Content uploaded by Sherry H Stewart
Author content
All content in this area was uploaded by Sherry H Stewart
Content may be subject to copyright.
Associations of Personality with Alcohol Use Behaviour
and Alcohol Problems in Adolescents Receiving Child
Welfare Services
Sherry Heather Stewart &Melissa McGonnell &
Christine Wekerle &Ed Adlaf &
The MAP Longitudinal Study Research Team
Published online: 29 June 2011
#Springer Science+Business Media, LLC 2011
Abstract Four specific personality factors have been theorized to put adolescents at risk for
alcohol abuse: hopelessness (HOP), anxiety sensitivity (AS), sensation seeking (SS), and
impulsivity (IMP). We examined relations of these personality factors to various alcohol-related
indices in a sample at high risk for alcohol problems—specifically, a child welfare sample.
Adolescents (n=197; mean age= 16.8 years; 43% males) receiving services through Ontario
Child Protective Services participated. Personality was assessed with the Substance Use Risk
Profile Scale (SURPS). Alcohol-related outcomes were assessed with the Ontario Student
Drug Use and Health Survey (OSDUHS). Results showed that, consistent with theory, HOP,
SS, and IMP were all positively correlated with overall drinking levels and overall alcohol
problems on the OSDUHS. Unexpectedly, AS was negatively correlated with overall drinking
levels and was unrelated to the OSDUHS overall alcohol problems factor. Consistent with
hypothesis, HOP was related to an increased likelihood of receiving treatment for an alcohol
problem. But, despite greater drinking levels and alcohol problems, IMP was related to a
decreased likelihood of receiving alcohol treatment. In addition, SS and HOP were related to
earlier onset drinking. Additional exploratory analyses revealed that AS was positively
associated with difficulties stopping drinking and negatively associated with discussing such
difficulties with school personnel. Implications for treatment and prevention of heavy
drinking and alcohol problems in child welfare youth are discussed.
Keywords Patterns of consumption .Alcohol abuse etiology .Personality .Child welfare
Int J Ment Health Addiction (2011) 9:492–506
DOI 10.1007/s11469-011-9339-0
S. H. Stewart (*)
Departments of Psychiatry and Psychology, Dalhousie University, 1355 Oxford Street, Halifax, Nova
Scotia, Canada B3H 4J1
e-mail: sstewart@dal.ca
M. McGonnell
Department of Psychology, Dalhousie University, 1355 Oxford Street, Halifax, Nova Scotia, Canada
B3H 4J1
C. Wekerle :The MAP Longitudinal Study Research Team
McMaster University, Hamilton, Ontario, Canada
E. Adlaf
Centre for Addiction and Mental Health, Toronto, Ontario, Canada
Adolescence is well-recognized as a period when individuals are significantly at-risk for the
development of a variety of challenges, including difficulties with substance use and abuse.
A large body of research points to the fact that the prevalence of substance use in
adolescents is high and that the substance most commonly used by teens is alcohol (e.g.,
Rehm et al. 2005). Studies of the trajectory of alcohol use point to an association between
early use and long-term alcohol problems (e.g., Colder et al. 2002) as well as decreased
rates of college completion and higher rates of violence and criminal behaviour (Tucker
et al. 2003). While not all youth who demonstrate early problematic drinking continue
this pattern, many do and there are some populations of adolescents who seem to be more
at-risk for the development of alcohol disorders. Adolescents who have been exposed to
violence and abuse are more likely to start drinking at younger ages and to drink larger
quantities of alcohol (Hamburger et al. 2008) and adolescents from disrupted families
have more difficulty with binge drinking (Tucker et al. 2003).
It has been theorized that certain personality factors reflect increased susceptibility to
specific reinforcing properties of alcohol and other drugs (Conrod et al. 2000; Pihl and
Peterson 1995). More specifically, disturbances in particular brain motivation systems are
said to place certain individuals at greater risk for seeking out and experiencing
reinforcement from alcohol. Four specific dimensions of personality have each been
posited to be related to alcohol use/misuse (Pihl and Peterson 1995): Hopelessness
(HOP; pessimism about the self, world, and future; Abramson et al. 1989), Anxiety
Sensitivity (AS; fear of anxiety-related sensations; Reiss et al. 1986), Sensation Seeking
(SS; preference for novel and intense activities; Zuckerman 1994), and Impulsivity (IMP;
action with insufficient forethought; Dawe and Loxton 2004).
Theoretically, those high in HOP are said to be sensitive to the threat of punishment as a
result of a dysregulation within the endogenous opiate system. High HOP youth are
hypothesized to be at risk for using drugs that have analgesic effects, including alcohol,
with the motivation of alleviating or managing depression. In contrast, those with high
levels of AS are said to have deficits in the gamma-aminobutyric acid (GABA)
neurochemical pathways that result in dysfunction in the fear motivation system.
Theoretically, youth with high AS are prone to abuse drugs with anxiolytic properties,
including alcohol, with the motivation of alleviating or managing anxiety. Youth high in SS
are said to be more likely to use drugs, including alcohol, with effects on the dopamine
brain reward system due to an increased sensitivity within this brain motivational system.
Youth high in IMP are said to have deficits in the serotonin system, and consequent
difficulties with self regulation, which may lead to increased risk for use of drugs with
immediately reinforcing effects, including alcohol (Pihl and Peterson 1995).
Research with youth has generally borne out predictions derived from the Pihl and
Peterson (1995) theory. For example, HOP has been related to using alcohol to relieve
depression (Woicik et al. 2009) and AS has been shown to be related to use of alcohol to
manage anxiety (Stewart and Kushner 2001). Both AS and HOP have been related to
heavier drinking behaviour and to greater alcohol-related problems (Krank et al. 2010;
Stewart and Kushner 2001). SS has been related to heavier drinking and to increased risk
for adverse drinking consequences (Conrod et al. 1997; Schall et al. 1992; Woicik et al.
2009) while IMP has been related concurrently and prospectively to a variety of
problematic substance use behaviours including heavy drinking, as well as alcohol
problems (Krank et al. 2010; Woicik et al. 2009).
