The Epidemiology of Psychiatric Disorders Among Repeat DUI ...

Journal of Consulting and Clinical Psychology
Copyright 2007 by the American Psychological Association
2007, Vol. 75, No. 5, 795– 804
0022-006X/07/$12.00
DOI: 10.1037/0022-006X.75.5.795
The Epidemiology of Psychiatric Disorders Among Repeat DUI Offenders
Accepting a Treatment-Sentencing Option
Howard J. Shaffer, Sarah E. Nelson,
Gabriel Caro
Debi A. LaPlante, Richard A. LaBrie, and
Cambridge Health Alliance
Mark Albanese
Harvard Medical School and Cambridge Health Alliance
Psychiatric comorbidity likely contributes to driving under the influence (DUI) of alcohol among repeat
offenders. This study presents one of the first descriptions of the prevalence and comorbidity of
psychiatric disorders among repeat DUI offenders in treatment. Participants included all consenting
eligible admissions (N
729) to a 2-week inpatient treatment facility for court-sentenced repeat DUI
offenders (i.e., offenders electing treatment in place of prison time) from April 17, 2005, to April 23,
2006. Participants completed the Composite International Diagnostic Interview, which assessed the
following disorders using criteria from the Diagnostic and Statistical Manual of Mental Disorders (4th
ed.; American Psychiatric Association, 1994): alcohol use and drug use, bipolar, generalized anxiety,
posttraumatic stress, intermittent expolosive, conduct, attention deficit, nicotine dependence, pathological
gambling, and major depressive. Repeat DUI offenders evidenced higher lifetime and 12-month preva-
lence of alcohol use and drug use disorders, conduct disorder, posttraumatic stress disorder, generalized
anxiety disorder, and bipolar disorder compared with the general population. Almost half qualified for
lifetime diagnoses of both addiction (i.e., alcohol, drug, nicotine, and/or gambling) and a psychiatric
disorder. Lifetime and past-year comorbidity rates were higher among participants than in the general
population. These results suggest that clinicians should consider multimorbidity within DUI treatment
protocols.
Keywords: alcohol abuse, comorbidity, epidemiology, driving under the influence
Like heart disease, cancer, domestic violence, and sexually
2004, alcohol was involved in 39% of fatal crashes and that
transmitted diseases, driving under the influence of alcohol (DUI)
248,000 people were injured or killed in alcohol-related crashes
continues to be a primary public health concern. The U.S. National
(NHTSA, 2006). These fatalities and injuries contribute to enor-
Highway Traffic Safety Administration (NHTSA) reported that, in
mous but avoidable financial and human costs. Taylor, Miller, and
Cox (2002) estimated that the annual economic cost of alcohol-
related accidents is roughly $51.1 billion. Furthermore, there is
evidence that the majority of DUI offenders go undetected. For
Howard J. Shaffer, Sarah E. Nelson, Debi A. LaPlante, Richard A.
example, in 1998, the Centers for Disease Control reported that
LaBrie, and Mark Albanese, Department of Psychiatry, Harvard Medical
only 1% of the DUI episodes reported by U.S. adults resulted in
School and Division on Addictions, Cambridge Health Alliance, Medford,
arrest (Centers for Disease Control, 2006). More than 80% of
Massachusetts; Gabriel Caro, Division on Addictions, Cambridge Health
alcohol-impaired drivers who caused accidents and were admitted
Alliance, Medford, Massachusetts.
to hospital emergency departments were not prosecuted for their
The National Institute of Alcohol Abuse and Alcoholism provided
offense (Orsay, Doan-Wiggins, Lewis, Lucke, & RamaKrishnan,
primary support for this study as part of Grant R01 AA014710-01A1. We
also received support from the Institute for Research on Pathological
1994), a statistic that reveals DUI is an even larger problem than
Gambling and Related Disorders and from bwin.com, Interactive Enter-
the arrest and conviction rate statistics indicate.
tainment, AG. We extend special thanks to the directors and counselors at
the Middlesex Driving Under the Influence of Liquor (MDUIL) treatment
Reducing DUI Behavior
program—Charles Karayianis, Michael Kennedy, Jim Barry, Christine
Breen, Daniel Gallo, Karen Horrigan, and Michael Jezylo—as well as to
NHTSA (2001) suggests that there are four legal approaches for
the entire MDUIL staff for their collaboration on this project. We also
addressing DUI offenses: (a) licensing sanctions, (b) vehicle sanc-
thank Leslie Bosworth, Andy Boudreau, Sarbani Hazra, Rachel Kidman,
tions, (c) mandatory alcohol abuse treatment and education, and
John Kleschinsky, Siri Odegaard, Allyson Peller, Michael Stanton, Chris-
(d) mandatory sentencing. These policies are meant to raise aware-
tine Thurmond, Audrey Tse, and Erinn Walsh for their support and work
ness about the dangers of drinking and driving or to increase
on this project. Sarah E. Nelson and Howard J. Shaffer had full access to
penalties for DUI behavior. Such DUI policies target (a) individ-
all of the data in the study and take responsibility for the integrity of the
uals who are not aware of these dangers and (b) individuals for
data and the accuracy of the data analysis.
whom harsh potential consequences should deter the decision to
Correspondence concerning this article should be addressed to Howard
J. Shaffer, Division on Addictions, 101 Station Landing, 2nd Floor, Med-
drink and to drive. Treatment programs for convicted DUI offend-
ford, MA 02155. E-mail: howard_shaffer@hms.harvard.edu
ers have been an integral feature of the criminal justice system for
795

