Maladaptive Personality Traits and Psychopathology in Childhood and Adolescence: The Moderating Effect of Parenting

Maladaptive Personality Traits and Psychopathology
in Childhood and Adolescence: The Moderating
Effect of Parenting
Barbara De Clercq, Karla Van Leeuwen, Filip De Fruyt,
Alain Van Hiel, and Ivan Mervielde
Ghent University
ABSTRACT
This study examines child and adolescent psychopathology
from a maladaptive trait perspective, incorporating both parental and
child ratings of parenting as a moderator of the personality-psychopa-
thology association. Hierarchical moderated regression analyses were
conducted on a combined sample of referred and nonreferred children
and adolescents (N 5 862, parental ratings of parenting and N 5 396,
child ratings of parenting). The results indicated positive main effects of
maladaptive traits on externalizing and internalizing problems, and pos-
itive main effects of parental negative control on externalizing problems.
Significant interactions were found for Disagreeableness and Emotional
Instability with parental Negative Control and for Disagreeableness Â
Positive Parenting in explaining externalizing problems. The discussion
focuses on the contribution of these findings to a better understanding of
the trait-psychopathology relationship at young age.
Child temperament/personality factors and parenting styles or
behaviors have been frequently found to be associated with the
development of child psychopathology (Nigg, 2006; Putnam,
Sanson, & Rothbart, 2002; Rettew & McKee, 2005; Rothbart &
Bates, 2006). This study, the first to broaden the trait perspective on
psychopathology towards the maladaptive side of personality at the
This research was supported by a PhD research grant awarded to the first author
(Grant 011D0201) under supervision of the second and fifth author.
Correspondence concerning this article should be addressed to Barbara De Clercq,
Department of Developmental, Personality and Social Psychology, Ghent University,
H. Dunantlaan 2, B-9000 Ghent, Belgium. E-mail: BarbaraJ.DeClercq@ugent.be.
Journal of Personality 76:2, April 2008
r 2008, Copyright the Authors
Journal compilation r 2008, Blackwell Publishing, Inc.
DOI: 10.1111/j.1467-6494.2007.00489.x

---

358
De Clercq, Van Leeuwen, De Fruyt, et al.
age of childhood, examines whether and how child problem behavior
is influenced by trait-like characteristics of psychopathology
observable at young age. The trait-psychopathology relationship is
described from a multivariate perspective (Tackett, 2006), relying on
a comprehensive set of hierarchically organized maladaptive traits
(De Clercq, De Fruyt, Van Leeuwen, & Mervielde, 2006) that are
hypothesized to represent developmental manifestations of person-
ality disorder pathology. The design further explores the moderating
role of parental behavior, directly building upon the notion that the
fit between a child’s temperament and the environmental context
influences adaptive and maladaptive developmental outcomes
(Thomas, Chess, Birch, Hertzig, & Korn, 1963) and incorporating
Shiner’s recommendation to include environmental measures
(Shiner, 2005) when studying the relation between temperament or
personality and psychopathology. The study, which may con-
tribute to the identification of trait-psychopathology associations
within the framework of maladaptive personality, investigates the
strength of maladaptive traits for the prediction of child psychopa-
thology by taking into account the moderating influence of parental
behavior.
Child Personality Traits and Psychopathology
Early individual differences have been traditionally conceived in
terms of temperamental characteristics (for a review of models, see
Mervielde & Asendorpf, 2000), whereas more stable trait-like differ-
ences observable later in life have been described as personality
traits. Although temperamental features appear from birth onward
and are presumed to have stronger biological roots than personality
traits, Caspi, Roberts, and Shiner (2005) recently argued that tem-
perament dimensions and the personality traits exemplified in the
Five-Factor Model (FFM; McCrae & Costa, 1999) share substantial
features and are thus much more alike than assumed previously.
The long-term continuity of these individual differences from
early childhood to adolescence (Abe, 2005), and even from child-
hood to adulthood (Caspi, 2000; Caspi & Roberts, 2001; Roberts &
Del Vecchio, 2000), has been convincingly demonstrated, further
indicating that early personality characteristics shape processes of
adaptation (Shiner, Masten, & Roberts, 2003) and maladaptation
(Caspi, 2000; Caspi et al., 2003) over time. Moreover, there is

