Ljungström et al.

Child Adolesc Psychiatry Ment Health (2020) 14:34 Child and Adolescent Psychiatry
https://doi.org/10.1186/s13034-020-00339-1
and Mental Health

RESEARCH ARTICLE Open Access

Bottom‑up and top‑down approaches

to understanding oppositional defiant disorder
symptoms during early childhood: a mixed
method study
Britt‑Marie Ljungström1* , Elisabeth Kenne Sarenmalm2,3 and Ulf Axberg4,5

Abstract
Background: Children with clinical levels of conduct problems are at high risk of developing mental health prob‑
lems such as persistent antisocial behavior or emotional problems in adolescence. Serious conduct problems in child‑
hood also predict poor functioning across other areas of life in early adulthood such as overweight, heavy drinking,
social isolation and not in employment or education. It is important to capture those children who are most at risk,
early in their development. The Diagnostic and Statistical Manual of Mental Disorders (DSM–5) is commonly used in
clinical settings, to identify children with conduct problems such as oppositional defiant disorder (ODD).This paper
presents a cross-sectional study in a clinical setting, and describes behaviors in 3- to 8-year-olds with ODD. Our aim
was to investigate whether there were problematic behaviors that were not captured by the diagnosis of ODD, using
two different methods: a clinical approach (bottom-up) and the nosology for the diagnosis of ODD (top-down).
Method: Fifty-seven children with clinical levels of ODD participated in the study. The mothers were interviewed
with both open questions and with a semi-structured diagnostic interview K-SADS. The data was analyzed using a
mixed method, convergent, parallel qualitative/quantitative (QUAL + QUAN) design. For QUAL analysis qualitative
content analysis was used, and for QUAN analysis associations between the two data sets, and ages-groups and gen‑
der were compared using Chi-square test.
Results: In the top-down approach, the ODD criteria helped to identify and separate commonly occurring opposi‑
tional behavior from conduct problems, but in the bottom-up approach, the accepted diagnostic criteria did not cap‑
ture the entire range of problematic behaviors-especially those behaviors that constitute a risk for antisocial behavior.
Conclusions: The present study shows a gap between the diagnoses of ODD and conduct disorder (CD) in younger
children. Antisocial behaviors manifest in preschool and early school years are not always sufficiently alarming to meet
the diagnosis of CD, nor are they caught in their entirety by the ODD diagnostic tool. One way to verify suspicion of
early antisocial behavior in preschool children would be to specify in the ODD diagnosis if there also is subclinical CD.
Keywords: Oppositional defiant disorder, Antisocial behavior, Mixed methods, K-SADS, Bottom–up-top–down

Background
Many children and adolescents in psychiatric care exhibit
conduct problems in the form of defiance, aggression
*Correspondence: [email protected]
1
and antisocial behavior [1]. Studies have shown that chil-
Department of Psychology, University of Gothenburg, Box 500, 405
30 Göteborg, Sweden dren who demonstrate serious conduct problems in early
Full list of author information is available at the end of the article childhood are at high risk to develop life-long difficulties

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​
mmons​.org/licen​ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat​iveco​mmons​.org/publi​cdoma​in/
zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 2 of 14

