer agents. Mostly this is case report literature; it is found in the literature for autism, schizophrenia, and mental retardation. Anticonvulsant mood stabilizing agents seem to be able to decrease aggressive outbursts and irritability. In one study, divalproex (DVPX)-treated individuals had improvement in terms of their aggressive/impulsive behavior when compared to the placebo group. In another 12-week double-blind, placebo-controlled study, we did not have a robust effect from DVPX. Both groups reduced their consumption of alcohol markedly so it was difficult to see a difference between the two. The DVPX group did experience less relapse of heavy drinking. We found differential effects of DVPX on irritability, lability and verbal assault—individuals on DVPX improved more than those on placebo.

Mr. A’s treatment plan was detoxification with a mood stabilizing anticonvulsant (valproate), and he began fluoxetine and increased to 40mg per day. He showed a marked improvement in mood and decrease in irritability. He was referred for intensive outpatient substance abuse treatment with special attention on anger management, marital therapy, parenting skills, and neuropsychological rehabilitation—interventions to lengthen the fuse from impulse to action.

Helping alcoholics become better parents is one way to interrupt the cycle of violence.

References:

1. Akiskal HS. The prevalent clinical spectrum of bipolar disorders: beyond DSM-IV. J Clin Psychopharmacol. 1996;6(2 Suppl 1), 4S-14S.
2. Brady KT, Pearlstein T, Asnis G, Baker D, Rothbaum B, Sikes C, Farfel G. Double-blind,placebo-controlled study of the efficacy and safety of sertraline treatment of postraumatic stress disorder. 1998.
3. Brady KT, Dansky BS, Saladin MS, Sonne SC. PTSD and cocaine dependence: Order of onset. Am J Addictions. 1998;(7):128-135.
4. Brady KT, Myrick H, McElroy S. The relationship between substance use disorders, impulse control disorders, and pathological aggression. Am J Addictions. 1998;7):221-230.
5. Cassano GB, Dell'Osso L, Frank E, Miniati M, Fagiolini A, Shear K, Pini S, Maser J. The bipolar spectrum: a clinical reality in search of diagnostic criteria and an assessment methodology. J Affect Disord. 1999;54(3):319-28.
6. McElroy SL, Keck PE, Stanton SP, Tugrul KC, Bennett JA, Strakowski, SM. A randomized comparison of divalproex oral loading versus haloperidol in the initial treatment of acute psychotic mania. J Clin Psychiatry. 1996;57(4):142-6.

Culture has become important again as we look at the changing color of America. The country is becoming more diverse and is about 49% non-Hispanic Caucasian with 51% other groups. The two fastest growing groups are Latinos and Asians. Many people are seeing themselves as multi-ethnic, that is two or three identities as indicated in the last census. It is well recognized that substances are used in different ways by different groups of people. African-American youth have low use of alcohol. However there is the perceived issue of the African-American use of crack versus powder cocaine. Crack is more common among lower income individuals, Latinos, and African-Americans. Designer drugs are heavily common in Caucasians. A recent study shows there is differential use on college campuses, which suggest that some of the drug use patterns are not simply related to income differences. Dr. Deas shows that on college campuses, African-Americans have heavy use of marijuana but less use of alcohol compared to Caucasians.

Multigenerational income is not controlled for in most studies. We forget that while Latinos and African-Americans’ income has increased, it is income gains of the last one or two generations. Family wealth may be only 1/10^th of the family wealth of Caucasians although median current family income may be 60-70% of the national average. The costs of direct services for substance abuse is often not calculated, but we presume that effective substance abuse treatment is out of the income range of these groups because they may not have the financial reserves necessary to have access to treatment.

One important issue in studies about African-Americans is that homicide rates are extremely high in terms of perpetrators and victims. The role of crack cocaine accounts for about 20% of the homicide cases. Turf battles are certainly a factor. When you control for income, the differences disappear. Latinos also show high rates, but their rates of female victimization are much lower than other groups.

Some cocaine users may have impaired cognitive functioning, especially the executive function that may be persisting. Impurities may make the difference between the toxic effects of crack cocaine versus powder; however this area has not been researched. The result is impaired judgment and/or impulsivity which may contribute to high rates of violence.

