Antisocial Behavior Implications for Diagnosis Robert Linda
K. Brooner, Ph.D., Chester J. Feich, M.A., and George
Objective: A pattern diagnosis ofantisocial the
adult
examined
criteria
ofchronic personality
are
met.
in intravenous
aggressive, volunteered
of Intravenous of Antisocial
adult antisocial disorder unless
The drug
utility
of the
abusers,
and criminal behavior. for an outpatient study
Drug Abusers: Personality Disorder
W. Schmidt, E. Bigelow,
behavior is not the early-onset
early-onset
a population
sufficient criteria
requirement
known
Method: The ofpsychopathology
M.D., Ph.D.
subjects
to have were and HIV
for the DSM-III-R are satisfied, even for
high
the
rates
diagnosis
if
was
of irresponsible,
23 7 drug abusers risk behavior and
who had infection.
They completed a structured psychiatric interview as part of their participation in that study. The adult antisocial behavior ofthe group that met both the early-onset and the adult criteria for antisocial personality disorder, the group that met only the adult criteria, and the group that met neither the early-onset nor the adult criteria was then compared. Results: Antisocial personality
disorder
sample; social
personality
social behavior all three groups. of drug behavior. possibility (Am
(meeting
an additional
abusers The
J
24%
the met
disorder
early-onset only
reported
than did the Conclusions:
adult
a more
other groups, Early onset
with important differences antisocial behavior of the
ofa late-onset Psychiatry 1992;
and
the
and less severe 149:482-487)
form
An earlier version of this paper was presented at the 143rd annual meeting of the American Psychiatric Association, New York, May 12-17, 1990. Received April 20, 1990; revisions received Dec. 5,
Medical Center. Address reprint requests to Dr. Brooner, ofPsychiatry,Johns Hopkins/Francis Scott Key Medical Eastern Ave., Baltimore, MD 21224.
Supported
by research
grants
DA-05127
National Institute on Drug Abuse. Copyright © 1992 American Psychiatric
482
and
Association.
The Johns Scott Key
Department Center, 4940
DA-000S0
from
and
was
group more
antisocial antisocial
diagnosed with
serious
the
in 44% diagnosis
pattern
ofadult
of the of antianti-
behavior was commonplace in behaviors identified a subset
in the extent and severity of their adult antisocial group that met only the adult criteria suggests the
ntisocial behavior is strongly linked to drug abuse (1-3). Ball et at. (3) interviewed 354 drug abusers regarding their day-to-day criminal activities. All subjects reported a history of adult antisocial behavior, including the illegal sale of drugs, illegal gambling, theft, and other criminal offenses. Despite the fact that adult antisocial behavior is commonplace among drug abusens, previous research indicates that most of these persons fail to satisfy the diagnostic criteria for antisocial personality disorder. For example, the prevalence of this disorder in drug abusers when the DSM-III (or DSM-III-R) definition is used has ranged from 35% (4)
21, 1991. From and the Francis
criteria) The
pervasive although of multiple
A
1990, and Sept. 10, 1991; accepted Oct. Hopkins University School of Medicine
adult
criteria.
the
ofantisocial
personality
disorder.
to a high of 54% (5), with an average prevalence rate of about 40% (6, 7). What accounts for the fact that most drug abusers fail to meet the diagnostic criteria for antisocial personality disorder? One potential explanation may be related to how DSM-III and DSM-III-R openationalize the disorden. Each of these systems uses definitions with a strong behavioral framework that require early onset (before the age of 15) of multiple antisocial behaviors for a patient to become eligible for the adult diagnosis. Thus, to be classified as having antisocial personality disorder, a person must have fulfilled at least three of 12 specific antisocial criteria (e.g., truancy, chronic lying, aggressiveness, and criminality) prior to the age of IS and fulfill at least four of the adult criteria (e.g., unstable work history, chronic lying, aggressiveness, criminality). As a result of the requirement for early onset, persons with high rates of adult antisocial behavior (including those who may satisfy the adult criteria for antisocial personality disorder) would not be diagnosed as having the disorder unless they also meet the early-onset criteria. This aspect of the current definition of the disorder has been criticized by Hare (8) and was discussed recently in a review by Gerstley et al. (6). These criticisms
Am
J
Psychiatry
1 49:4,
April
1992
BROONER,
raise important the early-onset cial personality who typically
questions about the clinical utility of requirement for the diagnosis of antisodisorder, particularly for drug abusers, report high rates of adult criminality, aggressiveness, irresponsibility, and unemployment. To examine this issue further, we compared the rates of specific adult antisocial behaviors among intravenous drug abusers to determine whether there were meaningful differences between those who satisfied both the earlyonset and adult criteria for antisocial personality disorden, those who satisfied the adult criteria but not the early-onset criteria, and those who satisfied neither the adult non the early-onset requirements for the disorder.