such as emotional problems and/or antisocial behavior for the DSM categories are met, each symptom in the
as well as poor functioning across other areas of daily life different diagnosis are judged absent or present. A top-
in early adulthood, such as heavy drinking, overweight, down approach is necessary to interpret and understand
social isolation and not in employment or education [2, what we perceive and is often used in research. However,
3]. In addition, if the antisocial behavior begins in early one criticism of the DSM is that its categories and criteria
childhood and persists through childhood and youth, are predetermined [14]. The top-down orientation makes
there is a great risk that it will stabilize and develop into it difficult to discover the unexpected by not allowing
antisocial personality disorder and criminality in adult- the formation of new knowledge. A bottom-up approach
hood [4]. Therefore, it seems essential to pay attention to uses individual cases to create new general knowledge;
childhood-onset of conduct problem and to constantly it progresses from individual elements to build a view
refine diagnostic instruments to promptly identify those of the whole, piecing data together until a larger pic-
children at highest risk for developing stable disruptive ture is formed [13]. Thus, a bottom-up approach is also
behavior [5, 6]. The present study will focus on conduct important since it can help us to refine and improve our
problems such as oppositional defiant disorder and con- knowledge systems. This type of information is of course
duct disorder in preschool children and early school age. more subjective and random, but it can be tested for
validity and utility as markers for important characteris-
Oppositional defiant disorder (ODD), conduct disorder (CD) tics, for robustness across samples and for developmen-
In the fifth edition of the Diagnostic and Statistical Man- tal course [13, 15]. Wakschlag and colleagues [9], argued
ual of Mental Disorders (DSM-5) severe conduct prob- the importance of integrating developmental psychology
lems are represented primarily by Oppositional Defiant and studies based on clinical research for generating and
Disorder (ODD) and Conduct Disorder (CD) [7]. ODD testing developmentally specified nosology. Their work
is defined mainly by irritable disposition and resistant is a bottom-up way of understanding behavioral prob-
interactions with authority figures, and is often concep- lems of young children. As a framework for a clinical and
tualized as a disorder of early childhood [8], while CD developmental understanding of disruptive behavior in
criteria are essentially designed to describe behavioral preschool-age children, they proposed a multi-dimen-
problems in older children and adolescents [9]. There sional model to define the core dimensions of disruptive
are no differences in the DSM diagnostic criteria for girls behavior problems (DBP) [9, 16]. The four core dimen-
and boys, but most studies show a higher prevalence of sions they created for disruptive behavior were Temper
ODD in boys than girls [10], and it has been questioned Loss (clinical indicator: frequent and intense temper
if the diagnostic criteria for ODD are as clinically useful tantrums), Aggression (frequent, hostile, and proactive
in girls as in boys [11]. CD is characterized by disregard aggression), Noncompliance (stubborn and pervasive
for social norms and rules, disregard for the rights and noncompliance), and Low Concern for Others (enjoying
well-being of others, and related serious aggression. The distressing and provoking others) [16].
DSM-5 defines three subtypes of CD: childhood-onset Oppositional defiant disorder (ODD) early in a
when at least one symptom of CD occurs in a child under child’s life should be considered an important marker
10 years; adolescent-onset when no symptoms occur of a wider pattern of dysfunctional regulation of behav-
prior to age 10; and unspecified onset [7]. In addition cal- ior and emotions that can predict the development of
lous-unemotional traits (CU-traits) through the Limited severe psychopathology [8, 17]. Many studies con-
Prosocial Emotions (LPE) specifier. ducted on ODD in children are top-down studies based
on defined diagnostic criteria [6, 18]. However, using
Different approaches to obtain information only a top-down perspective might impose limitations
Top-down and bottom-up approaches are two different in the information we obtain. To broaden the survey of
strategies for processing information [12]. A top-down problematic behaviors in preschool and early school-
approach uses an already formulated system or method aged children, our aim in the present study, is to go
and breaks the whole into smaller analytical pieces using more in-depth and beyond the diagnostic criteria by
existing general knowledge to understand the individual combining a bottom-up and a top-down approach in
case. The DSM, for example, offers a top-down process a mixed-method study, with a qualitative/quantitative
based on various diagnoses and criteria agreed upon (QUAL + QUAN) design. We will investigate whether
by groups of experts, the DSM committees [13]. In the there were problematic behaviors in children aged
text above, we have briefly described how children´s and 3–5 and 6–8 years that were not captured by the diag-
adolescent´s problems are classified according to diag- nosis of ODD, using two different methods: a clinical
nostic categories, where each diagnostic category implies approach (bottom-up) and the nosology for the diagno-
a taxonomic construct. To determine whether the criteria sis of ODD (top-down).
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 3 of 14

Aims Participants
The overall purpose of this study was to explore moth- To participate in the study, children had to be 3 to
ers’ descriptions of their children (bottom-up) and 8 years old and meet the DSM-IV criteria for ODD
compare them with descriptions that emerged through through the diagnostic interview K-SADS. Parents
a standardized semi-structured diagnostic interview had to understand Swedish well enough to fill out the
(top down). Specifically, we aimed to answer the follow- forms used in the study and give their permission for
ing questions: their child to participate in the study. Parents were
invited to participate in The Incredible Years training
1. How do mothers describe their children when they if their children’s problems met that program’s inclu-
are asked to identify the major problems they are sion criteria. Parents of 4 children declined to partici-
experiencing with their child (bottom-up)? How do pate after an initial telephone interview. Altogether, 62
mothers describe their children when they answer a children and their parents were included in the study.
standardized semi-structured diagnostic interview Parents of 5 children in the study dropped out. Parents
(top-down)? of 57 children (11 girls, 19%; 46 boys, 81%; 1:4) agreed
2. Are there any differences between gender and ages in to continue in the study. The ages of the children were
the mothers’ descriptions in the bottom-up and the 3–5 years, 35%; 6–8 years, 65%. Of the children in the
top-down approach? sample, 61% lived with their biological mother and
3. What kinds of convergences, divergences, contradic- father (or adoptive parents). Both of the parents of 91%
tions and associations in results are found between of the children had been born in Sweden. The educa-
the bottom-up and top-down approaches based tional level of mothers in the sample was in line with
on the mixed method in this study? Does it add the level of education of mothers in the general popu-
any additional dimension using mixed methods in lation. Most of the mothers, 77%, were employed or in
research on ODD? studies. According to Hollingshead´s four factor index
of social position [21], 41% were classified as low socio-
economic standard. Co-morbidity was common, 54% of
Methods the children with ODD, also met the criteria for atten-
Procedure tion deficit hyperactivity disorder (ADHD), one child
This study was part of the first RCT in Sweden to eval- met the criteria for CD and 14% an anxiety disorder.
uate the effectiveness of the American parent train- Other diagnoses were tics, Tourette`s syndrome, mal-
ing program, The Incredible Years [19]. The study was adaptive stress, enuresis, and encopresis. Some of the
approved by Gothenburg University and The Sahlg- children also had autistic traits. Twenty-one percent
renska Academy Ethics committee (D:nr Ö 669-03). fulfilled criteria for three or more diagnosis. Sixty-four
Informed consent was obtained from all. percent of the girls and 28% of the boys had ODD only
Parents of children with disruptive behavior signed (p = 0.038).
up for parent training groups, which were provided at
two Child and Adolescent Mental Health Services in Measures and interviews
Skaraborg. A semi-structured interview, The Kiddie- QUAL interviews with open‑ended questions
SADS-Present Version (K-SADS) [20], was completed Initially, the mothers were asked open-ended questions
with the mothers to assess whether the children met to describe the major problems they experienced with
the criteria for ODD and other psychiatric diagnoses. their children. These questions were asked before the
From the very beginning, the ambition was to interview structured K-SADS interview, i.e. before anything about
both parents, but very often only mothers appeared. diagnosis or diagnostic criteria was mentioned and
Therefore, following the usual procedure of the child formed the basis of the qualitative analysis in the present
and adolescent mental health services, were the study study.
was carried out, and to obtain a more homogeneous
sample, a decision was made early in the research pro- QUANT component with K‑SADS
cess to interview only mothers. These interviews were K-SADS is a semi-structured diagnostic interview with
conducted by three experienced clinical psychologists, three different parts: an initial background interview, a
trained in the K-SADS interview. Each interview took diagnostic screening interview, and a section in-depth
approximately 3 h. The interviews were recorded and questions about the different diagnoses [20]. The K-SADS
coded. consists of several questions aimed to capture the essence
of each criterion in the different DSM-IV diagnoses.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 4 of 14