Nonetheless, these tests of the Pihl and Peterson (1995) model have been conducted
almost exclusively with youth drawn from the general population (e.g., Krank et al. 2010;
Woicik et al. 2009). It remains important to test the utility of this model in clinical samples
Int J Ment Health Addiction (2011) 9:492–506 493
and in samples of youth at particularly high risk for the development of alcohol problems as
a next step in this line of research. As noted, one population at high-risk for the
development of alcohol abuse is adolescents who have been exposed to violence or who
come from disrupted families. Many adolescents in these situations end up in the care of
Child Protective Services or Children’s Aid Societies (CAS’s). These societies have the
legal right and responsibility to investigate situations where children are being maltreated or
are at risk for being maltreated (Gough 2005).
In 2008 in Canada, child welfare agencies investigated 235,842 reports of child
maltreatment (i.e., abuse [violence, harm, or mistreatment] or neglect). In more than one-
third (85,440) of these investigations, the reports of maltreatment were substantiated
(Trocmé et al. 2010). Ontario, the most populous province in Canada, has 53 child welfare
agencies, most of which are referred to as CAS’s (Gough 2005).
In a recent investigation of children in three CAS’s in Ontario, Finlay et al. (2007) noted
that adolescents represented the largest proportion of children in care and that they were
presenting with increasingly higher rates of mental health issues and were particularly likely
to struggle with substance abuse, including alcohol abuse. In an effort to better understand
the variables that contribute to more positive outcomes for youth involved in the child
welfare system in Ontario, the Maltreatment and Adolescent Pathways (MAP) longitudinal
study was initiated (Wekerle et al. 2009). This project collected longitudinal information
about maltreatment and exposure to violence and, at follow-up points, about personality
and patterns of substance use (including alcohol use) in these youth.
Information from the MAP longitudinal study was used in the present study to examine
relationships among personality factors and patterns and consequences of alcohol use in a
high-risk child-welfare sample. We predicted that the four personality factors specified in
the Pihl and Peterson (1995) model would be related to drinking behaviour patterns and
alcohol problems based on previous results obtained from teens in the general population
(e.g., Krank et al. 2010; Woicik et al. 2009) and on theoretical prediction (Conrod et al.
2000; Pihl and Peterson 1995). Specifically, our first hypothesis was that all four personality
factors (AS, HOP, SS, and IMP) would be related to greater drinking behaviour. Our second
hypothesis was that all four personality factors would be related to greater alcohol
problems. Third, we hypothesized that all four personality factors would be related to an
increased probability of receiving treatment for alcohol problems, given their hypothesized
heavier drinking and greater levels of alcohol problems. Finally, we hypothesized that the
two externalizing personality factors (SS and IMP) would be related to earlier onset of
alcohol use (Cloninger et al. 1996).
Method
Participants
Participants for the current study were a subset of participants from the MAP longitudinal
study. The MAP longitudinal was designed to improve knowledge about child-welfare
involved youth as they make the transition through adolescence to adulthood. The years of
data collection for the MAP were from 2002 to 2010. The study included five assessment
points: initial assessment plus 6 month, 1-year, 1.5 years, and 2 years follow-ups.
Participants in the MAP longitudinal study were recruited through Children’s Aid Societies
(CAS) in a large urban centre in Ontario, Canada. They were randomly selected from active
CAS cases. Most MAP longitudinal study participants were involved with CAS for
494 Int J Ment Health Addiction (2011) 9:492–506
relatively long periods of time (i.e., 6 months or more) and may represent a population that
is at higher risk for a variety of negative outcomes. Participation was voluntary and
participants were compensated for their time and effort at minimum wage rates. A total
of 561 adolescents agreed to participate and completed the initial questionnaire package
designed to collect information on key MAP longitudinal study variables (i.e., child
maltreatment history, global distress, substance use, dating violence) at the baseline
assessment. The initial recruitment rate was 70% of all eligible youth. The 197
participants in the current study were drawn from among those who participated at the
one-year follow-up. These 197 participants ranged in age from 15 to 20 (M=16.8; SD=1.1)
years. There were slightly more females (57%) than males. The present sub-sample (n=197)
was generally consistent with the larger MAP sample (n=561). However, as time
proceeded across follow-up assessments in the MAP longitudinal study, more of the
retained sample was female and Crown Wards (i.e., youth who receive the highest level of
child welfare services, and for whom the government of Ontario is the parent, as parental
rights have been terminated). Of these 197 participants, 126 reported that they had used
alcohol in the past 12 months. These 126 past year drinkers ranged in age from 15 to 20
(M=16.9; SD= 1.2) years and were 57% female.
Informed Consent
Ethics approval was first obtained from all participating CAS agencies. Next, ethics
approval was obtained from research ethics boards at the participating universities. All
MAP research team members who accessed the data signed confidentiality agreements with
participating CAS agencies. Youth participants aged 16 and over provided informed consent
themselves; any youth less than 16 years of age, had their parents or legal guardians
provide informed consent. Across the life of this longitudinal study, whenever youth turned
16 in the MAP, they provided their own informed consent.
Measures
Substance Use Risk Profile Scale (SURPS) The SURPS (Woicik et al. 2009) is a 23-item
self-report scale designed to measure four dimensions of personality (HOP, AS, SS, and
IMP) which have been shown to be related to risk for alcohol use/misuse. Participants are
asked to rate the degree to which they agree with each item. Ratings range from 1 (strongly
disagree)to4(strongly agree). (See Table 1for a list of SURPS items organized by
subscale.) Since the SURPS scales vary from 5 to 7 items per scale, the theoretical range of
scores can vary from 5 to 25 (for the AS and IMP scales), from 5 to 30 (for the SS scale),
and from 5 to 35 (for the HOP scale). The SURPS has been shown to have a stable four
factor structure and the four scales have been found to have good internal consistency
and good concurrent (Woicik et al. 2009) and predictive (Krank et al. 2010) validity in
non-clinical samples of adolescents recruited through Canadian schools.
Ontario Student Drug Use and Health Survey (OSDUHS) The OSDUHS (Adlaf and
Paglia-Boak 2007) was created at the Centre for Addiction and Mental Health (CAMH) in
Toronto, Canada, for use in an ongoing epidemiological study of substance use, mental
health, school and family involvement, medication, delinquency, and other issues in Ontario
youth. The OSDUHS has been conducted every 2 years since 1977 and is the longest
ongoing school survey in Canada (Rehm et al. 2005). We chose 20 items from the larger
OSDUHS survey for use in this study. These items were specifically related to alcohol and
Int J Ment Health Addiction (2011) 9:492–506 495
included frequency and quantity of drinking and binge drinking behaviour (alcohol use;
8 items), negative consequences related to alcohol use (alcohol problems; 10 items),
receiving treatment for alcohol problems (1 item), and age of onset of alcohol use (1 item).