METHOD The 237 subjects presented in this report were intravenous drug abusers participating in a longitudinal study of psychopathology and HIV high-risk drug use and sexual behavior. The sample included subjects who were in a hospital-based methadone treatment program (N=128) and subjects who were not in methadone treatment at the time of enrollment in the study (community sample, N=109). All subjects who were in treatment met the DSM-III-R criteria for current opioid dependence, and many also had at least one additional substance use disorder, primarily related to cocaine use (68%, N=87). These subjects had been in methadone treatment for at least 3 months and had been recruited by advertisements posted in the treatment clinic. Community subjects were eligible for inclusion in the study if they reported intravenous drug use within the 3 months before study enrollment, had a lifetime history of intravenous drug abuse for more than 3 continuous months, and had less than 12 months of cumulative methadone treatment in the 4 years before entering the study. Nearly all of the community subjects satisfied current diagnostic criteria for an opioid (N=89) and cocaine (N=104) use disorder. They were recruited by advertisements in selected health and research agencies and by chain-referral techniques. Each subject was paid $10 per hour for completing the intake assessment instruments, for a total of approximately 6 hours ($60). Informed consent was obtained from each subject after study procedures were fully described. Given the sensitive nature of these data, a confidentiality certificate was obtained for the project from the National Institute on Drug Abuse, which would protect the data from subpoena. The mean age of the subjects was 36.0 years (SD=6.6, nange=23-63); 70% (N=16S) were male. Forty percent (N=95) of the subjects were white, 59% (N=140) were black, and less than 1 % (N=2) were Hispanic or Native American. The mean educational level was 1 1 .0 years (SD=1 .9, range=5-1 9); 67% (N=1 59) were unemployed. Forty-six percent (N=109) had never married, 14% (N=33) were married, 35% (N=83) were separated/divorced, and 5% (N=12) were widowed. They reported onset of recurring life problems associated
AmJ
Psychiatry
149:4,
April
1992
SCHMIDT,
FELCH,
ET AL.
with any substance use at a mean age of 1 8.8 years (SD=4.9, range=1 1-45) and an average of 33.2 months (SD=S0.0, range=0-240) of cumulative (lifetime) methadone maintenance. Subjects who were in treatment were more likely to be white (66%, N=84, versus 9%, N=10; X2=76-OS, df=1, pA,N
44.6
6
10.0
B>A>N
Before age 15 (early onset) Truancy Suspensions Incorrigibility Running away Chronic tying Multiple sex partners Destruction of property Settingfires Use of weapons Instigating fights Arrests Other illegal acts
Reform
Other misdemeanors Felony arrests Other felonies aSubjects bSubjects CSubjects
meeting meeting meeting
both the early-onset and the adult criteria for antisocial only the adult criteria for antisocial personality disorder. neither the early-onset nor the adult criteria for antisocial
sets of criteria were compared by analysis of variance or chi-square tests. Analyses of the percentages of subjects who had engaged in the various childhood and adult antisocial behaviors were done by means of chisquare tests (with continuity correction) with post hoc testing of group differences by the Tukey-Kramer method. The frequencies of specific childhood and adult antisocial behaviors were compared for the three groups by means of analysis of covaniance and Tukey’s post hoc test; age at onset of recurring life problems associated with drug use was used as a covaniate because groups were significantly different on this vanable. While it would have also been appropriate to indude gender as a grouping factor in this analysis, the resulting cell sizes would have been too small for meaningful interpretation of the results. All analyses were conducted with the SPSS (12), and results were considened significant if the p value was less than 0.05.
RESULTS Of the 237 subjects, 44% (N=104) were classified as having antisocial personality disorder, since they met both the early-onset and the adult criteria (Both), 24% (N=56) met only the adult criteria (Adult Only), and 25% (N=60) satisfied neither the early-onset nor the adult criteria (Neither). An additional 7% (N=17) met the early-onset requirement but failed to meet the adult
484
9
personality personality
36 30 21 4
B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A,N B>A>N
B>A,N
B,A>N B,A>N B>A>N B,A>N
6.7
B>A>N
disorder. disorder.
criteria for the disorder. This last group was too small to permit meaningful statistical comparisons and was omitted from the analyses. Thus, all statistical companisons were made between subjects in the three remaining groups (N=220). More of the subjects in the Both and the Adult Only groups than in the Neither group were male (79.8%, 71.4%, and 51.7%, respectively; x2=l4.42 df=2, p= 0.001), and more of the subjects in the Neither group than in the Both and the Adult Only groups were marned (21 .7%, 9.6 %, and 5.4%, respectively; 1, df=2, p=O.O2). Those in the Both group also reported a significantly earlier age at onset of recurring life problems associated with substance use than those in the Adult Only and the Neither groups (17.4, 20.1, and 20.7 years, respectively; F=10.99, df=2, 215, pA,N B>A,N B>A,N B>A,N B>A,N
0.07
0.26
0.02
0.13
B>A,N
0.23 0.80 2.80 0.51 21.04
0.00 0.00 0.12 0.07 2.12
0.00 0.00 0.78 0.25 13.25
B>A,N B>A,N B>A,N B>A,N B>A,N
7.17 7.65 0.20 0.50
5.59 4.05 0.90 0.91
B>N BN B>N
Mean
ET AL.