---

796
SHAFFER ET AL.
about the past 30 years (e.g., Nochajski & Stasiewicz, 2006;
measure (Minnesota Multiphasic Personality Inventory) and the Al-
Williams, 2006). Some of these treatment programs define addic-
cohol Use Inventory, found that multiple offenders, compared with
tion as a central feature of DUI (Nochajski & Stasiewicz, 2006)
first-time offenders, had increased rates of mania, depression, hostil-
and of its treatment (e.g., Donovan, Marlatt, & Salzberg, 1983;
ity, alcohol and other drug consumption, alcohol- and drug-related
Marlatt, Baer, Donovan, & Kivlahan, 1988). The criminal justice
problems, traffic accidents, and nontraffic arrests. Similarly, Cavaiola
and addiction approach has met with some success (e.g., Glass,
et al. (2007) reported that elevated validity scale scores (L, lie or “fake
Chan, & Rentz, 2000; NHTSA, 2000; Nochajski & Stasiewicz,
good”; F, “fake bad”; and K, defensiveness) on the Minnesota Mul-
2006; Williams, 2006). These legal approaches and specific policy
tiphasic Personality Inventory predicted relapse among first-time DUI
interventions, such as lowering the legal blood alcohol concentra-
offenders.
tion and mandating stricter vehicle safety measures, have reached
their intended targets and have helped reduce alcohol-related fa-
Repeat DUI Offenders
talities by 21% since 1982 (NHTSA, 2006). Unfortunately, the
reduction in the proportion of traffic fatalities that are alcohol-
Repeat DUI offenders likely contribute to consistent rates of
related has leveled off during recent years: Since 1993, rates have
DUI accidents and deaths. Statistics (NHTSA, 2004b) reveal that
hovered between 39% and 43% (NHTSA, 2002, 2004a, 2006). The
up to one third of DUI arrestees are repeat offenders. Repeat DUI
recent stagnation in DUI fatality reduction after the earlier decline
offenders also are disproportionately responsible for DUI-related
suggests that targeting only individuals who are not aware of the
harms, as evidenced by their significant involvement in alcohol-
dangers of DUI and individuals who respond to harsh legal con-
related driving fatalities. Among drivers involved in fatal crashes,
sequences is insufficient. New and enhanced treatment-matching
those with blood alcohol concentration levels of 0.08% or higher
strategies (e.g., Wells-Parker, Dill, Williams, & Stoduto, 2006;
were nine times more likely to have a prior conviction for driving
Wells-Parker & Williams, 2002) are showing considerable prom-
while impaired than were drivers who had not consumed alcohol
ise, but these protocols are not yet widely available. The failure to
(NHTSA, 2006). Other research suggests that drivers who die in
continue reduction of DUI reveals that a subset of the DUI of-
alcohol-related crashes are more than four times more likely to
fender population is nonresponsive to the current intervention
have a history of DUI arrest than are drivers who die in non-
efforts and that different strategies will be necessary for further
alcohol-related crashes (Brewer et al., 1994; NHTSA, 2006).
reduction of DUI behavior (Yu, Evans, & Clark, 2006). In a recent
These statistics, coupled with the fact that repeat offenders per-
review, Williams (2006) noted that DUI remains a public health
petuate their behavior despite the negative consequences of past
concern and deserves our continuing attention. However, before
arrest, suggest that repeat offenders represent a group that is
public health professionals can develop new interventions, addi-
distinct not only from the general population but also, possibly,
tional research is necessary for determination of the characteristics
from the majority of first-time DUI offenders.
of DUI offenders who are resistant to traditional interventions.
During the past 30 years, various investigators have identified
and implicated comorbid psychiatric disorders and demographic
Psychiatric Comorbidity and DUI Behavior
attributes (e.g., gender, age, reduced acculturation) as causal in-
fluences for DUI relapse (Cavaiola et al., 2007; C’de Baca et al.,
Research suggests that psychiatric disorders can decrease the
2001; Glass et al., 2000; Hunter et al., 2006; Lapham et al., 2001;
effectiveness of substance abuse treatment (Albanese, 2001; Al-
Nochajski & Stasiewicz, 2006; Wells-Parker & Williams, 2002).
banese & Shaffer, 2003; Bradizza, Stasiewicz, & Paas, 2006).
However, only Lapham et al. (2006) reported the epidemiology of
Untreated or partially treated psychopathology might contribute to
psychiatric disorders for a repeat DUI offender population with a
the persisting rate of DUI events and fatalities. Some research
comprehensive measure of diagnoses from the Diagnostic and
suggests that rates of psychiatric disorders, including substance use
Statistical Manual of Mental Disorders (4th ed.; DSM–IV; Amer-
disorders and depression, are elevated among DUI offenders
ican Psychiatric Association, 1994). They reported that, in addition
(Lapham, C’de Baca, McMillan, & Lapidus, 2006; Lapham et al.,
to alcohol abuse or dependence, 50% of their sample had a lifetime
2001; Oslin, O’Brien, & Katz, 1999) and that there might be an
drug use disorder, more than 30% qualified for depression, and
association between various psychiatric disturbances and disorders
more than 15% had experienced posttraumatic stress disorder.
and DUI reoffense (e.g., Cavaiola, Strohmetz, & Abreo, 2007;
Though the rates of psychiatric comorbidity were similar to those
C’de Baca, Miller, & Lapham, 2001; Donovan et al., 1983; Glass
that these researchers found within their first-offender sample
et al., 2000; Hunter, Wong, Beighley, & Morral, 2006; McMillen,
(Lapham et al., 2001), the rate of drug use disorders was much
Adams, Wells-Parker, Pang, & Anderson, 1992; Nochajski &
higher among repeat offenders. Because repeat offenders are re-
Stasiewicz, 2006; Wells-Parker et al., 2006).
sponsible for a disproportionate public health burden, it is impor-
Lapham et al. (2001) conducted one of the first studies to assess
tant to continue examination of the extent and patterns of psycho-
psychiatric disorders among DUI offenders comprehensively. Using a
pathology among this specific population. Creation of a
sample of primarily first-time offenders, this research revealed that
comprehensive epidemiological profile of psychopathology among
85% of women and 91% of men referred to a DUI program met the
repeat DUI offenders mandated to treatment is fundamental to
diagnostic criteria for alcohol dependence or abuse; 32% of the
advancement of our understanding of DUI and development of
women and 38% of the men had a drug use disorder. For offenders
optimal clinical interventions. Because psychiatric disorders tend
with alcohol use disorders, 50% of women and 33% of men had at
to decrease the efficacy of substance abuse treatment and to
least one additional psychiatric disorder, primarily posttraumatic
increase relapse (Albanese, 2001; Albanese & Shaffer, 2003;
stress disorder or major depression. Although they did not focus on
Bradizza et al., 2006), unrecognized psychopathology likely limits
diagnostic categories, McMillen et al. (1992), using a personality
the value of clinical efforts to curtail DUI reoffense.