The interviewer assessed the parents’ answers about meaning units. The codes were then grouped into subcat-
the severity of the problem on a 4-point scale, on which egories, and the subcategories were grouped into larger
0 = not enough information, 1 = the problem does not categories. The analysis focused on the manifest content
exist, 2 = below the threshold and 3 = above the threshold and the codes; subcategories and categories were limited
(if the problematic behavior occurred at least 4–5 times to textual content and did not involve interpretation of
per week). If any diagnostic criterion was met in the the underlying meaning of the text [23]. One of the co-
screening interview, supplemental questions were asked authors was an expert on DBDs and the other co-author
to decide whether the child met the full criteria for the was an expert in the qualitative method. The process of
specific diagnosis. Because the present study focused on developing codes and creating subcategories and catego-
conduct problems, all the supplemental questions about ries was laborious as the material was passes back and
existing ADHD, ODD, and CD were asked. The interview forth between the three authors, followed by discussions
is designed for children aged 6 to 17, but the questions until consensus regarding the codes, subcategories, and
for ADHD and CD were adapted by the psychologists in categories was reached. Finally, based on the categories,
the criteria it required, for children aged 3 to 5 years (35% three different themes became clear.
of the sample). The questions about ODD symptoms
were unchanged because they were judged well adapted QUAN analysis
to children aged 3 to 5 years. To test interrater reliability among the three psychologist
who conducted the K-SADS interviews, one third of the
Data analyses interviews concerning the diagnostic parts were coded by
The present study is an exploratory study using two coders; intraclass correlation coefficient (ICC) was
mixed methods with a convergent parallel design calculated for the ODD total score and the total symptom
QUAL + QUAN, in which qualitative and quantita- score (rater 1 and 2, and rater 1 and 3; two-way mixed;
tive data were collected concurrently [22]. Figure 1 dis- absolute agreement; average measure). Interrater reliabil-
plays the mixed method convergent parallel design ity was excellent at over 0.95 for all combinations.
schematically. We made a description of frequencies from the data set
of the eight criteria for ODD. Girls and boys, and ages-
QUAL analyses groups were compared on the distribution of the eight
The qualitative content analysis was performed, follow- criteria using Chi-square test. We also compared boys
ing the steps suggested by Graneheim and Lundman [23] and girls, and age groups in the qualitative content analy-
and Elo and Kyngäs [24]: Mothers (n = 57) described sis using Chi-square.
their children’s major problems, and the interviewer
wrote down these descriptions. The transcripts of the Mixed‑method analysis
formulated problems were read through several times to The third step in the mixed method convergent par-
learn “what was going on” and to discern patterns. Each allel design was to merge the two data sets together
separate problem formed a meaning unit for the analy- (QUAL + QUAN) [22], to achieve a more complete
sis. The transcript of the children’s problems was then understanding of ODD. Even if each part of QUAL
condensed. The next step was to code these condensed and QUAN would be complete in themselves, we

Frequency
Qualitative
Qualitative distribution,
content Transform to Compare the
data compare
analysis, dataset data sets
Bottom-up gender and
deductive
ages
Associations
between the
two data sets
Frequency
Quantitative distribution,
Analysis of Compare the
data compare
data data sets
Top-down gender and
ages