Table 1 Factor pattern matrix for principal components analyses with varimax rotation on Substance Use
Risk Profile Scale (SURPS) items (n=157)
SURPS item and intended subscale Factors
Hopelessness Anxiety
sensitivity
Sensation
seeking
Impulsivity Communalities
Hopelessness subscale
I am content (R) −.694 −.037 .056 −.061 .490
I am happy (R) −.790 −.090 −.131 −.018 .650
I have faith that my future holds great promise (R) −.726 −.010 .071 −.065 .536
I feel proud of my accomplishments (R) −.696 −.053 .082 .020 .494
I feel that I’m a failure .523 .400 .124 .153 .472
I feel pleasant (R) −.711 .049 −.028 .032 .510
I am very enthusiastic about my future (R) −.687 .177 .117 −.164 .544
Anxiety sensitivity subscale
It’s frightening to feel dizzy or faint −.092 .650 −.111 .046 .445
It frightens me when I feel my heart beat change −.033 .647 −.144 .030 .442
I get scared when I’m too nervous −.086 .612 −.291 .227 .518
I get scared when I experience unusual body
sensations
.027 .695 .082 −.037 .491
It scares me when I’m unable to focus on a task .168 .726 −.029 .088 .563
Sensation seeking subscale
I would like to skydive .018 −.117 .758 −.151 .611
I enjoy new and exciting experiences even if they
are unusual
−.366 −.133 .474 .206 .419
I like doing things that frighten me a little −.016 .000 .678 .098 .470
I would like to learn how to drive a motorcycle .006 −.100 .686 .053 .484
I am interested in experience for it’s own sake even
if it is illegal
.246 .204 .418 .404 .440
I would enjoy hiking long distances in wild and
uninhabited territory
−.083 −.030 .608 −.146 .399
Impulsivity subscale
I often don’t think things through before I speak −.054 .023 −.110 .797 .652
I often involve myself in situations that I later
regret being involved in
.288 .138 −.122 .671 .568
I usually act without stopping to think .010 .027 −.060 .803 .650
Generally, I am an impulsive person −.006 .128 .211 .534 .346
I feel I have to be manipulative to get what I want .254 .499 .236 .343 .486
Percent variance (prior to rotation) 18.740 12.592 11.329 8.125
Scale internal consistency (coefficient alpha) .82 .72 .70 .73
First seven eigenvalues= 4.310, 2.869, 2.606, 1.869, 1.116, 1.042, 0.869. (R) = Items that should be reversed
scored in scoring of the SURPS subscales. Note that raw scores (i.e., no reverse scoring) were used in the
principal components analysis; reverse scoring was employed prior to calculation of alpha. Rotated factor
loadings greater than .4 shown in bold. The factor analyses used principle components extraction and
Varimax rotation. Items are grouped according to intended subscale and factors are labelled according to the
content of the highest loading factors. Hopeless factor loadings are multiplied by −1.000 since weightings in
principal components analysis are arbitrary; this facilitated interpretation of high factor scores as representing
high levels of HOP; the same was done with HOP factor scores used in subsequent analyses
496 Int J Ment Health Addiction (2011) 9:492–506
Table 2contains all of the OSDUHS items pertaining to alcohol use behaviour and Table 3
all of the OSDUHS items pertaining to alcohol problems. Table 4specifies the content of
the alcohol treatment history and age of onset of alcohol use items.
Data Analysis
We first conducted factor analyses of the independent variable (i.e., personality scores on
the SURPS) and multi-item dependent measures (i.e., alcohol use behaviour; alcohol
problems) to create factor scores on the variables of interest. Factor scores, as opposed to
subscale scores, allow for greater weighting to be given to items that better assess the
construct(s) of interest in the calculation of variables. In all three analyses, we employed an
exploratory factor analytic [EFA] approach (as opposed to a confirmatory factor analytic
[CFA] approach) since the SURPS and OSDUHS have never been used in a high risk child
welfare sample before. EFA is the factor analytic method of choice when first using a
measure in a new population (e.g., Floyd and Widaman 1995; Van Prooijen and Van der
Kloot 2001). In all three cases, an EFA was conducted on relevant item scores using
principle components analysis (PCA) extraction (see Woicik et al. 2009). For the SURPS,
where a multi-component solution was expected, a Varimax rotation was used which was
consistent with the hypothesized orthogonal structure of the measure (Woicik et al. 2009).
Moreover, an orthogonal rotation allowed for the creation of personality factor scores that
were uncorrelated with one another which permitted examination of the relatively pure
associations of each personality factor with the alcohol-related criterion variables in
hypothesis testing. For the two EFAs on relevant OSDUHS items, in each case, a single
factor was extracted (to represent global alcohol use levels and global alcohol problems,
Table 2 Component matrix for principal components analyses on Ontario Student Drug Use and Health
Survey (OSDUHS) alcohol use items (n=123)
OSDUHS alcohol use item Factor
loading
Communalities
How often do you have five or more drinks on one occasion? .911 .831
How many drinks containing alcohol do you have on a typical day
when you are drinking?
.827 .684
On average, how much hard liquor (for example rum, whiskey, vodka, coolers)
do you usually drink at one time?
.822 .675
In the last 12 months, how often did you drink alcohol such as liquor
(rum, whiskey, etc.), wine, beer, coolers?
.804 .647
During the last 4 weeks, how often did you drink alcohol
(liquor, wine, beer, or coolers)?
.733 .537
On average, how much beer do you usually drink at any one time? .721 .520
On average, how much wine do you usually drink at any one time? .417 .173
How many times in the last 4 weeks have you had five or more
drinks of alcohol on the same occasion?
.318 .101
Percent variance 52.100
Scale internal consistency (coefficient alpha) .82
First three eigenvalues= 4.168, 1.018, 0.958. Factor loadings greater than .4 shown in bold. The factor
analyses used principle components extraction. Only past year drinkers with no missing values on any of the
OSDUHS alcohol use items (n=123) were included in this analysis
Int J Ment Health Addiction (2011) 9:492–506 497
Table 3 Component matrix for principal components analyses on Ontario Student Drug Use and Health
Survey (OSDUHS) alcohol problem items (n=120)
OSDUHS alcohol problem item Factor
loading
Communalities
How often during the last 12 months have you not done things you were
supposed to because of drinking?
.711 .505
How often during the last 12 months have you been unable to remember
what happened the night before because you had been drinking?