---

PSYCHIATRIC DISORDERS AMONG REPEAT DUI OFFENDERS
797
Admissions
1,063 MDUIL admissions
between 4/17/05 and 4/23/06
85 clients w/ language barrier
(8% of admissions)
33 CIDIs incomplete (discharged
Exclusions
prior to interview completion)
(3% of admissions)
(by MDUIL)
25 clients w/ cognitive difficulties
(2% of admissions)
Eligible
920 completed CIDI intake
interviews (87% of admissions)
Clients
Recruitment
729 consenting participants
(79% of eligible clients)
Rate
Figure 1.
Study eligibility and recruitment. MDUIL
Middlesex Driving Under the Influence of Liquor
Program; CIDI
Composite International Diagnostic Interview.
Current Study
different intake assessment and were not eligible for this study. Of the
920 study-eligible admissions, we successfully recruited 729 clients
This article represents one of the first studies to use systematic
(i.e., 79% of eligible admissions and 69% of all consecutive admis-
methods and measures to report comprehensively the psychiatric
sions) into this study. Figure 1 summarizes the CIDI completion and
epidemiology and multimorbidity among repeat DUI offenders
study recruitment rates.
attending court-mandated treatment in place of incarceration. Pre-
Participants were 81% male. Information available from the
vious studies of repeat and first-time DUI offenders have used
MDUIL program indicated that the gender proportions of the
disparate methods and measures, a fact that limits their compara-
sample were similar to those of all admissions (i.e., 82% male)
bility. The current study measures psychiatric disorders with a
during the study time period. The sample was 88% Caucasian, 5%
widely used comprehensive diagnostic instrument that has been
Hispanic, 5% African American, 1% Native American, and 1%
employed both in Lapham’s recent work with repeat offenders and
Asian. The gender and race distribution in the counties served by
in general population studies. The current study builds upon the
MDUIL was 48%–53% female, 65%–96% Caucasian, 2%–18%
work of Lapham and her colleagues (2001, 2006) by including
Hispanic, 2%–25% African American, 1%– 8% Asian, and less
more possible diagnoses, comparing the rates for repeat offenders
than 1% Native American. MDUIL’s race distribution was most
with rates obtained in the general population using an identical
similar to the less metropolitan counties.
instrument, and analyzing the level of comorbidity among repeat
Participants ranged in age from 19 to 77 years (M
39.7
age
offenders. We hypothesized that repeat offenders would exhibit all
years, SD
11.6). Of the participants, 19% were married, 28%
measured Axis I psychiatric disorder at rates higher than those of
were divorced or separated, 2% were widowed, and 52% had never
the general population and that comorbidity would be common in
married. Personal income ranged from $0 to $100,000 . Among
this population. By providing a psychiatric epidemiology of repeat
those sampled, 32% had an income below $20,000, 42% had an
offenders in treatment, this research holds the potential to inform
income between $20,000 and $49,999, 13% had an income be-
clinical efforts that address psychopathology and that consequently
tween $50,000 and $74,999, and 9% had an income of $75,000 or
help to curtail recidivism.
more. (Four percent did not report their income.) Seventy-two
percent of the sample had a high school education or less, and 63%
Method
were employed at the time of the study. Sixty-two percent of the
sample reported 2 DUI convictions, and 36% reported more than
Participants
2 (range
3–10).1 The mean number of DUI convictions reported
The Middlesex Driving Under the Influence of Liquor Program
by participants was 2.5 (SD
1.0).
(MDUIL), a mandated 2-week inpatient treatment program for repeat
DUI offenders who elect a treatment-sentencing option instead of
Procedures
incarceration, had 1,063 consecutive admissions between April 17,
MDUIL, a 2-week inpatient DUI treatment facility for court-
2005, and April 23, 2006. Of those admitted, 920 (87%) completed
sentenced repeat DUI offenders, implemented a structured, com-
the Composite International Diagnostic Interview (CIDI) at intake.
MDUIL did not complete CIDIs with clients who required interpret-
ers (8% of admissions), who were discharged early (3% of admis-
1 Just 2% of participants reported one DUI arrest; however, the DUI
sions), or who, in the program director’s judgment, had severe cog-
program they attended admits only offenders with two or more DUI arrests
nitive difficulties (2% of admissions); these clients were given a
on their record.

---

798
SHAFFER ET AL.
puterized mental health assessment instrument (i.e., the CIDI) as
of disorders. The CIDI is employed worldwide; it is one of the
part of its program intake. During this study, MDUIL was one of
most comprehensive and thorough instruments currently available
two residential facilities in the state licensed by the Massachusetts
for use in diagnosis of substance use and of psychiatric disorders
Department of Public Health for provision of services to repeat
(e.g., Kessler, Abelson, et al., 2004; Kessler, Demyttenaere, et al.,
DUI offenders. In the state of Massachusetts, under Chapter 90,
2004; Kessler & Ustun, 2004; Kessler et al., 1998). The CIDI’s
Section 24, of the General Laws of Massachusetts (2005), repeat
favorable psychometric properties include (a) high concurrent va-
DUI offenders (i.e., offenders who have been convicted of more
lidity for substance use disorders, with a kappa of .83 measured
than one DUI offense in Massachusetts) can receive an alternative
against ICD-10 criteria (Janca, Robins, Cottler, & Early, 1992); (b)
sentence to incarceration. Specifically, the alternative requires 2
relative stability of the CIDI’s time-related symptom items, with a
years of supervised probation and a treatment condition of no less
test–retest concordance of 72.7% (intraclass correlation coeffi-
than 14 days in a residential treatment program, such as MDUIL,
cient
0.86) for age of abuse– dependence onset and 86.0%
followed by outpatient aftercare programming for the duration of
agreement for recency items (Wittchen et al., 1989); and (c) good
probation. If this option is not elected, offenders must serve no less
reliability and convergent validity with other diagnostic screening
than 30 days in prison (General Laws of Massachusetts, 2005).
procedures (Andrews & Peters, 1998; Lachner et al., 1998; Peters
Clients who attend MDUIL receive group and individual counsel-
& Andrews, 1995; Ustun et al., 1997). The National Comorbidity
ing sessions that provide education on methods for abstinence
Survey Replication (NCS-R; Kessler, Berglund, et al., 2004) used
from alcohol and from other drugs and on the physical effects of
the version of the CIDI employed in the current study as its
alcohol and of drug abuse. Participants are required to attend an
primary instrument; this circumstance provided us with a norma-
Alcoholics Anonymous meeting, two group counseling sessions,
tive community comparison group for the MDUIL data analyses.
and two to three educational classes during each day of the
The CIDI includes 21 modules that assess DSM–IV diagnostic
program. Individual counseling sessions are conducted several
criteria for Axis I disorders. MDUIL initially selected 8 of these
times each week.
modules and later added 2 more modules, on the basis of its needs
MDUIL served 53% of the repeat offenders in Massachusetts
and of an assessment of the modules clients tended to screen into
who were sentenced to treatment and who agreed to it in place of
during the pretest period. The CIDI modules used at MDUIL
prison time. MDUIL received admissions from 37 of the 39 district
provided DSM–IV-based lifetime and past-year diagnoses for al-
courts, which represented all of the counties, both urban and rural,
cohol dependence, alcohol abuse, drug dependence, drug abuse,
in the eastern half of the state during the study period. The 47% of
nicotine dependence, pathological gambling, major depression,
repeat offenders sentenced to treatment who did not attend
mania, hypomania, dysthymia, generalized anxiety disorder, post-
MDUIL received sentences in the western half of the state and
traumatic stress disorder, intermittent explosive disorder, conduct
attended treatment at the other state-licensed facility, mentioned
disorder, and attention-deficit disorder. The CIDI employs diag-
above.
nostic exclusion rules for each of the following disorders: alcohol
This study received approval from the Cambridge Health Alli-
abuse—no diagnosis of alcohol dependence; drug abuse—no di-
ance Institutional Review Board. Project staff (i.e., Sarah E. Nel-
agnosis of drug dependence; pathological gambling—no diagnosis
son, Howard J. Shaffer, and Gabriel Caro) trained MDUIL coun-
of mania; major depression—no diagnosis of mania or hypomania;
selors in the use of the CIDI prior to the study. An initial 3-month
hypomania—no diagnosis of mania; dysthymia—no diagnosis of
period was defined as the pretest period for use of the CIDI
major depression during the first 2 years of dysthymic symptoms
technology at MDUIL. During this period, only selected clients
and no diagnosis of mania or hypomania; generalized anxiety—no
were interviewed, as MDUIL staff learned how best to administer
temporally overlapping diagnosis of major depression; intermittent
the instrument and to integrate the CIDI into the program intake
explosive disorder—no diagnosis of mania, conduct disorder, or
process. After the pretest period, counselors administered the CIDI
attention-deficit disorder.
to all of their eligible clients within the 1st week of each 2-week
Two of the CIDI modules—intermittent explosive disorder and
patient cohort. (See Figure 1.) The average length of the interview
attention-deficit hyperactivity disorder—were administered start-
was 90 min; interviews ranged between 45 and 210 min.
ing with the fifth MDUIL cohort (i.e., after data had already been
Prior to discharge of clients from the MDUIL program, research
collected from 107 participants), because MDUIL reassessed the
staff met with them to obtain written informed consent for the use
time available for intakes and suspected an elevated presence of
of their intake information and to enroll them in the study. Clients
these disorders in its population. In addition to administering the
who agreed to participate received a $25 gift card from a grocery
CIDI modules, MDUIL counselors collected client demographic
or a department store.
information and arrest histories.
Instrument
Analysis Plan
We employed Version 19.101 of the computerized CIDI
We applied SAS-based diagnostic algorithms, which were pro-
(Kessler & Ustun, 2004) as the intake assessment at MDUIL. The
vided by the developers of the computerized CIDI (see www.hcp
CIDI is a comprehensive, standardized, and computer-guided in-
.med.harvard.edu/wmhcidi), to the data; this procedure yielded
strument for the assessment of substance use and other mental
diagnoses for both lifetime and past 12-month time frames for the
disorders in accordance with the definitions and criteria of the
disorders listed in Table 1. We then used SPSS to conduct descrip-
International Classification of Diseases, Tenth Revision (ICD-10;
tive analyses for the entire sample, identifying lifetime and 12-
World Health Organization, 1992), and of the DSM–IV. It applies
month prevalence rates for each of the disorders. Because of the
these criteria to measure the presence, age of onset, and remission
wide age range of our sample, we conducted logistic regressions to