Fig. 1 Convergent parallel design; QUAL + QUAN

Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 5 of 14

were interested in finding and analyzing the interface Problematic behaviors

between the two [25]. This process was carried out in The largest group of reported problems were Problem-
two different ways. First, we put the data sets next to atic behaviors that mothers experienced as very stress-
each other and compared for contradictions, diver- ful. During the analysis, six categories of problematic
gences, and convergences. In this way, we achieved a behavior emerged: defiant, emotionally externalizing,
triangulation of the results of the bottom-up and top- norm-breaking, provocative, aggressive, and hyperactive
down analyses [26]. Next, we looked for associations behaviors. The problems that most mothers experienced
between the two data sets using Chi-square test. One as very difficult were defiant behaviors (see Fig. 3).
of the ideas of the study was to explore possible asso- Only a few mothers described hyperactive behavior as
ciations between the data sets, therefore we chose not a major problem, even though 54% of the children met
to correct for multiple comparisons. To allow individ- the criteria for ADHD in K-SADS interview.
ual comparisons between mothers’ descriptions; their We found three different qualities of Defiant behav-
most serious problems with their children (bottom-up) ior in the analysis; disobedience, inflexibility and
and their responses to the diagnostic questions (top- rebelliousness. The mothers described Emotionally
down), we registered each child with the themes and externalizing behavior as aggressive outbursts, severe
categories that their problems belonged. outbursts of anger, temper tantrums, and screaming
Because the analyzed data were categorical or ordi- and shouting. Norm-breaking behavior was divided in
nal, we calculated the frequencies and proportions and two subcategories: dishonesty (lying, fibbing, and blam-
used non-parametric statistical tests for comparisons ing others) and negative behavior (breaking things,
between the groups, Mann–Whitney test for ordinal running away from home, and drawing on walls).
data, and Chi-square test, or Fisher test for categorical Aggressive behavior was sorted into three subcategories
data, when cells were smaller than 5. according to whether children violated siblings, par-
ents, or peers. Provocative behavior included teasing,
but more commonly conflict-seeking, conflict-seeking
Results and -initiating, fighting with siblings, and fighting with
QUAL results peers. The mothers described Hyperactive behaviors
Of 225 problems formulated from mothers’ descrip- as over-activity and restlessness. These categories are
tions, we used 220 in the qualitative content analy- described in Table 1.
sis. The five problems discarded were judged to
differ from research question. Because the qualitative
analysis was limited to textual content, we have been Problematic traits
responsive to in which way mothers verbally described Many of the problem’s mothers formulated expressed
their children. The more familiar the material became, three main categories of Problematic traits within the
the clearer it was that the mothers described three child: defiant, negative emotional, and hyperactive/
main things, which formed the three different themes impulsive traits. More mothers thought that defiant
in the study. Mothers described Problematic behav- traits were a bigger problem than negative emotionality
ior—what the children do (e.g., constantly fighting, and hyperactivity/impulsivity. The lower part of Table 1
attacking those who pass), the Problematic traits of shows the qualitative content analysis of problematic
the children—how the children are (e.g., is frustrated traits in children.
and angry), and the child’s Difficulties—what they
could not do (e.g., has difficulty mastering his tem- Difficulties
per). Sometimes these themes seemed to interweave Of all the problems the mothers formulated, nearly one
as above; is frustrated and angry and has difficulty third described Difficulties within the child. We sorted
mastering his temper, but in the analysis we separate these difficulties into five different categories: behavioral
them as mothers´ descriptions of how they described regulation, emotional regulation, flexibility, cognitive (the
their child more in terms of traits (is frustrated and largest category), and social interaction difficulties. The
angry) and how they saw their child´s difficulties (has five categories with their subcategories and examples of
difficulty mastering his temper). Of these, three main problems are shown in Table 2.
themes were Problematic behavior (44%), Problematic
traits (25%), and Difficulties (31%.) Fig. 2 shows how QUAN results
the problems mothers described were analyzed and A descriptive compilation was made based on the diag-
sorted into themes, each with various categories. nostic questions in the K-SADS. Table 3 shows the dis-
tribution of the various ODD criteria. We also compared
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 6 of 14

Problematic
behavior

Emotionally Norm-
Defiant Provocative Aggressive Hyperactive
externalizing breaking
behavior behavior behavior behavior
behavior behavior

Problematic
traits

Negative Hyperactive
Defiant traits emotional and impulsive
traits traits

Difficulties

Behavioral Emotional Social

Flexibility Cognitive
regulation regulation interaction
difficulties difficulties
difficulties difficulties difficulties

Fig. 2 Results from the qualitative content analysis showing three different themes; Problematic behavior, Problematic traits, and Difficulties and
the categories within each

There were significant differences in criterion 3, Often

Problemac behavior actively defies or refuses; boys defied or refused signifi-
cantly more often than girls (p = 0.046), and in criterion
Defiant behavior 4, Often deliberately annoys others, with girls deliber-
Emoonally externalizing ately annoying others significantly more often than boys
behavior (p = 0.037). There were significant differences between
Aggressive behavior age groups; children aged 3–5 years lost their tempers
Norm-breaking behavior more often but were less often touchy or easily annoyed
Provocave behavior
than the other age group (p = 0.039 vs. p = 0.030).
In the distribution of the themes and categories there
Hyperacve behavior were some differences between boys and girls. In the
themes, girls were significantly more often described to
have Problematic behaviors (p = 0.049). In the categories,
Fig. 3 Description of problematic behaviors
girls were significantly more often described to have defi-
ant traits (p = 0.006) and defiant behavior (p = 0.017).
When comparing the different age groups, we found no
boys, girls and ages in terms of the distribution of diag- significant differences between ages 3–5 and ages 6–8.
nostic criteria (Table 3) and the themes and categories in
the qualitative content analysis (Table 4).
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 7 of 14

Table 1 Problematic behavior (PB) and Problematic traits (PT) are two of the three themes from the qualitative content
analysis. The figure presents the different categories with their subcategories and some examples of different problems,
described by the mothers
Category (PB) Subcategory/dimensions Examples of problems