.663 .439
How often during the last 12 months have you needed a first drink in the
morning to get yourself going after a heavy drinking session?
.657 .432
Has a relative or friend or a doctor or other health care worker been
concerned about your drinking or suggested you cut down?
.635 .403
Have you or someone else been injured as a result of your drinking? .628 .394
How often during the last 12 months have you found that you were not
able to stop drinking once you had started?
.549 .301
How often during the last 12 months have you had a feeling of guilt
or remorse after drinking?
.430 .185
Have you ever talked to a school counsellor, school nurse or teacher
because you had a problem as a result of your use of alcohol?
.394 .155
Have you ever been warned by the police because of your use of alcohol? .374 .140
Have you ever seen a doctor or been in a hospital because you had
been drinking alcohol?
.374 .140
Percent variance 30.922
Scale internal consistency (coefficient alpha) .72
First four eigenvalues= 3.098, 1.288, 1,032, 0.948. Factor loadings greater than .4 are shown in bold. The
factor analyses used principle components extraction. Only past year drinkers with no missing values on any
of the OSDUHS alcohol problems items (n=120) were included in this analysis
Table 4 Correlations of Substance Use Risk Profile Scale (SURPS) personality factors with the alcohol-
related criterion variables from the Ontario Student Drug Use and Health Survey (OSDUHS)
OSDUHS score NSURPS personality factor score
Hopelessness Anxiety
sensitivity
Sensation
seeking
Impulsivity
Alcohol use levels:
Factor score from PCA in Table 297 .168* −.241** .430** .248**
Alcohol problems:
Factor score from PCA in Table 395 .186* −.021 .249** .309**
Alcohol treatment history item:
Have you been in a treatment program
because of your alcohol use?
99 .177* −.002 .008 −.169*
Age of drinking onset item:
When, if ever, did you first drink alcohol? 105 −.203* .135 −.283** −.043
**p<.01; *p< .05 (one-tailed tests); SURPS scores are factor scores (not subscale scores). The n’s vary
because of missing values on either the SURPS or OSDUHS. Effect sizes of the correlation coefficients
determined as follows: small, r=0.1–0.23; medium, r=0.24–0.36; large, r= 0.37 or larger (Cohen 1988,1992)
498 Int J Ment Health Addiction (2011) 9:492–506
respectively) and factor scores were retained for later use in hypothesis testing. For the PCA
of SURPS items, we used scores for the entire sample (n= 197) in this analysis as there was
no reason to restrict the sample to past year drinkers. For the PCAs of OSDUHS items, we
restricted the sample to the past year drinkers (n=126).
Lastly, for hypothesis testing, we correlated the four personality factor scores (HOP, AS,
SS, and IMP) with the four drinking-related dependent measures (i.e., factor scores on the
drinking levels and alcohol problems factors, respectively, and item scores on the age of
drinking onset and alcohol treatment history, respectively, all from the OSDHUS).
Again, for this analysis, we restricted the sample to past year drinkers (n=126). The
statistical significance of the correlations was examined using one-tailed tests given that
directional predictions had been made a priori. The clinical significance of the findings
was inferred by examining the effect sizes of the correlations using criteria recommended
by Cohen (1988,1992).
Results
Principal Components Analysis of the SURPS
While six factors had eigenvalues greater than 1.00 (see bottom of Table 1), Kaiser’s(1961)
eigenvalue >1 criterion for determining the number of factors to retain has been criticized
for resulting in factor over-extraction (Longman et al. 1989). Thus, we used the more
stringent parallel analysis (Longman et al. 1989) for determining the number of components
to retain. Comparison of obtained eigenvalues with: (a) mean eigenvalues and (b) 95th
percentile eigenvalues, both indicated support for retaining four components. Thus, a
four factor solution was extracted, consistent with the four personality risk factors
obtained in previous research with non-clinical adolescents recruited through the school
system (Krank et al. 2010;Woiciketal.2009). In the present study, the four factors
together accounted for a total of 50.8% of the variance in SURPS item scores.
Examination of item content indicated that the first factor corresponded to HOP, the
second to AS, the third to SS, and the last to IMP.
With salient factor loadings defined as ≥.400 (see Woicik et al. 2009), all but one
item showed its highest salient loading on its hypothesized factor. There were only two
complex loadings (i.e., items with salient loadings on more than one factor), no
hyperplaneitems(i.e.,itemsfailingtoloadonanyfactor),andanadequatenumberof
salient loadings per factor (i.e., 5+ salient loadings per factor). Thus, the four-factor
solution showed adequate ‘simple structure’(Thurstone 1947). The communalities
indicated that the four factors together explained from 34.6% to 65.2% of the variance
in item scores, across SURPS items. The internal consistency values (coefficient alphas)
for the four scales were acceptable to good (alphas ranged from 0.70 to 0.82; see Table 1).
Principal Components Analysis of the OSDUHS Alcohol Use Behaviour Items
While two factors had eigenvalues greater than 1.00 (see bottom of Table 2), comparison of
obtained eigenvalues with: (a) mean eigenvalues and (b) 95th percentile eigenvalues, using
the more stringent parallel analysis (Longman et al. 1989), both indicated support for
retaining a single component. Thus, a one factor solution was extracted, consistent with a
global alcohol use levels variable. In the present study, the single factor accounted for a
total of 52.1% of the variance in OSDUHS alcohol use behaviour item scores.
Int J Ment Health Addiction (2011) 9:492–506 499
With salient factor loadings defined as ≥.400, all but one item showed a salient loading
on this global alcohol use levels factor. Given that there was only one hyperplane item, and
an adequate number of salient loadings on this single global factor (i.e., 7 salient loadings),
the single-factor solution can be said to display adequate ‘simple structure’(Thurstone
1947). The communalities indicated that the single factor explained from 10.1% to 83.1%
of the variance in item scores, across OSDUHS alcohol use behaviour items. The internal
consistency value for the resultant scale was good (coefficient alpha=0.82; see Table 2).
Principal Components Analysis of the OSDUHS Alcohol Problems Items
While three factors had eigenvalues greater than 1.00 (see bottom of Table 3),
comparison of obtained eigenvalues with: (a) mean eigenvalues and (b) 95th percentile
eigenvalues, using the more stringent parallel analysis (Longman et al. 1989), both
indicated support for retaining a single component. Thus, a one factor solution was
extracted, consistent with a global alcohol problems variable. In the present study, the
single factor accounted for a total of about 30.9% of the variance in OSDUHS alcohol use
behaviour item scores.