---

Child Maladaptive Personality and Parental Behavior
359
growing evidence for childhood-onset psychopathology to evolve in
more maladaptive pathways compared to adolescent-onset psycho-
pathology (Tackett, 2006). In this respect, Frick et al. (2003) argue
that early onset psychopathology includes a larger trait component,
supporting the relevance of studying the relationship between traits
and psychopathology at young age. From this trait perspective on
the developmental course of psychopathology, the chronic nature
of child psychopathology (Tackett, 2006), indeed, may suggest
the presence of enduring trait characteristics that make some
children more likely to experience stable forms of externalizing or
internalizing psychopathology than others.
Within the externalizing spectrum, dysfunctional behavior such as
aggressive and delinquent behavior, can be related to the concept of
‘‘difficult temperament’’ (Tackett, 2006), with the FFM traits of
low Conscientiousness, low Benevolence, and high Neuroticism
(Krueger, Caspi, Moffitt, Silva, & McGee, 1996; Prinzie et al.,
2003; Tackett, 2006; Van Leeuwen, Mervielde, Braet, & Bosmans,
2004), early childhood impulsivity (Leve, Kim, & Pears, 2005; Tackett,
2006) and lack of self-control or disinhibition (Tackett, 2006).
Considering child internalizing problems, substantial linkages of
anxiety, shyness, and emotional reactivity to novel stimuli (Colder,
Mott, & Berman, 2002; Kagan, Snidman, Zentner, & Peterson, 1999;
Leve et al., 2005; Prior, Smart, Sanson, & Oberklaid, 2000) with in-
ternalizing problems have been described. Also, the FFM traits
Emotional Instability or Neuroticism and Introversion (Van Lee-
uwen et al., 2004), and the personality construct Behavioral Inhibi-
tion (Tackett, 2006) are significantly associated with (later)
internalizing problems. Extending the application of the tripartite
model (Clark & Watson, 1991) to children and adolescents, Lonigan,
Vasey, Philips, and Hazen (2004) posit that children with both anx-
iety and depression have a common personality profile of high
Neuroticism and low Conscientiousness.
So far, the associations between child psychopathology and
personality have been mainly addressed from normal trait variation,
with only few researchers focusing on the role of specific maladaptive
traits in childhood (see for example Salekin & Frick, 2005). The
present study introduces a more comprehensive perspective on mal-
adaptive personality at young age, exploring broad maladaptive
trait-psychopathology associations from an age-specific approach.
The maladaptive trait description relies on the Dimensional

---

360
De Clercq, Van Leeuwen, De Fruyt, et al.
Personality Symptom Itempool (DIPSI; De Clercq et al., 2006),
developed within the search for a dimensional representation of
personality disorder antecedents. The construction procedures of the
DIPSI are extensively described in De Clercq et al. (2006), resulting
in the 172-item taxonomy that is organized in 27 lower-level facets,
each representing a cluster of developmental maladaptive trait symp-
toms. The 27 facets are structured in four higher-order dimensions,
i.e. Disagreeableness (including extreme low-end variants of Benev-
olence such as Dominance/Egocentrism and Irritable/Aggressive
traits, high-end variants of Extraversion such as Hyperexpressivity
and Hyperactivity, and low-end variants of Conscientiousness such
as Distraction and Disorderliness), Emotional Instability (referring
to both Anxious and Depressive traits, and also including a Depen-
dency component), Introversion (describing extreme low-end vari-
ants of Extraversion, such as Withdrawn traits and Shyness), and
Compulsivity (including the high extremes of Conscientiousness
traits such as Perfectionism and Extreme Order). The nature of
these four higher-order maladaptive trait factors is reflected in their
correlations with normal-range personality traits (see De Clercq
et al., 2006), representing the extremes of these traits and providing a
more detailed description of pathological trait features that is not
fully accounted for by general trait or temperamental models. The
four-factor structure can further be represented within the common
framework of the FFM (Markon, Krueger, & Watson, 2005), and at
both the higher and lower level, the DIPSI structure is conceptually
similar (De Clercq et al., 2006) to adult maladaptive trait models
such as the DAPP-BQ (Livesley, 1990) and the SNAP (Clark, 1993).
Personality Traits and Psychopathology: Contamination of
Measures
Historically, childhood studies on temperamental traits or personality
and psychopathology have been conducted rather independently
(Frick, 2004; Tackett, 2006). The integration of both research areas
has been hampered by a number of issues, including the boundaries
between disciplines that study temperament versus psychopathology,
conceptual problems in defining the relationship between both
constructs and possible item overlap between measures of each
construct. A small number of researchers have addressed this last
issue empirically and have demonstrated significant relationships