Defiant behavior Disobediance Does not obey; Does not obey when choosing clothes; Causes trouble when eating;
Does not do homework; Creates conflicts in everyday routine situations; Does not
listen; Does not mind me
Inflexibility Has his own rules; Wants to decide everything himself; Wants to do things in his own
way; Has strange ideas; Has fixed ideas
Rebelliousness Powerstruggles with mother daily; Constant contrary to parents; Test limits; Protest‑
ers; Refuses to follow rules; Refuses to follow corrections; Refuses to stop; Refuses to
cooperate; Constant struggles each time when doing things
Emotionally externalizing behavior Outbursts Aggressive outbursts; Severe outbursts of anger; Temper tantrums
Noise Screaming and shouting
Norm-breaking behavior Dishonesty Lying; Fibbing; Blaming others
Negativ behavior Running away from home;
Breaking things; Drawing on walls
Provocative behavior Teasing Teases little brother; Treats brother bad
Conflict seeking Seeking and initiating conflict; Fighting with siblings; Fighting with peers
Aggressive behavior Violates siblings Hits little brother; Attacks younger sister; Physical fights with siblings
Violates parents Attacks mother; Wants to hurt dad
Violates peers Attacks others, Constantly fighting; Attacks those who pass
Hyperactive behavior High activity Overactivity; Anxiety with restlessness
Category (PT)
Defiant traits Disobediant traits Is obstinate; Bossy; Insistent; Loudmouthed; Contentious
Inflexible traits Is stubbhorn; Headstrong; Strong-willed
Rebellious traits Is impertinent; Disrespectful
Provocative traits Is provocative; Disorderly
Negative emotional traits Internalized traits Is tense and anxious; Afraid; Constantly Worried; Has fears; Is afraid to fail; Is sad; Does
not feel well; Has low self-esteem
Negativistic traits Is unpleasant; Jealous; Negative; Is grumpy; Frustrated; Never satisfied; Whiny
Aggressive traits Is angry; Aggressive; Frustrated and angry; Aggressive with no provocation
Hyperactive and impulsive traits High activity traits Is restless; Active; Intense; Extremely impatient
Impulsive traits Is intrusive; Impulsive

Mixed methods result or refuse to comply with adult’s requests or rules (50% vs.
Associations between the two data sets 80%).
There was an association between the two variables Often We also expected that children who fulfilled the criteria
angry and resentful (DSM criteria) and the category of Often spiteful and vindictive or Often angry and resent-
norm-breaking behavior (p = 0.029) in which children ful should be associated with the category of aggressive
with norm-breaking behavior seemed to be less angry behavior, but no such association was found (p = 0.219;
and resentful (0% vs. 72%). p = 0.390).
There was also an association between Often argues
with authority figures and the category of difficulties Discussion
with cognitive skills (p = 0.022) in which children with The overall purpose of this study was to explore moth-
cognitive difficulties seem to argue less with authority ers’ descriptions of their children (bottom-up) and com-
Figs. (71% vs. 95%). pare them with the descriptions that emerged through a
A tendency for association was found between the standardized diagnostic interview (top-down). To make
criterion Often actively defies or refuses to comply with this comparison possible, we used a mixed method
adults’ requests or rules and the category of aggressive design. When comparing problem descriptions and diag-
behavior (p = 0.100). Children who showed aggressive noses between bottom-up and top-down approaches, we
behavior toward others seemed less often to actively defy found both convergent and divergent results.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 8 of 14

Table 2 Five different categories of Difficulties (D) with subcategories and problems, as described by the mothers
in the study
Category (D) Subcategory/dimension Examples of problems

Difficulties with behavioral regulation Difficulties with hyperactivity Cannot sit still; Does not know when to stop
Trouble with compliance Has difficulty being corrected; Has difficult to accept no; Gets upset being
told off
Difficulties with independence Cannot play alone and self-entertain; Requires constant attention; Always
need to be in centre
Difficulties with emotional regulation Difficulties regulating anxiety Is concerned about mother; Wants to be very near parents when going to
sleep; Difficult to separate when going to school
Difficulties regulating mood Has difficulty mastering temper; Easily angry; Loses control over temper
Difficulties with flexibility Difficulties with changes Difficulties with different transitions; Needs much preparation; Sensitive to
change; Unsure in new situations
Becomes fixated Hard to break off activities; Rigid in many situations
Cognitive difficulties Lack of concentration Cannot concentrate; Is careless with everything
Inattention Has difficulty listening; Is difficult to reach and talk to
Difficulties with intellectual capacity Does not always understand; Does not understand consequences
Difficulties with social Interaction Limited social skills Can only interact with one person at a time; Has difficulty with social skills
Limited in acting in groups Has difficulty acting in a group; Has difficulty interacting with peers; Has
difficulty with social interactions in school