With salient factor loadings defined as ≥.400, all but three items showed a salient
loading on this global alcohol use levels factor. Given that there were only three
hyperplane items, and an adequate number of salient loadings on this single global
factor (i.e., 7 salient loadings), the single-factor solution can be said to display adequate
‘simple structure’(Thurstone 1947). The communalities indicated that the single factor
explained from 14.0% to 50.5% of the variance in item scores, across OSDUHS alcohol
problem items. The internal consistency value for the resultant scale was acceptable
(coefficient alpha= 0.72; see Table 3).
Association of Personality Factors with Alcohol Outcomes
Drinking Levels Scores on the alcohol use levels factor from the OSDUHS were correlated
with the four personality factor scores from the SURPS. Consistent with the first
hypothesis, HOP, SS, and IMP were all significantly positively correlated with overall
alcohol use levels (see Table 4), suggesting that child welfare adolescents with higher levels
of these traits were using alcohol more frequently and heavily than others. In direct contrast
with the first hypothesis, AS was significantly negatively correlated with overall alcohol
use levels (see Table 4), suggesting that child welfare adolescents with higher levels of AS
were using alcohol less frequently and heavily than others. The correlation with HOP was
small in magnitude, those with AS and IMP were moderate in magnitude, and the
correlation with SS was large (Cohen 1988,1992).
Drinking Problems Scores on the alcohol problems factor from the OSDUHS were
correlated with the four personality factor scores from the SURPS. Consistent with the
second hypothesis, HOP, SS, and IMP were all significantly positively correlated with
overall alcohol problems (see Table 4), suggesting that child welfare adolescents with
higher levels of these traits were experiencing more severe alcohol-related problems than
others. In contrast with the second hypothesis, AS was not significantly correlated with
overall alcohol problems (see Table 4), suggesting that child welfare adolescents with
higher levels of AS were experiencing similar severity of alcohol problems to others. The
correlation with HOP was small in magnitude whereas the correlations with SS and IMP
were moderate (Cohen 1988,1992).
500 Int J Ment Health Addiction (2011) 9:492–506
Given the failure to support the second hypothesis regarding a positive association
between AS levels and overall alcohol problems, additional exploratory analyses were
conducted to examine the associations of AS levels to each of the ten specific alcohol
problem items on the OSDUHS. These correlations revealed that, consistent with the
second hypothesis, AS factor scores were significantly positively correlated with difficulties
stopping drinking once started (r(98)=.213, p< .05); this correlation was small in
magnitude (Cohen 1988,1992). However, in direct contrast to the second hypothesis, AS
factor scores were significantly negatively correlated with having discussed a drinking
problem with school personnel (r(99)=−.234, p= .01); this correlation was moderate in
magnitude (Cohen 1988,1992). None of the other correlations of the AS factor with
individual OSDUHS alcohol problems items proved statistically significant.
Alcohol Treatment Participants were asked whether they had ever received treatment for
an alcohol problem. Scores on this alcohol treatment history item from the OSDUHS
(see Table 4) were correlated with each of the SURPS factor scores. Consistent with the
third hypothesis, HOP factor scores were positively correlated with alcohol treatment
history (see Table 4), indicating that child welfare youth with higher levels of HOP were
more likely to have received treatment for an alcohol problem than others. In direct
contrast to the third hypothesis, IMP was significantly negatively correlated with alcohol
treatment history (see Table 4), indicating that more IMP child welfare youth were less
likely than others to have received treatment for an alcohol problem despite their higher
levels of alcohol problems as indicated in the test of the second hypothesis. Contrary to
the third hypothesis, AS and SS were unrelated to history of treatment for alcohol use
problems (see Table 4). The correlations with HOP and IMP were both small in
magnitude (Cohen 1988,1992).
Age of Onset Participants were asked in what grade they had first used alcohol. Scores on
this age of onset of alcohol use item from the OSDUHS (see Table 4) were correlated with
each of the SURPS factor scores. Consistent with the fourth hypothesis, SS factor scores
were negatively correlated with age of onset of alcohol use (see Table 4), indicating that
child welfare youth with higher levels of SS started drinking at an earlier age than
others. Contrary to the fourth hypothesis, IMP was unrelated to age of onset of alcohol
use(seeTable4). An unexpected finding was that HOP factor scores were also
significantly negatively correlated with age of onset of alcohol use (see Table 4),
indicating that more HOP child welfare youth stated drinking at an earlier age than others.
The correlation with HOP was small in magnitude whereas the correlation with SS was
moderate (Cohen 1988,1992).
Discussion
In past research, the four personality risk factors of HOP, AS, SS, and IMP specified in Pihl
and Peterson’s(1995) theoretical model of alcohol abuse risk were related to alcohol
outcomes in theoretically expected ways in non-selected samples of adolescents recruited
through the schools (Krank et al. 2010; Woicik et al. 2009). The present study sought to
determine if these prior findings were generalizable to a high-risk sample of adolescents
recruited through child welfare services. Consistent with theoretical expectation (e.g.,
Conrod et al. 2000) and with prior findings with general population youth (Krank et al.
2010; Woicik et al. 2009), each of the four personality factors was related to one or more of
Int J Ment Health Addiction (2011) 9:492–506 501
the alcohol outcomes assessed on the OSDUHS. We look at each personality factor in turn
and its relations to the alcohol outcomes in the present child welfare sample.
Individuals with higher HOP levels were found to be earlier onset drinkers. This earlier
onset of drinking behaviour was contrary to expectations that it would be the externalizing
personality characteristics only (SS and IMP) that would be associated with earlier onset
drinking (Cloninger et al. 1996). The finding of an earlier onset of drinking in HOP
adolescents in child welfare has serious clinical implications in that it suggests that this
group may be at particular risk for the negative outcomes associated with earlier onset
alcohol use (Colder et al. 2002). Moreover, higher levels of HOP were associated with more
frequent and heavier drinking behaviour as well as with more severe alcohol problems. The
relationship between high levels of HOP and greater drinking behaviour has been found
previously in non-clinical adolescents (Krank et al. 2010; Woicik et al. 2009). Similarly, the
relationship between high levels of HOP and problem drinking has been found previously
in non-clinical adolescents (e.g., Krank et al. 2010). Thus, the present study replicated these
prior findings among child welfare youth. Consistent with hypothesis, higher HOP
individuals in the present study were more likely to have received treatment for an alcohol
problem, which is consistent with their heavier drinking and more severe alcohol problems.