---

PSYCHIATRIC DISORDERS AMONG REPEAT DUI OFFENDERS
799
Table 1
Lifetime and Past-Year Disorder Prevalence (N
729)
MDUIL
MDUIL
Gender-adjusted NCS-R ratesa
Effect size ( )
Lifetime
Past-year
Lifetime
Past-year
Past
Disorder
prevalence (%)
prevalence (%)
prevalence (%)
prevalence (%)
Lifetime
year
Addiction-related disorders
Alcohol abuse
56.9*
42.4*
10.2
2.3
.35
.50
Alcohol dependence
40.7*
31.1*
7.1
1.7
.30
.43
Drug abuse
25.9*
4.9*
6.4
1.4
.22
.08
Nicotine dependence
15.9*
12.5*
8.9
4.3
.08
.12
Drug dependence
14.7*
5.1*
3.9
0.5
.16
.15
Pathological gambling
1.9
1.1
—
—
—
—
Psychiatric disorders
Conduct disorder
17.8*
2.3
11.1
1.5
.08
.03
Posttraumatic stress disorder
13.3*
11.5*
4.8
2.4
.12
.16
Major depressive disorder
11.7
8.2*
14.2
5.4
.02
.03
Generalized anxiety disorder
8.4*
6.6*
4.8
2.2
.04
.07
Attention-deficit disorderb
8.4
5.8
9.2
4.2
.01
.03
Bipolar I or II disorder
7.3*
5.3*
4.3
2.9
.04
.04
(I with MDE)
(3.2)
(3.2)
—
—
—
—
(I without MDE)
(3.9)
(2.1)
—
—
—
—
(II)
(0.1)
(0.0)
—
—
—
—
Intermittent explosive disorderb
5.6
2.7
8.5
4.5
.03
.02
Dysthymia
3.3
2.7*
2.0
1.2
.02
.04
Note.
Dashes indicate that no data are available. MDUIL
Middlesex Driving Under the Influence of Liquor Program; NCS-R
National Comorbidity
Survey Replication; MDE
major depressive episode; DUI
driving under the influence (of alcohol).
* Significant difference between repeat DUI offenders and general population ( p
.01).
a NCS-R rates (Kessler, Berglund, Demler, Jin, & Walters, 2005; Kessler, Chiu, Demler, & Walters, 2005) have been adjusted by gender to reflect the
gender composition of the MDUIL sample: 81% male, 19% female. We used the tables available at http://www.hcp.med.harvard.edu/ncs/ftpdir/
table_ncsr_by_gender_and_age.pdf to perform this adjustment. The NCS-R sample for addiction-related disorders and posttraumatic stress disorder
includes 5,692 cases. NCS-R sample for conduct disorder and attention-deficit disorder includes 3,199 cases. NCS-R sample for other psychiatric disorders
includes 9,282 cases.
b Percentage for repeat DUI offender sample based on subsample of 623 cases.
determine whether prevalence rates differed by age. For analysis,
disorder (17.8%) and posttraumatic stress disorder (13.3%). Table
we collapsed the diagnoses for (a) mania and (b) hypomania with
1 summarizes prevalence rates for all lifetime disorders. Logistic
the presence of a major depressive episode into the category
regressions demonstrated that 4 of the 14 measured lifetime dis-
“Bipolar I or II”; this strategy replicates the NCS-R analytic
orders varied by age in our sample. Younger offenders were more
algorithm (Kessler, Berglund, Demler, Jin, & Walters, 2005;
likely to qualify for lifetime diagnoses of drug abuse (Wald’s
Kessler, Chiu, Demler, & Walters, 2005). We measured both
statistic
13.1, p
.001, Exp(B)
1.03), drug dependence
lifetime and 12-month comorbidity as the number of co-occurring
(Wald’s statistic
11.3, p
.01, Exp(B)
1.03), and conduct
disorders. We used regression analyses to determine whether co-
disorder (Wald’s statistic
11.7, p
.01, Exp(B)
1.03). Older
morbidity varied by age in our sample. We also compared preva-
offenders were more likely to qualify for nicotine dependence
lence rates and number of comorbid disorders between those
(Wald’s statistic
6.6, p
.05, Exp(B)
0.98).
offenders in our sample who reported two DUI offenses and those
Past 12 months.
The past 12-month rate of alcohol use disor-
who reported more than two. In addition to describing the preva-
ders (73.5%) was slightly lower than was the lifetime rate; the past
lence and the comorbidity of psychiatric disorders in our sample,
12-month prevalence rate of drug use disorders (10.0%) was much
we conducted chi-square analyses that compared our sample of
lower than was the lifetime rate. During the previous 12 months,
repeat DUI offenders with the NCS-R sample.
28.9% of the sample had experienced symptoms that qualified
them for a psychiatric disorder that was not substance related or
Results
gambling related. Posttraumatic stress disorder (11.5%) was the
most prevalent of these disorders. Table 1 summarizes the preva-
Prevalence of Psychiatric Disorders
lence rates for all disorders that participants experienced within the
Lifetime.
Of those sampled, 97.6% qualified for an alcohol use
previous 12 months. Logistic regressions demonstrated that 5 of
disorder, and 40.6% qualified for a drug use disorder. Forty-five
the 14 measured past-year disorders varied by age in our sample.
percent (44.5%) of the sample qualified for a psychiatric disorder
Younger offenders were more likely to qualify for past-year diag-
that was not substance related (i.e., alcohol, nicotine, or other drug)
noses of alcohol abuse (Wald’s statistic
6.7, p
.05, Exp(B)
or gambling related. The most prevalent of these were conduct
1.02), drug abuse (Wald’s statistic
10.0, p
.01, Exp(B)