Convergence between bottom‑up and top‑down resentful. Mothers reported as a major problem that their
For the ODD diagnosis as a whole, bottom-up and top- children had aggressive traits and described them as angry,
down converged. Two of the themes from the qualitative frustrated and angry, aggressive, and prone to unprovoked
content analysis, Problematic behavior and Problematic aggression, but no mother described their child as resent-
traits, correlated well with the DSM’s construction of the ful, which is included in the diagnostic description. Both
ODD diagnosis in terms of behaviors (e.g., often loses approaches, however, highlighted a group of children with
temper, often argues, often blames others) and descrip- ODD who have an ongoing angry mood.
tions of traits (e.g., is often touchy, spiteful, angry, and There was also convergence between top-down and
resentful). The third theme, Difficulties, described regu- bottom-up approaches in the category of Difficulties. A
latory difficulties many of these children had, showing large group of children with ODD are immature in their
comorbidity with other DSM diagnoses such as ADHD, cognitive, social, and emotional development [27, 28].
depression, anxiety, and autistic traits, which converged When mothers in the present study shared their experi-
with the diagnostic interviews. There was also conver- ences of the largest problems, 68% highlighted the vari-
gence between top-down and bottom-up descriptions of ous constraints their children faced (74% of the boys, 46%
the essential qualities of ODD, expressing both behavio- of the girls). They described deficits in emotional regula-
ral and emotional disturbances. tion, attention regulation, and behavioral control. Some
Three of the criteria that seem to be core symptoms children also had limitations in intellectual capacity and/
in the ODD diagnosis overlapped very well with the cat- or difficulty with flexibility. Assessment on the K-SADS
egories and subcategories in the bottom-up analysis. The interview revealed that 54% met the criteria for ADHD or
criterion Often actively defies or refuses to comply with ADHD UNS, showing deficits in attention and behavio-
adults’ requests or rules equated with the category defiant ral regulation. When children with regulation deficits and
behavior, which was the most common problem moth- neurological immaturity are exposed to requirements
ers reported, describing children who create conflicts in they are not yet mature enough to cope with, their defi-
everyday routine situations and who refuse to follow cor- ant behavior often occurs at the intersection of demands
rective instruction. Often loses temper was also one of the to self-regulate and their inability to do so [29]. Further-
two most common criteria in DSM met by children with more, the inflexibility in the bottom-up category of defi-
ODD, with good agreement with emotionally externalizing ant traits and the children´s aggressive behaviors could
behaviors. This category was the second largest group of also be early signs of autism. It is important not to con-
problems mothers mentioned, describing tantrums, out- sider the children as simply brutal and defiant; instead we
bursts, and screaming and shouting, which matched the should be aware that many may have developmental neu-
criterion well. The third criterion with good agreement rological difficulties and to a large extent depend upon
between top-down and bottom-up was Often angry and adaptions and support from their environment.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 9 of 14

Table 3 The distribution of the different ODD criteria. Comparisons between boys and girls and between age-groups
Criteria, n (%) Study group, n = 57 Girls, n = 11 Boys, n = 46 p-valuea 3–5 years, n = 21 6–8 years, n = 36 p-valueb

Often loses temper

No (1–2) 7 (12.3) 2 (18.2) 5 (10.9) 0.507 0 7 (19.4) 0.039*
Yes (3) 50 (87.7) 9 (81.8) 41 (89.1) 21 (100) 29 (80.6)
Often argues with authority figures
No (1–2) 7 (12.3) 1 (9.1) 6 (13.0) 0.720 2 (9.5) 5 (13.9) 0.628
Yes (3) 50 (87.7) 10 (90.9) 40 (87.0) 19 (90.5) 31 (86.1)
Often actively defies or refuses to comply.
No (1–2) 13 (22.8) 5 (45.5) 8 (17.4) 0.046* 6 (28.6) 7 (19.4) 0.428
Yes (3) 44 (77.2) 6 (54.5) 38 (82.6) 15 (71.4) 29 (80.6)
Often deliberately annoys others
No (1–2) 23 (40.4) 1 (9.1) 22 (47.8) 8 (38.1) 15 (41.7)
Yes (3) 34 (59.6) 10 (90.9) 24 (52.2) 0.037* 13 (61.9) 21 (58.3) 0.791
Often blames others for his or her mistakes.
No (1–2) 23 (40.4) 4 (36.4) 19 (41.3) 6 (28.6) 17 (47.2)
Yes (3) 34 (59.6) 7 (63.6) 27 (58.7) 0.764 15 (71.4) 19 (52.8) 0.166
Is often touchy or easily annoyed
No (1–2) 15 (26.3) 4 (36.4) 11 (23.9) 0.455 9 (42.9) 6 (16.7) 0.030*
Yes (3) 42 (73.7) 7 (63.6) 35 (76.1) 12 (57.1) 30 (83.3)
Is often angry and resentful
No (1–2) 19 (33.3) 4 (36.4) 15 (32.6) 0.812 9 (42.9) 10 (27.8) 0.244
Yes (3) 38 (66.7) 7 (63.6) 31 (67.4) 12 (57.1) 26 (72.2)
Is often spiteful or vindictive
No (1–2) 40 (70.2) 7 (63.6) 33 (71.7) 0.717 17 (81.0) 23 (63.9) 0.235
Yes (3) 17 (29.8) 4 (36.4) 13 (28.3) 4 (19.0) 13 (36.1)
1 = No problem, 2 = below threshold, and 3 = above threshold
a
p-value considers comparisons between boys and girls
b
p-value considers comparisons between the two age groups
* Statistically significant according to Chi-2 test (Fisher’s test when n < 5 in some of the cells)