Overall, HOP youth appear a very high risk group within the child welfare sample given the
associations of HOP with heavier alcohol use, more alcohol problems, greater rates of
alcohol treatment, and earlier onset drinking. Nonetheless, despite their statistical
significance, all correlations of HOP with the alcohol outcomes were small in magnitude.
In contrast with theoretical prediction (Pihl and Peterson 1995), higher levels of AS were
associated with less alcohol use behaviour overall. While inconsistent with prior findings in
young adults showing a positive association between AS and alcohol use (e.g., Stewart et
al. 1995), this finding is actually consistent with prior findings in non-clinical adolescents
(Krank et al. 2010; Woicik et al. 2009) suggesting that AS may be protective from heavier
drinking in the teenage years. Unexpectedly, based on theory (Pihl and Peterson 1995) and
prior findings (Woicik et al. 2009), AS was unrelated to the overall index of alcohol
problems. However, additional exploratory analyses indicated that high levels of AS were
associated with greater difficulties stopping drinking once started. It seems then that these
individuals may drink less overall, but when they do drink they find that it is difficult to
control their intake (possibly due to their increased sensitivity to the anxiolytic properties of
alcohol; MacDonald et al. 2000). It may appear counter-intuitive that AS is related to lower
consumption and yet also to greater difficulties controlling intake. Work by Cooper (1994)
on adolescents’reasons for drinking may help clarify this pattern. Specifically, adolescent
coping motives (a reason for drinking closely associated with AS; Comeau et al. 2001) have
been shown to be associated with increased drinking problems even after controlling
consumption levels. Because AS teens drink to cope with negative emotions, despite their
lighter consumption behavior, they may be more likely to develop a psychological
dependence on alcohol to cope, which, in turn, could result in difficulties controlling their
alcohol intake. Individuals with higher AS in the present study also indicated that they were
less likely to seek help from school personnel. This could simply be the result of the fact
that, at least at this point in their lives, their drinking is not causing them great difficulty
and, therefore, they do not need assistance. Alternatively, they may avoid help-seeking due
to social sensitivities (Stewart et al. 1997) about being judged negatively by school
personnel should they reveal concerns about their alcohol problems (e.g., their loss-of-
control drinking). High AS individuals may then, drink less to avoid negative consequences
(hangover symptoms, social censure); however, their difficulty controlling their drinking
and their reluctance to seek help may point to potential for future problems with alcohol.
502 Int J Ment Health Addiction (2011) 9:492–506
As expected, high SS was associated with earlier onset of drinking (Cloninger et al.
1996). Interestingly, of the two externalizing personality characteristics, SS (and not IMP)
was related to earlier onset drinking. This pattern points to the importance of separating SS
from IMP when assessing personality risk for alcohol misuse, since only SS appears related
to early onset drinking. Also, consistent with hypothesis, SS was related to overall greater
drinking behaviour—an association of large magnitude (Cohen 1988,1992). This pattern of
association of SS with higher quantity and frequency of drinking has previously been
observed in studies using non child welfare samples (Conrod et al. 1997,2006; Schall et al.
1992; Woicik et al. 2009). In the present study, SS was also positively associated with the
overall index of drinking-related problems (see also Conrod et al. 1997; Schall et al. 1992).
But despite their earlier onset drinking, their greater drinking behaviour, and their more
severe alcohol problems, high SS child welfare youth were no more likely than others to
receive treatment for an alcohol use problem.
High IMP demonstrated a pattern of association with drinking behaviours that was
similar in some respects, and different in others, to the pattern displayed by high SS youth.
Like SS, IMP was significantly positively correlated both with alcohol use behaviour and
drinking-related problems—both associations of moderate magnitude. Previous research
(Krank et al. 2010; Woicik et al. 2009) has also noted an association between IMP and
drinking frequency as well as negative consequences of drinking. Unlike high SS, high IMP
was not associated with early onset alcohol use in the present study. Also unlike high SS
youth, high IMP youth in the child welfare sample were less likely than others to be
receiving treatment for a drinking problem. This lower probability of receiving treatment is
concerning given the significant association of IMP to both heavier drinking and more
drinking related problems. It is not clear whether IMP youth’s alcohol problems are going
unrecognized, or if they are being appropriately referred for treatment but, possibly due to
their difficulties with planning and organization, higher IMP youth are having greater
difficulty following through on recommendations for alcohol treatment.
Overall, our results indicate that the major predictions of the Pihl and Peterson (1995)
model of personality risk for alcohol abuse were supported among a high risk sample of
adolescents involved with the Ontario child welfare system. In terms of clinical
implications, these finding suggest that the SURPS might be useful in determining who
is likely to experience problematic drinking behaviours and as a means of determining
appropriate interventions for adolescents involved with child welfare. These potential uses
are particularly important because the SURPS provides information that is functionally
independent of current alcohol use (see Woicik et al. 2009). This is theoretically important
because of the potential for the SURPS to determine likely involvement with alcohol. It
also has practical value, however, in that it means that the SURPS can be used to define
subsets of populations which are at high-risk for alcohol abuse without requiring
individuals to admit to alcohol use. This means that this tool could be used to identify
adolescents before they begin to use alcohol and adolescents who are either unaware that
their use is problematic or who are reluctant to admit that they have difficulties with alcohol
use. These adolescents could then be encouraged to participate in intervention activities
which can also be targeted to specific high risk personality factors (Conrod et al. 2006).
Given the associations between all four SURPS personality factors and frequency and
quantity of drinking and/or overall or specific drinking-related problems, it seems that
intervention is warranted for child welfare adolescents with elevations on any of these four
personality risk factors.
Typically, alcohol intervention programs for youth rely on education about the potential
consequences of using alcohol and/or on the development of skills to increase adolescents’
Int J Ment Health Addiction (2011) 9:492–506 503
ability to resist peer pressure to drink. These programs have shown some benefit, but
generally, these programs are only moderately effective at reducing problematic alcohol use
in adolescents (e.g., Wagner et al. 1999). More recently, attention has been given to the
possibility of matching treatment to individual characteristics related to alcohol misuse
(Conrod et al. 2000; Stewart and Conrod 2008). The recognition of the importance of
personality factors in alcohol abuse patterns has lead to the development of cognitive-
behavioural interventions which target specific personality risk variables in adolescents
(see Watt et al. 2008). When matched to personality profiles, these interventions have
been shown to decrease drinking quantity and frequency, as well as binge drinking rates
(Conrod et al. 2006), at least in non-clinical adolescents recruited through the school system.