Drug Abusers Di vided Into Three Groups
Neither’ (N=60) (N)
Only” (A)
FELCH,
Before
age IS (early onset) Truancy Suspensions Incorrigibility Runningaway Multiple sex partners Destruction of property
4.49 3.19 5.44 1.28 2.55 3.73
5.20 4.97 10.64 2.65 5.22 8.92
Reform
0.51
0.93
0.58 2.38 4.38 1.40 24.03
1.62 6.56 9.52 2.98 45.17
school
Setting fires Use of weapons Instigating fights Arrests Other illegal acts
After
1.21
0.55
0.05 0.1 1 0.52 0.18 4.98
age IS (adult)
Different
Years
employers
Use of weapons Misdemeanor
Other
arrests
,
misdemeanors
Felony
arrests
Other
felonies
aSubjects
10.18
of monogamy
meeting
both
the the
bSubjects
meeting
only
‘Subjects
meeting
neither
early-onset adult
criteria
the early-onset
4.39
8.43
3.46
6.45
4.45 3.61
6.43 7.59
3.27 2.84
8.27 5.09 8.70 6.21
50.23
89.19
42.69
92.87
10.10
26.88
B>N
3.73
4.37
1.84
3.26
0.57
1.08
B>A,N
1.60
7.32
B>N
40.51 and
92.12
the
for nor
10.13
adult
antisocial the
adult
criteria
15.11 for
personality criteria
antisocial
J
Psychiatry
1 49:4,
April
1992
disorder.
disorder. for antisocial
interview. As expected, significantly more of the subjects in the Both group than in the Adult Only and Neithen groups had been involved in each of the childhood behaviors, including having been placed in reform school. There were no significant differences between the percentages of subjects in the Adult Only and the Neither groups who had been involved in these early life behaviors, with the exception that more of the Adult Only group had engaged in “other illegal acts.” On measures of adult antisocial behavior, there were also consistent differences between the Both and the Neither groups (table 1 ). With the exception of current unemployment, a greaten percentage of subjects in the Both group had been involved in each of the adult behaviors. In contrast, comparisons between subjects in the Both and the Adult Only groups yielded mixed resuits. Similar percentages of subjects in the Both and the Adult Only groups reported employment problems, difficulty maintaining a monogamous relationship, homelessness, and misdemeanor arrests. However, compared to the Both group, significantly fewer subjects in the Adult Only group had been arrested for a felony, committed a felony not resulting in arrest, committed a misdemeanor not resulting in arrest, used weapons in fights, on developed a pattern of chronic lying. It is interesting that significantly more subjects in the Adult Only group than in the Neither group had been involved in each of the adult behaviors, with the exception of felony arrest and of unemployment at the time of study entry. In short, subjects in the Adult Only group appeared to be behaviorally different from and intermediate between those in the Both and the Neither groups.
Am
42.44 personality
personality
disorder.
Table 2 shows the comparison of the three groups with respect to the number of times subjects reported committing specific antisocial acts. These comparisons were restricted to those questions which provided data on the number of times a specific behavior occurred (e.g., number of times arrested, number of times weapons were used in fights). As would be expected, the group with antisocial personality disorder (Both) reported significantly higher frequencies on each of the measures of childhood antisocial behavior than did the Adult Only and the Neither groups. Importantly, the Adult Only and Neither groups did not differ significantly on the frequency of any of these early life behaviors. The relationships were notably different with respect to the adult behavior of these groups. Although the Both group reported uniformly higher frequencies than the Neither group on all of the antisocial behaviors (including fewer years of monogamy), comparisons between the Both and the Adult Only groups produced more variable results. In general, the antisocial behavion of the Adult Only group was intermediate between that of the Both and Neither groups and most often did not differ significantly from that of either group. Compared to the Both group, the Adult Only group reported a significantly lower number of felony arrests (q=4.69, df=3, 216, p=O.Ol) and felonies not resulting in arrest (q=3.23, df=3, 216, p=O.OS) and more continuous years of monogamy (q=4.30, df=3, 216, p=O.Ol). In contrast, the two groups did not differ significantly on number of misdemeanor arrests, misdemeanors not resuiting in arrest, or number of different employers since the age of 1 8 years.
485
ANTISOCIAL
BEHAVIOR
OF
DRUG
ABUSERS
Although subjects in the Adult Only group were more likely than the Neither group to be involved in most measures of adult antisocial behavior (table 1 ), they did not typically differ from each other in the frequencies of these behaviors (table 2). The one exception to this general finding was that subjects in the Adult Only group used weapons more frequently in fights than did those in the Neither group (q=3.73, df=3, 216, p=O.OS). Finally, the three groups were compared on the previously described summary measure of severe criminality and physical aggressiveness. The Both group obtamed a significantly higher mean score (48.13) than that of the Adult Only group (20.18) (q=3.47, df=3, 216, p