Divergences between bottom‑up and top‑down more active refusals and resistance against parents.
In contrast to the DSM, which employs a categorical sys- Mothers described power struggles, refusals, and behav-
tem, the bottom-up process allows more multi-dimen- iors associated with a strong negative affect and more
sional thinking, as seen in the case of the category defiant aggression in refusing to comply with adults’ requests or
behavior. Mothers in the study described three differ- rules, which is in the line with Wakschlag et al. [9]. Our
ent qualities or dimensions of defiance in their children. findings show notable variations in the quality, severity,
The first dimension, disobedience, included children’s and intrinsic degrees of aggressiveness and inflexibility in
attempts to ignore and/or act against their parents’ direc- all three dimensions of defiant behavior. In children with
tives. This typically defiant behavior occurs when chil- defiant behavior, further investigations are warranted to
dren disobey, do not listen, and generate conflicts during access the quality of the behavior, which could contribute
everyday routine situations. The second defiant dimen- to better adapted interventions for children with more
sion was inflexibility. These children had difficulty adapt- serious defiant behaviors.
ing to the demands of their surroundings. They had their Second, in the diagnostic interview, 74% of moth-
own rules, wanted to do things their own way, and had ers responded affirmatively to the item Is often touchy
strange and fixed ideas. They expected to live on their or easily annoyed as a problem on the clinical level.
own terms and became very frustrated when things did In the bottom-up description, however, no mother
not go their way. They seemed to be more rigid in their said spontaneously that her child was touchy or eas-
defiant behavior. The third dimension of defiance was ily annoyed. Instead, they described their children as
rebelliousness. This dimension seems to be the most grudging (unpleasant, jealous, negative, grumpy) and
severe of the three. The rebellious dimension includes displeased (frustrated, never satisfied, whiny). We called
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 10 of 14

Table 4 Results from the qualitative content analysis, with themes and categories. The table also shows comparisons
between boys and girls and between the two age-groups
Theme and Category, n (%) Girls, n = 11 Boys, n = 46 p-valuea 3–5 years, n = 21 6–8 years, n = 36 p-valueb

Problematic behaviors 11 (100) 32 (69.6) 0.049* 15 (71.4) 28 (77.8) 0.591

Problematic traits 8 (72.7) 24 (52.2) 0.315 10 (47.6) 22 (61.1) 0.322
Difficulties 5 (45.5) 34 (73.9) 0.068 13 (61.9) 26 (72.2) 0.419
Norm-breaking behavior 2 (18.2) 2 (4.3) 0.095 1 (4.8) 3 (8.3) 0.878
Provocative behavior 0 (0) 6 (13.6) 0.334 2 (9.5) 4 (11.1) 0.823
Aggressive behavior 2 (18.2) 4 (8.7) 0.586 2 (9.5) 4 (11.1) 0.823
Defiant behavior 10 (90.9) 23 (50.0) 0.017* 11 (52.4) 22 (61.1) 0.440
Emotionally externalizing behavior 6 (54.5) 16 (34.8) 0.248 7 (33.3) 15 (41.7) 0.480
Hyperactive behavior 1 (9.1) 1 (2.2) 0.357 1 (4.8) 1 (2.8) 0.710
Negative emotional traits 3 (27.3) 18 (39.1) 0.728 5 (23.8) 16 (44.4) 0.150
Defiant traits 6 (54.5) 7 (15.2) 0.006* 5 (23.8) 8 (22.2) 0.935
Hyperactive traits 0 (0) 6 (13.0) 0.334 2 (9.5) 4 (11.1) 0.823
Social difficulties 1 (9.1) 10 (21.7) 0.434 2 (9.5) 9 (25.0) 0.179
Cognitive difficulties 1 (9.1) 16 (34.8) 0.144 7 (33.3) 10 (27.8) 0.708
Inflexibility difficulties 1 (9.1) 8 (17.4) 0.671 4 (19.0) 5 (13.9) 0.715
Emotional regulation difficulties 1 (9.1) 6 (13.0) 0.703 3 (14.3) 4 (11.1) 0.754
Behavioral regulation difficulties 1 (9.1) 10 (27.1) 0.434 4 (19.0) 7 (19.4) 0.931
a
p-value considers comparisons between boys and girls
b
p-value considers comparisons between the two age groups
* Statistically significant according to Chi-2 test (Fisher’s test when n < 5 in some of the cells)