The importance of early recognition of alcohol abuse and of intervening before risky
drinking becomes a stable behavioural pattern is increasingly recognized (Stewart et al.
2005). One method for assessing potential for problematic alcohol use is screening for
personality risk-factors. A decision to screen high-risk adolescent populations in this way
should be predicated not only on theory but on direct information about the relevance of the
theory to the population in question. This study has shown that the Pihl and Peterson (1995)
personality risk model appears to be a useful model upon which to base future clinical
interventions for treating or preventing alcohol problems among child welfare youth.
There are several limitations to this study. First, there may be generalizability issues with
the present results. Our participants were only recruited in Ontario. Therefore, the present
results may not generalize to adolescents involved with child welfare in other parts of
Canada or other parts of the world. Participation in this study was also restricted to children
involved in the child welfare system for long periods (i.e., more than 6 months). The
sample may, therefore, not be representative of all adolescents in the child welfare system.
In addition, in the MAP study, as time went on, more of the retained sample was female and
Crown Wards, suggesting that the present sample may not be entirely representative even of
child welfare youth in Ontario. That said, the findings of this study, on the associations of
the four personality variables (Pihl and Peterson 1995) to various alcohol outcomes, have
also been established in samples of adolescents recruited through the school system across
Canada (Krank et al. 2010; Woicik et al. 2009). Thus, the present findings extend these
prior results with non-clinical youth to high risk youth recruited through the CAS’s active
caseload.
A second potential limitation concerns the fact that all variables were assessed via youth
self-report. This requires youth to report truthfully about their personality and their alcohol
involvement, and to be sufficiently aware of these characteristics in order to report
accurately. While research does suggest that alcohol self-reports are highly accurate
when respondents are assured of confidentiality (Sobell and Sobell 1990), future
research might nonetheless consider including caregiver reports to supplement and
validate youth self-reports.
A third potential limitation pertains to the cross-sectional nature of the present findings.
While Pihl and Peterson’s(1995) model posits personality as a factor that increases risk for
alcohol use/misuse and shapes risk for alcohol problems, the present findings cannot
address either temporality or causality. For example, it is possible that heavier drinking
increases IMP. Future research should utilize a longitudinal methodology in order to test
whether the four personality factors from the Pihl and Peterson (1995) model predict
increases in alcohol use or alcohol problems over time in child welfare youth (see
Krank et al. 2010, for a study that demonstrates this in youth from the general population).
A final possible limitation refers to the clinical significance of the findings. Most of the
observed statistically significant relations between personality and the alcohol outcomes
504 Int J Ment Health Addiction (2011) 9:492–506
were small to moderate in magnitude. One exception was the large magnitude relation
between SS and alcohol use levels. This suggests that other variables in addition to
personality (e.g., parental history of alcohol use disorders; maltreatment history) need to
be considered in determining risk for heavy drinking and alcohol problems in child
welfare youth.
Acknowledgements The MAP Longitudinal Study was supported by research funds from the Canadian
Institutes of Health Research (CIHR), Institute of Gender and Health; Ontario Ministry of Child and Youth
Services; and the Provincial Centre of Excellence in Child and Youth Mental Health at the Children’s
Hospital of Eastern Ontario. The MAP Research Team: Principal Investigator—C. Wekerle; Co-Investigators—
M. Boyle, D. Goodman, B. Leslie, E. Leung, H. MacMillan, B. Moody, N. Trocmé, & R. Waechter. We thank the
MAP youth, MAP youth caseworkers, MAP community advisory team, and supports to Dr. Wekerle
(CIHR/Ontario Women’s Health Council Mid-Career award) and policy position (PHAC Interchange
Canada Assignment, 2009–2010). Dr. Stewart was supported through a Killam Research Professorship
from the Faculty of Science at Dalhousie University and Meli ssa McGonnell was supported through a
doctoral fellowship from the Social Sciences and Humanities Research Council of Canada, at the time
this research was conducted. An earlier version of this paper was presented at the 34th Annual Alcohol
Epidemiology Symposium of the Kettil Bruun Society (KBS) in Victoria, Canada, June, 2008. We thank
Dr. Emmanuel Kuntsche for his helpful comments as discussant on this paper at the KBS meeting.
Melissa McGonnell conducted this study as a comprehensive requirement toward the PhD in clinical
psychology degree at Dalhousie University, under the supervision of Dr. Stewart.
References
Abramson, L. Y., Metalsky, G. I., & Alloy, L. B. (1989). Hopelessness depression: a theory-based subtype of
depression. Psychological Review, 96, 358–372.
Adlaf, E. M., & Paglia-Boak, A. (2007). The Ontario Student Drug Use and Health Survey (OSDUHS).
Toronto: Centre for Addiction and Mental Health.
Cloninger, C., Sigvardsson, S., & Bohman, M. (1996). Type I and type II alcoholism: an update. Alcohol
Health and Research World, 20,18–23.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Mahwah: Erlbaum.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Colder, C. R., Richardson, J. L., Campbell, R. T., Ruel, E., & Flay, B. R. (2002). A finite mixture model of
growth trajectories of adolescent alcohol use: predictors and consequences. Journal of Consulting and
Clinical Psychology, 70, 976–985.
Comeau, N., Stewart, S. H., & Loba, P. (2001). The relations of trait anxiety, anxiety sensitivity, and sensation
seeking to adolescents’motivations for alcohol, cigarette, and marijuana use. Addictive Behaviors, 26,
803–825.
Conrod, P. J., Peterson, J. B., & Pihl, R. O. (1997). Disinhibited personality and sensitivity to alcohol
reinforcements: independent predictors of drinking behavior. Alcoholism: Clinical and Experimental
Research, 21, 1320–1332.
Conrod, P. J., Pihl, R. O., Stewart, S. H., & Dongier, M. (2000). Validation of a system of classifying female
substance abusers based on personality and motivational risk factors for substance abuse. Psychology of
Addictive Behaviors, 14, 243–256.
Conrod, P. J., Stewart, S. H., Comeau, M. N., & Maclean, A. M. (2006). Efficacy of cognitive-behavioral
interventions targeting personality risk factors for youth alcohol misuse. Journal of Clinical Child and
Adolescent Psychology, 35, 550–563.