---

Child Maladaptive Personality and Parental Behavior
361
between child temperamental traits and psychopathology, even after
controlling for measurement confounding between both constructs
(Lemery, Essex, & Smider, 2002; Lengua, West, & Sandler, 1998;
Oldehinkel, Hartman, De Winter, Veenstra, & Ormel, 2004).
However, the problem of overlapping item content may especially
hamper studies that focus on psychopathology from a maladaptive
perspective on personality because both measures focus on
pathological functioning.
Parental Behavior and Psychopathology
It is generally assumed that child psychopathology is not solely
driven by intrinsic child factors but is also affected by environmental
factors such as parental behavior. Several studies indicated that neg-
ative parenting in (young) children—such as inconsistent, harsh dis-
cipline strategies and limited warmth, lack of responsiveness or
rejection—is a risk factor for persistent externalizing problems (Bel-
sky, Hsieh, & Crnic, 1998; Rothbaum & Weisz, 1994). From a life-
span perspective, Johnson et al. (2001) demonstrated that children
who experienced maternal verbal abuse were more than three times
as likely to develop borderline, narcissistic, obsessive-compulsive,
and paranoid personality symptoms in adulthood, compared to chil-
dren who did not experience this abuse. From a FFM perspective,
Rogosch and Cicchetti (2004) showed that maltreated children were
considered less agreeable, less conscientious, and less open to
experiences compared to nonmaltreated children and that these
differences remained stable across time.
Compared to the externalizing spectrum, fewer studies focused on
the relation between parenting and child internalizing psychopathol-
ogy. In their meta-analysis, Wood, McLeoad, Sigman, Hwang, and
Chu (2003) found observed parental control (operationalized as
oversolicitous behavior, overprotection, and reduced granting of
autonomy) to be consistently associated with shyness and child anx-
iety disorders across studies. Van Leeuwen and colleagues (2004)
found main effects of high negative control (operationalized as re-
strictiveness in order to place limits on the child’s behavior) and
main effects of low positive parental behavior on child internalizing
problems.
Beyond the main effects of personality traits and parental behav-
ior on the child’s maladjustment, it is very likely that an interaction

---

362
De Clercq, Van Leeuwen, De Fruyt, et al.
between traits and parenting (O’Connor, 2002; Thomas et al., 1963)
explains the development of behavioral problems. Rothbart and
Bates (2006) have described a number of temperament  environ-
ment interactions, including interactions between temperamental
fearfulness (inhibition) and parental control to account for the de-
velopment of conscience (Kochanska, 1993), and between the tem-
perament construct Lack of Control and a single-parent home to
account for later criminal behavior (Henry, Caspi, Moffitt, & Silva,
1996). Prinzie and colleagues (2003) and Van Leeuwen and col-
leagues (2004) found significant interactions between parental neg-
ative control and the personality domains Benevolence and
Conscientiousness for explaining externalizing problem behavior.
These findings indicate that the effect of personality on a child’s de-
velopmental outcome often depends upon environmental measures
such as parenting. Frick (2004) therefore advocates that our under-
standing of childhood psychopathology would benefit from includ-
ing measures of the child’s context in the joint study of
temperamental traits and psychopathology.
The present study addresses this suggestion and studies child
psychopathology from both a maladaptive trait and parenting (rep-
resented by a negative, controlling parenting and a positive parent-
ing dimension) perspective. Relying on hierarchical moderating
regression analyses, we examine main effects of both childhood mal-
adaptive personality traits (Disagreeableness, Emotional Instability,
Introversion and Compulsivity) and parental behavior (Negative
Control and Positive Parenting) on the broad internalizing and ex-
ternalizing dimensions of psychopathology. We predict that both
personality and parental behavior have important main effects on
child psychopathology, with positive main effects of maladaptive
traits and parental negative control and a negative main effect of
positive parenting. We additionally consider parenting as a moder-
ator of the personality-psychopathology association, thus focusing
on how parental behaviors may act as risk or protective factors in
the development of the child’s (mal)adjustment (Putnam et al.,
2002).
In line with previous studies, we specifically expect negative
control to interact with maladaptive personality in explaining child
psychopathology, with high levels of negative control enhancing the
aversive effect of maladaptive traits on child psychopathology.
It is furthermore hypothesized that positive parenting also shows