---

800
SHAFFER ET AL.
1.06), drug dependence (Wald’s statistic
15.6, p
.001,
Table 2
Exp(B)
1.07), and conduct disorder (Wald’s statistic
9.9, p
Lifetime and Past-Year Comorbidity Patterns Among Repeat
.01, Exp(B)
1.09). Older offenders were more likely to qualify
DUI Offenders (N
729)
MDUIL
for pathological gambling (Wald’s statistic
4.3, p
.05,
Exp(B)
0.94).
Lifetime
Past-year
Pattern
prevalence (%)
prevalence (%)
Comparison with NCS-R.
To compare repeat DUI offenders
with a normative community sample, we examined lifetime and
No disordersa
1.1
16.5
past-year rates for those disorders, which were available from the
AA or AD onlya
35.0
42.7
general household population surveyed as part of the NCS-R
Only disorders other than
addiction-related disordersa
0.7
5.9
(Kessler, Berglund, et al., 2005; Kessler, Chiu, et al., 2005). We
Addiction-related disorders only
adjusted the NCS-R rates to approximate the gender composition
(AA/AD, DA/DD, ND, PG)
19.5
11.9
in our sample (i.e., 19% female, 81% male, as compared with 53%
Addiction-related disorders
1
female, 47% male) using the tables available on the NCS website
other disorder
26.1
14.8
(http://www.hcp.med.harvard.edu/ncs/ftpdir/table_ncsr_by_
Addiction-related disorders
2
other disorders
10.3
5.2
gender_and_age.pdf).2 Criteria for all diagnoses except alcohol
Addiction-related disorders
3
and drug disorders were identical, because both studies employed
or more other disorders
7.4
3.0
the same CIDI instrument. We used a slightly modified computer
algorithm to calculate alcohol and drug abuse and dependence in
Note.
DUI
driving under the influence (of alcohol); MDUIL
Mid-
dlesex Driving Under the Influence of Liquor Program; AA
alcohol
order to accurately capture both lifetime and past-year rates.3 In
abuse; AD
alcohol dependence; DA
drug abuse; DD
drug depen-
Table 1, asterisks mark statistically significant differences between
dence; ND
nicotine dependence; PG
pathological gambling.
prevalence estimates derived from our sample and from the gen-
a Past-year rates for these instances of no or single disorder occurrence
eral population.
(i.e., no disorders, alcohol disorders only, disorders other than addiction
As Table 1 shows, repeat DUI offenders had significantly higher
only) are higher than lifetime rates because overall comorbidity in the
sample is lower past year than lifetime.
lifetime rates than did the general population for alcohol and drug
use disorders, as well as for nicotine dependence. Similarly, con-
duct disorder, posttraumatic stress disorder, mania, and general-
addiction-related disorders. The majority of participants who had
ized anxiety disorder were significantly elevated in our sample.
only comorbid addiction-related disorders exhibited alcohol abuse/
There were no differences between the two samples for major
dependence and drug abuse/dependence. Participants who experi-
depression, dysthymia, intermittent explosive disorder, or
enced only one other comorbid disorder with their addiction-
attention-deficit disorder. Table 1 also shows that differences in
related disorders were most likely to have conduct disorder. No
past-year diagnoses reveal a pattern similar to that of lifetime
single pattern of lifetime comorbidity was more common than
diagnoses, though diagnoses of past-year major depression and of
dysthymia were significantly elevated in our sample, which was
not the case for the lifetime diagnoses of these disorders.
2 We also compared our sample and the NCS-R sample using NCS-R
rates unadjusted for gender. The only differences between these two sets of
Comorbidity of Psychiatric Disorders With Addiction-
analyses were that (a) the lifetime rate of depression in the NCS-R sample
Related Disorders
was significantly greater than that in the MDUIL sample, (b) the past-year
rate of conduct disorder in the MDUIL sample was significantly greater
To measure comorbidity, we first calculated the number of
than that in the NCS-R sample, and (c) the differences between past-year
disorders reported as co-occurring for each individual. We then
rates of major depression and dysthymia in the two samples were no longer
analyzed comorbidity patterns for each individual and collapsed
significant when NCS-R rates were left unadjusted.
3
these into six primary comorbidity categories: (a) no disorders, (b)
The only variations between diagnoses in the current study and those
only alcohol-related disorders, (c) alcohol disorder comorbid with
in the NCS-R involved alcohol and drug abuse and dependence. First, we
modified the CIDI gateways that determined whether participants would be
other addictions, (d) addiction-related disorder(s) comorbid
asked about symptoms of alcohol use disorder and drug use disorder. As
with one other disorder, (e) addiction-related disorder(s) comor-
programmed, Version 19.101 of the CIDI screens people out of these
bid with two other disorders, and (f) addiction-related disorder(s)
modules prior to asking about symptoms if participants indicate that their
comorbid with three or more other disorders. Table 2 summarizes
heaviest drinking period was in the past 12 months. We fixed this gating
the results for these categories. As before, we present these de-
problem to correctly include these people if the frequency and the amount
scriptive analyses for both lifetime and past-year diagnostic time
of their drinking in the past 12 months met documented CIDI criteria. We
frames. For lifetime diagnoses, comorbidity does not indicate
reported this problem, and the next version of the World Mental Health–
temporal co-occurrence. For past-year diagnoses, comorbidity in-
CIDI (post-NCS-R) included this revision. The second variation between
dicates co-occurrence of disorders within that 12-month time
our reporting and that from the NCS-R involved the distinction between
frame.
abuse and dependence. For alcohol abuse and drug abuse, the findings
reported from the NCS-R did not exclude individuals with alcohol depen-
Lifetime.
As shown in Table 2, 35.0% of the sample qualified
dence and drug dependence, respectively. Therefore, to assure compara-
only for lifetime diagnosis of alcohol abuse or dependence, and
bility, we adjusted the NCS-R rates reported here for substance abuse by
19.5% qualified only for addiction-related disorders (i.e., alcohol
subtracting that sample’s dependence rates. (For example, the NCS-R
abuse/dependence and drug abuse/dependence, nicotine depen-
reports a lifetime alcohol abuse rate of 13.2% and a lifetime alcohol
dence, and/or pathological gambling). More than 43% of partici-
dependence rate of 5.4%; for this comparison, the NCS-R lifetime alcohol
pants qualified for one or more lifetime disorders in addition to
abuse rate is adjusted to 7.8%.)