Cooper, M. L. (1994). Motivations for alcohol use among adolescents: development and validation of a four-
factor model. Psychological Assessment, 6,117–128.
Dawe, S., & Loxton, N. J. (2004). The role of impulsivity in the development of substance use and eating
disorders. Neuroscience and Biobehavioral Reviews, 28, 343–351.
Finlay, J., Greco, K., Kerr, J., Erbland, J., & Cooke, D. (2007). We are your sons and daughters: The Child
Advocate’s Report on the Quality of Care of 3 Children’s Aid Societies. Office of Child and Family
Service Advocacy, Government of Ontario.
Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of clinical
assessment instruments. Psychological Assessment, 7(3), 286–299.
Int J Ment Health Addiction (2011) 9:492–506 505
Gough, P. (2005). Ontario’s child welfare system. CECW Information Sheet #31E. Toronto, ON: University of
Toronto, Faculty of Social Work. Retrieved August 11, 2008, from www.cecw-cepb.ca/DocsEng/
OntChildWelfareSystem31E.pdf.
Hamburger, M. E., Leeb, R. T., & Swahn, M. H. (2008). Child maltreatment and early alcohol use among
high-risk adolescents. Journal of Studies on Alcohol and Drugs, 69, 291–295.
Kaiser, H. F. (1961). A note on Guttman’s lower bound for the number of common factors. Multivariate
Behavioral Research, 1, 249–276.
Krank, M., Stewart, S. H., O’Connor, R., Woicik, P. B., Wall, A.-M., & Conrod, P. B. (2010). Structural,
concurrent, and predictive validity of the Substance Use Risk Profile Scale in early adolescence.
Addictive Behaviors, 36,37–46.
Longman, R. S., Cota, A. A., Holden, R. R., & Fekken, G. C. (1989). A regression equation for the parallel
analysis criterion in principal components analysis: mean and 95th percentile eigenvalues. Multivariate
Behavioral Research, 24,59–69.
MacDonald, A. B., Baker, J. M., Stewart, S. H., & Skinner, M. (2000). Effects of alcohol on the response to
hyperventilation of participants high and low in anxiety sensitivity. Alcoholism: Clinical and
Experimental Research, 24, 1656–1665.
Pihl, R. O., & Peterson, J. B. (1995). Alcoholism: the role of different motivational systems. Journal of
Psychiatry and Neuroscience, 20, 372–396.
Rehm, J., Monga, N., Adlaf, E., Taylor, B., Bondy, S. J., & Fallu, J.-S. (2005). School matters: drinking
dimensions and their effects on alcohol-related problems among Ontario secondary school students.
Alcohol and Alcoholism, 40, 569–574.
Reiss, S., Peterson, R. A., Gursky, D. M., & McNally, R. J. (1986). Anxiety sensitivity, anxiety frequency and
the predictions of fearfulness. Behaviour Research and Therapy, 24,1–8.
Schall, M., Kemeny, A., & Maltzman, I. (1992). Factors associated with alcohol use in university students.
Journal of Studies on Alcohol, 53, 122–136.
Sobell, L. C., & Sobell, M. B. (1990). Self-report issues in alcohol abuse: state of the art and future
directions. Behavioral Assessment, 12,77–90.
Stewart, S. H., & Conrod, P. J. (Eds.) (2008). Anxiety and substance use disorders: The Vicious cycle of
comorbidity. New York: Springer.
Stewart, S. H., & Kushner, M. G. (2001). Introduction to the special issues on ‘Anxiety sensitivity and
addictive behaviours’.Addictive Behaviors, 26, 775–785.
Stewart, S. H., Peterson, J. B., & Pihl, R. O. (1995). Anxiety sensitivity and self-reported alcohol
consumption rates in university women. Journal of Anxiety Disorders, 9, 283–292.
Stewart, S. H., Taylor, S., & Baker, J. M. (1997). Gender differences in dimensions of anxiety sensitivity.
Journal of Anxiety Disorders, 11, 179–200.
Stewart, S. H., Conrod, P. J., Marlatt, G. A., Comeau, M. N., Thush, C., & Krank, M. (2005). New
developments in prevention and early intervention for alcohol abuse in youth. Alcoholism: Clinical and
Experimental Research, 29, 278–286.
Thurstone, L. L. (1947). Multiple factor analysis. Chicago: The University of Chicago Press.
Trocmé, N., Fallon, B., MacLaurin, B., Sinha, V., Black, T., Fast, E., et al. (2010). Canadian incidence study of
reported child abuse and neglect—2008: Major Findings. Ottawa, Canada: Public Health Agency of
Canada. Retrieved May 24, 2011, from http://www.phac-aspc.gc.ca/ncfv-cnivf/pdfs/nfnts-cis-2008-rprt-eng.
pdf.
Tucker, J. S., Orlando, M., & Ellickson, P. L. (2003). Patterns and correlates of binge drinking trajectories
from early adolescence to young adulthood. Health Psychology, 22,79–87.
Van Prooijen, J.-W., & Van der Kloot, W. A. (2001). Confirmatory analysis of exploratively obtained factor
structures. Educational and Psychological Measurement, 61(5), 777–792.
Wagner, E. F., Brown, S. A., Monti, P. M., Myers, M. G., & Waldron, H. B. (1999). Innovations in adolescent
substance abuse intervention. Alcoholism: Clinical and Experimental Research, 23, 236–249.
Watt, M. C., Stewart, S. H., Conrod, P., & Schmidt, N. B. (2008). Personality-based approaches to treatment
of co-morbid anxiety and substance use disorder. In S. H. Stewart & P. J. Conrod (Eds.), Anxiety and
substance use disorders: The vicious cycle of comorbidity (pp. 201–219). New York: Springer.
Wekerle, C., Leung, E., MacMillan, H. L., Boyd, M., Trocmé, N., & Waechter, R. (2009). The contribution of
childhood emotional abuse to teen dating violence among child protective services-involved youth.
Child Abuse and Neglect, 33,45–58.
Woicik, P. A., Conrod, P. J., Stewart, S. H., & Pihl, R. O. (2009). The Substance Use Risk Profile Scale: a
scale measuring traits linked to reinforcement specific substance use profiles. Addictive Behaviors, 34,
1042–1055.
Zuckerman, M. (1994). Behavioral expressions and biosocial bases of sensation seeking. New York:
Cambridge University Press.
506 Int J Ment Health Addiction (2011) 9:492–506