---

Child Maladaptive Personality and Parental Behavior
363
interactive effects, with a high level of positive parenting weakening
the aversive effect of maladaptive traits on child psychopathology.
Before conducting the regression analyses, the present study
empirically investigates item overlap between the maladaptive trait
and psychopathology constructs relying on factor-analytic methods
(Lengua et al., 1998) and explores whether maladaptive trait-psy-
chopathology associations are valid or largely due to the contami-
nation of predictor and criterion measures. We assume a significant
number of overlapping items that confound the measurement of the
trait and psychopathology constructs and expect the largest item
overlap for the Disagreeableness-externalizing and Emotional Insta-
bility-internalizing combinations, given the previously found, strong
correlations between these constructs (De Clercq et al., 2006).
METHOD
Participants
Nonreferred children and adolescents.
Second- and third-year undergrad-
uate psychology students of Ghent University recruited children and
adolescents from the general population. Both samples are described sep-
arately, because of the differences in the number of administered ques-
tionnaires to the children and adolescents, respectively. The sample of
children (N 5 193) included 93 boys and 100 girls with a mean age of 10.82
years (SD 5 1.86), ranging from 7 to 15 years (Sample 1). The early-ad-
olescent sample (N 5 453) consisted of 216 boys and 237 girls with a mean
age of 13.9 years (SD 5 1.14), ranging from 10.5 to 16.7 years (Sample 2).
Almost all participants of both samples attended normal primary or sec-
ondary education, with only 1% attending special education for pupils
with learning disabilities. The educational level of mothers of Sample 1
included secondary education (40.8%), higher nonuniversity education
(49.7%), and university (9.4%), whereas 1.1% of the fathers followed
primary education, 46% secondary education, 35.8% higher nonuniver-
sity education, and 16% university. For Sample 2, the educational level of
mothers and fathers included secondary education (73.3% and 65.5%,
respectively), higher nonuniversity education (13.3% and 12.9%, respec-
tively) and university education (13.3% and 21.4%, respectively).
Referred children.
Two hundred and eighteen children who were
consulting general mental health services served as the referred sample.
Exclusion criteria were the presence of a physical disability or a condition

---

364
De Clercq, Van Leeuwen, De Fruyt, et al.
of chronic disease. Participants were recruited by third-year undergrad-
uate psychology students of Ghent University by contacting a psychol-
ogist or psychiatrist from an online list with registered providers of
mental care. Psychologists and psychiatrists randomly selected one of
their clients, following the chronological order of their appointment
schedule. Two hundred and five participants (this subsample is the
same as the referred sample in De Clercq et al., 2006) consulted outpa-
tient treatment programs in psychiatric clinics, pediatric units of hospi-
tals, services for school counselling, private psychotherapy services, and
psychomedical services for children with developmental and learning dis-
orders, whereas 13 participants were enrolled in an inpatient treatment
program at a pediatric psychiatric hospital. The children’s mean age was
9.9 years (SD 5 1.93), ranging from 5.2 to 14.6 years. The period of psy-
chological treatment ranged from 0 to 56 months, with a mean duration
of 10 months (SD 5 10.13). The treating psychologist/psychiatrist pro-
vided information on the initial reason for counselling: 21.0% of the
children had anxiety and depressive symptoms, 24.9% externalizing prob-
lem behavior (lying, aggression and temper tantrums), 15.6% exhibited
enduring behavioral and emotional difficulties related to major stress
events (divorce or decease of the parents), 6.8% struggled with psycho-
somatic complaints (pain without physical cause, eating and sleeping
problems), 9.3% suffered from attention and concentration problems
without neurological dysfunction, 11.2% demonstrated behavioral prob-
lems due to developmental disorders such as ADHD, Tourette syndrome
and autism spectrum disorder, 8.3% showed withdrawn behavior or de-
fective social skills, 1% exhibited obsessive-compulsive behavior, and
0.5% had symptoms of automutilation and suicidal thoughts. For 1.5%
of the children, no information on the primary reason for referral was
available. Almost all children attended normal primary education or kin-
dergarten (86.8%), with 13.2% attending special education for pupils
with learning disabilities. The educational level of mothers and fathers
ranged from secondary education (89.1% and 77.5% respectively), higher
nonuniversity education (11.4% and 8.8% respectively), and university
degree (13.7% and 8.5% respectively).
Procedure
Trained psychology students visited the samples of nonreferred children
(N 5 193) and adolescents (N 5 453) at home, providing detailed oral and
written instructions on how to complete the questionnaires. All mothers
were administered three inventories: a measure of child maladaptive
traits, a measure on child psychopathology, and a measure of parental