---

PSYCHIATRIC DISORDERS AMONG REPEAT DUI OFFENDERS
801
others for participants who qualified for two or more non-
disorder, obsessive-compulsive disorder, and oppositional defiant
addiction-related disorders. Regression analyses demonstrated that
disorder), the current comparisons, illustrated in Figure 2, should
younger offenders qualified for significantly more comorbid dis-
be considered conservative.
orders than did older offenders,
.09, t(728)
2.5, p
.05,
Despite the conservative estimates of co-occurring disorders in
but age accounted for less than 1% of the variance in comorbidity.
the repeat DUI sample, comorbidity among repeat DUI offenders
Past 12 months.
As summarized in Table 2, 16.5% of the
was significantly higher than it was in the general population,
sample qualified for no disorders, and 42.7% qualified only for
within both a lifetime and a past-year time frame.
past-year alcohol abuse or dependence. More than 30% of the
sample reported a co-occurrence of disorders within the past year,
Multiple DUI Offending and Prevalence and Comorbidity
with 11.9% qualifying only for comorbid past-year addiction-
of Psychiatric Disorders
related disorders and 23.0% qualifying for one or more past-year
disorders in addition to past-year addiction-related disorders. Past-
If, as we suggest, it is important to consider DUI offenders who
year anxiety-related disorders and bipolar disorders continued to
reoffend separately from first-time offenders, the question arises
co-occur with addiction-related disorders at prevalence levels sim-
whether mental health issues increase as DUI arrests increase. In
ilar to those for the corresponding lifetime estimates. Postraumatic
our sample, 35% of offenders have more than two arrests. Chi-
stress disorder was the most commonly occurring disorder comor-
square analyses demonstrated that these offenders were more
bid with the addictions during the past 12 months. Regression
likely to qualify for a lifetime diagnosis of alcohol dependence,
analyses demonstrated that younger offenders qualified for signif-
2(1, N
729)
4.8, p
.05,
.08, and were less likely to
icantly more past-year comorbid disorders than did older offend-
qualify for a past-year diagnosis of alcohol abuse,
2(1, N
ers,
.09, t(728)
2.5, p
.05; again, age accounted for
729)
8.3, p
.01,
.11, than were offenders with two DUI
less than 1% of the variance in comorbidity.
arrests. This group did not differ on rates of other disorders from
Comparison with NCS-R.
In papers analyzing the NCS-R,
those with two DUI offenses. Offenders with four or more arrests
Kessler and colleagues report the prevalence of lifetime and 12-
demonstrated the same pattern: They were more likely than the rest
month comorbidity. For lifetime comorbidity, they use the cate-
of the sample to qualify for a lifetime diagnosis of alcohol depen-
gories “any disorder,” “two or more disorders,” and “three or more
dence, 2(1, N
729)
5.0, p
.05,
.08, and, consequently,
disorders” (Kessler, Berglund, et al., 2005). For past-year comor-
were less likely to qualify for a lifetime diagnosis of alcohol abuse,
bidity, they use the categories “any disorder,” “one disorder,” “two
2(1, N
729)
3.9, p
.05,
.07. This finding was
disorders,” and “three or more disorders” (Kessler, Chiu, et al.,
observed just below statistical significance when we considered
2005). For purposes of comparison, we reanalyzed our data using
offenders with three or more arrests. Offenders with four or more
the NCS-R classification schema and including only disorders that
arrests also were less likely to qualify for past-year alcohol abuse,
were measured in the NCS-R and reported in available published
2(1, N
729)
8.5, p
.01,
.11, than was the rest of the
papers. Because the NCS-R also included seven disorders not
sample. Overall, 37.8% of offenders who reported two arrests,
measured in the current repeat DUI study (i.e., panic disorder,
43.2% of offenders who reported three arrests, and 51.7% of
specific phobia, social phobia, agoraphobia, separation anxiety
offenders who reported four or more arrests qualified for a lifetime
3+ Past Year Disorders
*
2 Past Year Disorders
*
1 Past Year Disorder
*
Any Past Year Disorder
*
3+ Lifetime Disorders
*
2+ Lifetime Disorders
*
Any Lifetime Disorder
*
0%
20%
40%
60%
80%
100%
Lifetime Disorder Prevalence
Repeat DUI Offenders
General Population
Figure 2.
Comparison between repeat DUI offenders and general population (NCS-R) comorbidity of psy-
chiatric disorders. Counts for repeat DUI offender sample are based only on disorders measured by both this
study and the NCS-R. DUI
driving under the influence (of alcohol); NCS-R
National Comorbidity Survey
Replication. * p
.001.