In terms of suicide, culturally it has been rare among many minorities. It has only been recently that good, reliable data has been available that shows that suicide rates increasing with age is primarily seen in Asians and non-Hispanic Caucasians. Most groups show a peak among young adolescents, but not all groups show an age increase. Native Americans, African-Americans, and Latinos do not show increasing rates in the elderly. Additional studies are needed to look at the relationship between substance abuse and suicide.

There are two books that review the literature on ethnicity and psychopharmacology, but have little discussion on substance abuse. There is an under-recognition of mental disorders among many ethnic minorities. This may lead to an increased risk for substance abuse and violence due to the lack of treatment.

For many psychotropic medications, in Asians and African-Americans there are fewer rapid metabolizers, fewer poor metabolizers, and more slow metabolizers. The common doses given of common psychotropic medications may result in more side effects. These side effects may include akathisia, which can lead to substance abuse to relieve symptoms and to increasingly violent behavior.

Bipolar illness is often under recognized in African-Americans. Substance abuse impacts the course of the disorder. We found that age of onset is greatly affected by substance abuse. Alcohol-only cases had early onset and individuals with alcohol and drug abuse problems had even earlier ages of onset. Alcoholism and substance abuse also affect outcomes.

In contrast to the general perception that affective disorders were non-existent in certain minority populations, the 1993 Epidemiological Catchment Area Study and the 1994 National Comorbidity Study found that affective disorders were prevalent in African-Americans. These disorders are either under treated or treated inappropriately. Lithium and other antimanic agents are delayed or not given because of misdiagnosis, under diagnosis or subtypes other than Bipolar I are undiagnosed. Bipolar I may be more common in African-Americans than previously thought.

African-Americans are frequently perceived as being more hostile and this reticence to cooperate is often based on past experience in terms of interracial encounters. Sometimes it is unfamiliarity with the idioms involved. They are more willing to report intrapsychic experiences, have a protective wariness or healthy paranoia of the system, and have experienced misinterpretation of culturally-bound experiences.

In one inpatient study with a group with similar diagnosis and on a fixed medication dose, we found consistently that African-American patients showed less objective evidence of violence, but that the staff consistently reported the African Americans as being more violent.

Studies have found Latinos and African-Americans receive differential treatment. They are given higher doses of psychotropic medication, more different medications, more injections of medications, and they are more likely to receive depot medication and less likely to receive antidepressants or psychotherapy.

There are some differences in availability of treatment for African-Americans, Latinos and Asians than the rest of the population. There is less likelihood of inpatient care. African-Americans and Latinos are much likely to be referred to the correctional system when they have the same symptoms as non-Hispanic Caucasians. They are more likely to be involuntarily committed and more likely to leave against medical advice. African-Americans, in particular, are less likely to get specifically identified psychotherapy, i.e., cognitive behavioral therapy.

Post-traumatic stress disorder has been found to be more common among African-American combat veterans. It is extremely common among inner-city children and adolescents. This latter group has not been studied very well but has important implications. It is often misdiagnosed; flashbacks may be confused with hallucinations. Emotional blunting is confused with flattened affect, antisocial coldness; avoidant behavior can be confused with paranoia. Autonomic hyperreactivity can be confused with rage.

Substance abuse as self-medication for non-recognized mental disorders is an area that needs additional exploration. There is a subset of the population that has underlying treatable mental disorders especially among underrepresented ethnic groups who are not being treated. These individuals may be at increased risk for violent behavior as well as substance abuse.

References:

1. Chung H; Mahler JC; Kakuna.T. Racial differences in the treatment of psychiatric inpatients. Psychiat Services. 1995;46: 586-591.
2. Flaherty JA; Meagher R. Measuring racial bias in inpatient treatment. Am J Psychiatry. 1980;137: 679-682.
3. Lawson WB. Clinical issues in the pharmacotherapy of African-Americans. Psychopharmacology Bull. 1986;2: 275-281.
4. Strakowski SM; Shelton RC; Kolbrener ML. The effects of race and comorbidity on clinical diagnosis in patients and psychosis. J. Clin Psychiatry. 1993;54:96-102.
5. Price N; Glazer W; Morgenstern H. Demographic predictors of the use of injectable versus oral antipsychotic medications in outpatients . Am J Psychiatry. 1985;142:1491-1492.
6. Segal SP; Bola J, Watson M. Race, quality of care, and antipsychotic prescribing practices in psychiatric emergency services. Psychiatric Services. 1996;47:282-286.