---

Child Maladaptive Personality and Parental Behavior
365
behavior. The children sample (N 5 193) also supplied ratings on parental
behavior.
Families of the referred sample (N 5 218) were contacted via the
treating psychologist or psychiatrist who introduced the study and
subsequently asked both the child and the mother to participate.
Mothers were requested to provide ratings on the same three inventories
administered to the nonreferred samples. The children of the outpatient
group (N 5 205) also described parental behavior.
The instructions were similar for all samples and underscored that
mother and child should complete the questionnaires independently. Par-
ticipants were not assisted in order to guarantee that their independent
opinion was assessed. Both mothers and children were assured that all
information would be treated as confidential and would serve only re-
search purposes. Written informed consent was obtained from all moth-
ers and children and from the psychologists/psychiatrists at the moment
of assessment.
Measures
Dimensional Personality Symptom Itempool for Children (DIPSI).
All
mothers rated their child on the DIPSI item pool (De Clercq et al., 2006),
judging the applicability of the 172 items referring to concrete, trait-re-
lated symptoms. In the DIPSI, 27 maladaptive personality facets (for an
overview, see De Clercq et al., 2006) are hierarchically organized in four
dimensions, to be calculated as the mean scores of their facets. Disagree-
ableness includes items like ‘‘Manipulates other children repeatedly to
have his way’’ (low-end Benevolence), as well as ‘‘Exhibits his/her inner
feelings at all occasions’’ (high-end Extraversion) and ‘‘Never takes care
of his/her belongings’’ (low-end Conscientiousness). Emotional Instabil-
ity includes facets that refer to both Anxious and Depressive traits with
sample items such as ‘‘Panics very easily’’ and ‘‘Feels often empty inside’’
and also includes a Dependency component represented by items like
‘‘Clings to other people’’ (item of the Insecure Attachment facet), ‘‘Needs
someone around all the time’’ (Dependency item), and ‘‘Always submits
to other children’’ (Submissiveness item). Introversion describes extreme
low-end variants of Extraversion with sample items such as ‘‘Cannot ex-
press feelings of affection,’’ whereas Compulsivity represents the high
extremes of Conscientiousness traits, including items such as ‘‘Wants life
to be perfectly organized.’’
The reliability analysis of the four higher-order maladaptive dimen-
sions indicates good internal consistency across samples with alpha co-
efficients of .97 for Disagreeableness and .96 for Emotional Instability,
.87 for Introversion, and .86 for Compulsivity. Alpha coefficients for the

---

366
De Clercq, Van Leeuwen, De Fruyt, et al.
27 facets range from .69 (Perfectionism) to .88 (Irritability, Affective
Lability, and Disorderliness), with a median value of .83.
The Ghent Parental Behavior Scale (GPBS).
The GPBS (Van Leeuwen
& Vermulst, 2004) describes parental rearing behavior and consists of
nine scales, hierarchically structured in two dimensions of parental be-
havior. The first factor represents the dimension of Negative Control
(calculated as the mean score on the scales of Discipline, Ignoring of
Unwanted Behavior, and Harsh Punishment) and the second factor de-
scribing Positive Parenting (calculated as the mean score on the scales
Positive Parental Behavior, Teaching Rules, and Autonomy). During the
construction phase, Van Leeuwen and Vermulst (2004) found the three
remaining scales (Monitoring, Material Rewarding, and Inconsistent Dis-
cipline) did not consistently load on the same dimension and suggested
they be dropped from the higher-order structure.
Both parental dimensions are strongly related to two constructs that
are frequently mentioned in the literature on parental behavior (Gallag-
her, 2002; Maccoby & Martin, 1983), with ‘‘Negative Control’’ referring
to efforts of parents to influence their child’s behavior, such as setting and
enforcing standards of behavior (control or demandingness), and ‘‘Pos-
itive Parenting’’ describing the affective nature of the parent-child rela-
tionship
(warmth
or
responsiveness,
to
be
differentiated
from
overprotectiveness or oversolicitousness). Psychometric properties of
the questionnaire were examined by Van Leeuwen and Vermulst (2004),
indicating good internal consistencies for the scales and satisfactory con-
firmatory factorial validity. A self-report form is available for parents
rating their own behavior, as well as a version for the child to describe the
behavior of his/her parents. The reliability analysis on the present com-
bined samples resulted in internal consistencies for parental ratings of
a 5 .81 (Positive Parenting) and a 5 .73 (Negative Control), and a 5 .79
(Positive Parenting) and a 5 .65 (Negative Control) for child ratings.
The Child Behavior Checklist (CBCL).
The CBCL (Achenbach, 1991;
Verhulst, Van der Ende, & Koot, 1996) assesses child psychopathology,
distinguishing eight syndrome scales (Withdrawn, Somatic Complaints,
Anxious/Depressed, Social Problems, Thought Problems, Attention
Problems, Delinquent Behavior, and Aggressive Behavior), and two
broadband scales (Internalizing and Externalizing Problems), aggregat-
ed in a total problem score. The CBCL syndrome scales have satisfactory
Cronbach’s alpha coefficients for the combined sample, ranging from .65
(Thought Problems) to .93 (Aggressive Behavior), with a median value of
.75, and alphas of .93 for the Externalizing and .91 for the Internalizing
Problem scale.

---