---

802
SHAFFER ET AL.
diagnosis of alcohol dependence. Offenders with more DUI arrests
study included clinical significance criteria (i.e., participants were
were not more likely to exhibit comorbidity (i.e., to qualify for
required to report qualifying symptoms and to acknowledge that
more lifetime or past-year disorders), whether we compared of-
these symptoms have had a serious impact on their ability to
fenders with three or more reported arrests or offenders with four
function) that were not included in the original NCS. Given the
or more reported arrests with the rest of the sample.
comparability of our instrument to that used in the NCS-R, com-
parison of our results with those for participants with alcohol use
Discussion
disorders will be interesting. Though the current data do not inform
us about the temporal patterns of anxiety and of alcohol use, it is
Almost 100% of our repeat DUI sample qualified for a lifetime
likely that some of these repeat offenders use alcohol to cope with
diagnosis of alcohol abuse or dependence, and three quarters of the
the trauma and anxiety they experience. It also is possible that
sample met criteria for one of those diagnoses within the past year.
anxiety disorders emerge as a result of excessive alcohol use and
This finding is consistent with Lapham et al.’s recent study (2006)
of dependence.
of repeat DUI offenders. In their sample of primarily first-time
Compared with the rate for the general population, the lifetime
DUI offenders, Lapham et al. (2001) observed lifetime rates of
rate of depression in this sample is not elevated; this seems
alcohol-related disorders that were similar to the rates reported
surprising, given past work that links substance use disorders to
here; however, the past-year rates for their first-time sample
depression (Kessler et al., 1997, 2003). There are several possible
(34.4% for women and 40.0% for men) were considerably lower
explanations for this finding. Our sample of repeat offenders might
than in both their repeat DUI offender sample (70.6%) and our
regard their depressive symptoms as physical consequences of
repeat DUI offender sample (73.5%). This finding suggests that
their drinking behavior, which would preclude a diagnosis for
alcohol use disorders reported by repeat DUI offenders in treat-
depression in the CIDI. It is also possible that repeat offenders who
ment persist, unlike those reported by first offenders. In addition,
experience depressive episodes tend to experience mania; this
number of DUI offenses in our sample was associated specifically
circumstance qualifies them for bipolar disorder instead of major
with greater likelihood of alcohol dependence. We can speculate
depression. Extending this interpretation, repeat offenders in treat-
about the many reasons for this association; one possibility is that
ment appear to “act out” rather than “act in” (e.g., Krueger, Caspi,
alcohol dependence not only increases the likelihood of DUI but
Moffitt, & Silva, 1998). The symptom patterns of repeat offenders
might increase the likelihood of getting caught, because of the
tend to qualify them for externalizing disorders, such as substance-
increased severity and duration of alcohol-related morbidity (Yu et
related disorders and conduct disorder. These behaviors might be
al., 2006). However, it also is possible that when offenders are
responses to negative emotions that others internalize and do not
arrested multiple times for DUI, it increases their recognition of
discharge. However, Lapham et al. (2006) found a much higher
alcohol-related problems and reduces their resistance to reporting
lifetime rate of depression (30.9%) among their sample of repeat
symptoms.
DUI offenders. Given that they used a similar instrument, this
Our findings confirm and expand the past findings of psychiatric
higher rate might be due to regional differences—Lapham et al.
comorbidity among DUI offenders that were discussed earlier.
obtained their sample in a northwestern city— but the divergent
More than 60% of our sample qualified for a lifetime mental
findings clearly merit further investigation. Results from the
disorder in addition to alcohol-related problems. Strikingly, almost
NESARC study indicate that, in a nationally representative sam-
half of the sample (45%) met criteria for a lifetime diagnosis of a
ple, depression is more strongly associated with drug use disorders
mental disorder that was not substance related. Comorbidity was
than it is with alcohol use disorders; within alcohol use disorders,
considerably elevated in this sample compared with the general
depression is associated more strongly with dependence than it is
population, and either lifetime or past-year prevalence rates were
with abuse (Hasin, Goodwin, Stinson, & Grant, 2005). Further
significantly higher than were those for the general population for
examination of our study sample might reveal elevated depression
10 of 12 disorders. In particular, repeat offenders in treatment had
among subgroups with alcohol or drug dependence. As we men-
elevated rates of posttraumatic stress disorder, both current and
tioned earlier, now that the data are available for the NCS-R, we
lifetime, and were more likely than was the general population to
will need to compare our results with those for participants with
have a lifetime diagnosis of conduct disorder or of bipolar disor-
alcohol use disorders in that sample.
der.
Age is an important factor for those considering repeat DUI
The extent of anxiety-related disorders experienced by this
offenders: We found that younger offenders were more likely to
sample suggests an interaction between alcohol and anxiety among
have lifetime and past-year drug use disorders than were their
repeat DUI offenders. This is not surprising, given previous re-
older counterparts. Kessler et al. (Kessler, Berglund, et al., 2005;
search on the co-occurrence of anxiety and of alcohol use disorders
Kessler, Chiu, et al., 2005; see also http://www.hcp.med.harvard
(Kessler et al., 1997), though the most recent national survey, the
.edu/ncs/ftpdir/table_ncsr_by_gender_and_age.pdf) observed sim-
National Survey on Alcohol and Related Conditions (NESARC),
ilar patterns of both lifetime and 12-month comoribidity. Studies
reported that this comorbidity was limited to alcohol dependence
have confirmed that young people tend to have more externalizing
(Grant et al., 2004, 2005). One might wonder why the rates are not
patterns of behavior (e.g., substance use) than do their older
higher; the National Comorbidity Study (NCS) found that 23% of
counterparts, a finding that suggests different diagnostic popula-
participants with lifetime alcohol abuse and 36% of participants
tion segments. However, it also is possible that the observation of
with lifetime alcohol dependence also evidenced lifetime anxiety-
this pattern among lifetime prevalence estimates reveals a self-
related disorders. However, the NCS included multiple anxiety
report artifact: Older participants might not recall or report as
disorders not included in the current study (e.g., phobias and panic
many symptoms as do younger participants. More research is
disorder). In addition, the version of the CIDI that we used for this
necessary for clarification of these findings.

---

PSYCHIATRIC DISORDERS AMONG REPEAT DUI OFFENDERS
803
Treatment Implications
References
Our findings support the proposal that conventional DUI treat-
Albanese, M. J. (2001). Assessing and treating comorbid mood and sub-
ments and interventions that focus primarily on education and on
stance use disorders. Psychiatric Times, 18, 55–58.
punishment likely are insufficient for dealing with the observed
Albanese, M. J., & Shaffer, H. J. (2003). Treatment considerations in
level of psychiatric comorbidity among repeat DUI offenders.
patients with addictions. Primary Psychiatry, 10, 55– 60.
American Psychiatric Association. (1994). Diagnostic and statistical man-
Despite the evidence that DUI offenders are a heterogenous group
ual of mental disorders (4th ed.). Washington, DC: Author.
and that cognitive– behavioral and treatment-matching approaches
Andrews, G., & Peters, L. (1998). The psychometric properties of the
have shown promise for limiting relapse within contemporary
Composite International Diagnostic Interview. Social Psychiatry & Psy-
treatment programs (e.g., Donovan et al., 1983; Glass et al., 2000;
chiatric Epidemiology, 33, 80 – 88.
Wells-Parker et al., 2006; Wells-Parker & Williams, 2002), current
Bradizza, C. M., Stasiewicz, P. R., & Paas, N. D. (2006). Relapse to
DUI-related public policy in theory and action inadvertently tends
alcohol and drug use among individuals diagnosed with co-occurring
to support a one-size-fits-all approach to treatment for DUI of-
mental health and substance use disorders: A review. Clinical Psychol-
fenses. Limited resources combined with the demands of public
ogy Review, 26, 162–178.
Brewer, R. D., Morris, P. D., Cole, T. B., Watkins, S., Patetta, M. J., &
policy governing DUI treatment encourage current DUI treatment
Popkin, C. (1994). The risk of dying in alcohol-related automobile
programs to focus on the obligatory education about drinking (i.e.,
crashes among habitual drunk drivers. New England Journal of Medi-
alcohol abuse and dependence) and driving; even the most com-
cine, 331, 513–517.
prehensive DUI programs often do not have the resources neces-
Cavaiola, A. A., Strohmetz, D. B., & Abreo, S. D. (2007). Characteristics
sary for addressing the variety of psychiatric needs among their
of DUI recidivists: A 12-year follow-up study of first time DUI offend-
patients. However, the extant scientific literature and this study
ers. Addictive Behaviors, 32, 855– 861.
describe a heterogeneous offender population with extensive psy-
C’de Baca, J., Miller, W. R., & Lapham, S. C. (2001). A multiple risk
chiatric comorbidity that requires more comprehensive treatment
factor approach for predicting DWI recidivism. Journal of Substance
Abuse Treatment, 21, 207–215.
options for maximizing clinical outcomes and for reducing the
Centers for Disease Control. (2006). Alcohol factsheet— general informa-
incidence of renewed DUI offenses (C’de Baca et al., 2001;
tion. Retrieved April 19, 2006, from http://www.cdc.gov/alcohol/
Reynolds, Kunce, & Cope, 1991; Wieczorek & Miller, 1992).
factsheets/general_information.htm
Donovan, D. M., Marlatt, G. A., & Salzberg, P. M. (1983). Drinking
behavior, personality factors and high-risk driving. A review and theo-
Limitations and Future Directions
retical formulation. Journal of Studies on Alcohol, 44, 395– 428.
General Laws of Massachusetts. Motor vehicles and aircraft, Part 1, Title
The current data are derived from a sample of repeat DUI
XIV, Chapter 90, Section 24. Retrieved July 15, 2005, from http://
offenders that was confined to a single treatment program in
www.state.ma.us/legis/laws/mgl/90 –24.htm
Massachusetts. Despite the evidence showing that this sample is
Glass, R. J., Chan, G., & Rentz, D. (2000). Cognitive impairment screening
representative of repeat DUI offenders in Massachusetts who elect
in second offense DUI programs. Journal of Substance Abuse Treat-
a treatment sentence option, these findings need to be replicated in
ment, 19, 369 –373.
additional settings and in communities in other states. In Massa-
Grant, B. F., Hasin, D. S., Stinson, F. S., Dawson, D. A., Ruan, W. J.,
chusetts, offenders mandated to treatment can refuse this sentenc-
Goldstein, R. B., et al. (2005). Prevalence, correlates, co-morbidity, and
comparative disability of DSM–IV generalized anxiety disorder in the
ing option and can serve 30 or more days in prison instead. Thus,
USA: Results from the National Epidemiologic Survey on Alcohol and
our findings are specific to offenders who agree to attend a 2-week
Related Conditions. Psychological Medicine, 35, 1–13.
inpatient program and outpatient aftercare program in place of
Grant, B. F., Stinson, F. S., Dawson, D. A., Chou, S. P., Dufour, M. C.,
30
days in prison. Further, the language and cognitive exclusions
Compton, W., et al. (2004). Prevalence and co-occurrence of substance
required for CIDI administration likely skewed the results in a
use disorders and independent mood and anxiety disorders. Archives of
conservative direction. It is reasonable to expect higher rates of
General Psychiatry, 61, 807– 816.
psychiatric disorder and of comorbidity among those excluded for
Hasin, D. S., Goodwin, R. D., Stinson, F. S., & Grant, B. F. (2005).
cognitive difficulties.
Epidemiology of major depressive disorder: Results from the National
Epidemiologic Survey on Alcoholism and Related Conditions. Archives
The current research provides a comprehensive and systematic
of General Psychiatry, 62, 1097–1106.
study of psychiatric comorbidity in a sample of repeat offenders.
Hunter, S. B., Wong, E., Beighley, C. M., & Morral, A. R. (2006).
Understanding the role that mental health issues play in this
Acculturation and driving under the influence: A study of repeat offend-
population, for which current legal sanctions are not completely
ers. Journal of Studies on Alcohol, 67, 458 – 464.
successful, is particularly important for prevention and treatment
Janca, A., Robins, L. N., Cottler, L. B., & Early, T. S. (1992). Clinical
of DUI behavior. Given the prevalence of psychiatric disorders
observation of assessment using the Composite International Diagnostic
among repeat DUI offenders in treatment, future research needs to
Interview (CIDI): An analysis of the CIDI field trials—wave II at the St
(a) identify the causal relationships among psychiatric disorder,
Louis site. British Journal of Psychiatry, 160, 815– 818.
alcohol problems, and DUI behavior; (b) investigate whether the
Kessler, R. C., Abelson, J. M., Demler, O., Escobar, J. I., Gibbon, M.,
Guyer, M. E., et al. (2004). Clinical calibration of DSM–IV diagnoses in
patterns of psychiatric comorbidity among repeat DUI offenders
the World Mental Health (WMH) version of the World Health Organi-
differ for men and for women; (c) examine the factors that distin-
zation (WHO) Composite International Diagnostic Interview (CIDI).
guish repeat offenders from onetime offenders; and (d) structure
International Journal of Methods in Psychiatric Research, 13, 122–139.
and evaluate treatment to address the psychiatric issues that con-
Kessler, R. C., Berglund, P., Chiu, W. T., Demler, O., Heeringa, S., Hiripi,
tribute to repeat DUI behavior.
E., et al. (2004). The U.S. National Comorbidity Survey Replication

---

804
SHAFFER ET AL.
(NCS-R): Design and field procedures. International Journal of Methods
facts 2001. [Fact sheet]. Retrieved November 20, 2002, from http://
in Psychiatric Research, 13, 69 –92.
www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSF2001/2001alcohol.pdf
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D. S., Merikan-
National Highway Traffic Safety Administration. (2004a). Traffic safety
gas, K. R., et al. (2003). The epidemiology of major depressive disorder:
facts 2003: Alcohol. [Fact sheet]. Retrieved January 10, 2005, from
Results from the National Comorbidity Survey Replication (NCS-R).
http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSF2003/809767.pdf
Journal of the American Medical Association, 289, 3095–3105.
National Highway Traffic Safety Administration. (2004b). Traffic safety
Kessler, R. C., Berglund, P., Demler, O., Jin, R., & Walters, E. E. (2005).
facts—Repeat intoxicated drivers laws. [Fact sheet]. Retrieved April 19,
Lifetime prevalence and age-of-onset distributions of DSM–IV disorders
2006, from http://www.nhtsa.dot.gov/people/injury/new-fact-sheet03/
in the National Comorbidity Survey Replication. Archives of General
RepeatIntoxicated.pdf
Psychiatry, 62, 593– 602.
National Highway Traffic Safety Administration. (2006). Motor vehicle
Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005). Preva-
and crash fatality counts and estimates of people injured for 2005. [Fact
lence, severity, and comorbidity of 12-month DSM–IV disorders in the
sheet]. Retrieved April 19, 2006, from www.nhtsa.gov
National Comorbidity Survey Replication. Archives of General Psychi-
Nochajski, T. H., & Stasiewicz, P. R. (2006). Relapse to driving under the
atry, 62, 617– 627.
influence (DUI): A review. Clinical Psychology Review, 26, 179 –195.
Kessler, R. C., Crum, R. M., Warner, L. A., Nelson, C. B., Schulenberg, J.,
Orsay, E. M., Doan-Wiggins, L., Lewis, R., Lucke, R., & RamaKrishnan,
& Anthony, J. C. (1997). Lifetime co-occurrence of DSM–III–R alcohol
V. (1994). The impaired driver: Hospital and police detection of alcohol
abuse and dependence with other psychiatric disorders in the National
an