this category of answers negativistic traits. This category of 8 years. The present study has shown that aggressive,
seemed to express emotions of dissatisfaction (displeased norm-breaking, and provocative behaviors manifest in
and grudging) rather than a touchy mood. preschool and early school years are not caught in their
There was also a divergence between approaches in entirety by the ODD diagnostic tool, nor are they always
the results for antisocial behavior. The mothers reported sufficiently alarming in early childhood to meet the diag-
antisocial behavior in the bottom-up approach, but the nosis of CD (three criteria need to be met for a diagno-
ODD criteria did not catch these behaviors. Aggressive sis). Furthermore, some of the children also displayed
behavior (physical violation), together with provocative traits, similar to CU-traits and impulsive traits, which
behavior (seeking and initiating physical conflict), and/or might increase the risk to develop conduct problems
norm-breaking (destroying things, running away, lying), and later an antisocial personality [35]. DSM diagnostic
were presented as major problems in about one quarter tools seem sometimes to be too blunt and not adapted to
of the children in this study. Mothers described children capture emerging symptoms of severe conduct problems
who had physically attacked parents, peers, and siblings. in very young children; this study shows a gap between
Several studies have attempted to identify the character- ODD and CD criteria and diagnoses for the young-
istics of children with ODD who develop CD and anti- est children. In the top-down approach the ODD crite-
social personality [30–32]. Subthreshold CD symptoms ria were found to help identify and separate commonly
have been identified as predictive [8], and persistent occurring oppositional behavior from conduct problems,
physical fighting is particularly important [33, 34]. but in the bottom-up approach the accepted criteria did
Clinical psychologists sometimes seem to lack sup- not capture the entire range of problematic behaviors,
port in identifying such children. However, it might be especially aggressive, provocative, and norm-breaking
inappropriate to use the criteria for a diagnosis of CD in behaviors that risk developing into persistent antisocial
young children. Approximately a quarter of CD symp- behavior. One way to verify suspicion of early antisocial
toms seem to be developmentally impossible in early behavior in preschool children would be to have the pos-
childhood and approximately one third are developmen- sibility to specify in the ODD diagnosis, if there is also
tally improbable in preschoolers [9]. In the present study, subclinical CD (one or two criteria), in the same way as
only one of 57 (< 2%) children with ODD also met the full in the CD diagnosis where it is possible to retrospectively
criteria for CD, and this child was in the oldest age group
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 11 of 14

specify whether the individual has shown at least one CD the content analysis. Thus, ODD seems to be expressed
symptom before the age of ten (child-hood onset type). quite similarly in boys and girls. However, some small dif-
Comparing the content of reported problem behaviors ferences were found that is needed to be mentioned.
(bottom-up), we found similarity between our descriptions In the top-down approach, boys actively defied or
of Problematic behaviors and the dimensions described refused to comply with adult’s requests or rules more
by Wakschlag and collegues [16]. Wakschlag’s four core often than girls. In the bottom-up analysis, it was exactly
dimensions for disruptive behavior: Noncompliance, Tem- the opposite. Defiant behavior and defiant traits were
per Loss, Aggression, and Low Concern for Others corre- more often reported as major problems among girls than
spond well with our four categories of defiant, emotionally boys. These contradictory results might be explained by
externalizing, aggressive, and provocative behaviors. How- greater social acceptance of defiance in boys than in girls,
ever, in addition to the dimensions proposed by Wakschlag even when boys are more often and more strongly defi-
and colleagues [16], and the criteria for ODD, we also ant. If defiant behavior and traits in girls are less socially
found a group with norm-breaking behavior. accepted and considered more problematic, disobedience
in girls is probably more noticeable and more likely to be
Associations reported by mothers as a major problem, even though
An important part of the mixed-method analysis is to these behaviors are more frequent among boys. These
merge the two data sets [22]. Although the sample size results, however, do not fully correspond with Wright
was small, there appeared to be a distinction between and colleagues, who found that only fathers (not moth-
overt and covert externalizing behavior in this study. ers) had less tolerance for daughters’ than for sons’ DBD
These results are similar to those of Loeber and Burke behaviors [36]. However, due to the small sample seize
[32], who described three different pathways into antiso- in the present study the results should be interpreted
cial behavior: 1. Overt Pathway, beginning with annoy- cautiously.
ing others and bullying, moving on to more aggressive Another difference between boys and girls was found
behaviors such as physical fighting, gang fighting, and for the criterion Often deliberately annoys others. The
rape; 2. Covert Pathway, beginning with lying and shop- mothers´ bottom-up descriptions didn´t tally with their
lifting, leading to vandalism, pick-pocketing, and serious answers on the diagnostic question about annoying oth-
delinquency (theft, burglary); and 3. Authority Conflict ers. In the diagnostic interview, girls deliberately annoyed
Pathway, beginning with stubborn behavior, then disobe- others significantly more often than boys. Studies have
dience, staying out late, running away, and finally truancy shown that girls are significantly more relationally aggres-
and avoiding authority. sive than boys, and this sex difference is apparent as early
In the present study, several of the behaviors described as the preschool years [27]. Girls tend to engage in higher
above, such as annoying others, fighting, lying, destroy- levels of both proactive and reactive relational aggression
ing things, disobedience, running away and avoiding compered to physical actions [37]. This indicates that
authority, were represented among the children and were even in early childhood, relational aggression appears
clearly visible in the qualitative content analysis. Further- to be the modal type of aggression for girls. Comparing
more, children with norm-breaking behavior seemed to results from the bottom-up and top-down analyses, there
be less angry and resentful, while children who showed seem to be several levels of annoying severity; teasing
aggressive behavior toward others seemed less often to and annoying others are on the border of more norma-
actively defy or refuse to comply with adults’ requests or tive behavior levels, while seeking and initiating conflict
rules. The data are based on small groups and these find- (for boys) and relational aggression (for girls) might be
ings merit further research. However, they suggest that on a more clinical level. The differences between boys’
maybe it could be possible to trace different pathways and girls’ disruptive behaviors might be most visible on
of externalizing behavior even in children as young as this criterion. The different provocative behaviors, boys’
3–8 years. and girls’ distinct expressions need to be further explored
and perhaps more clearly defined based on diagnostic
Differences between girls and boys criteria.
While, relatively few studies have investigated differences
between boys and girls when it comes to ODD criteria, Differences between ages
one of our aims in this study was to examine whether In the bottom-up description, we found no significant
there seemed to be any differences between gender and differences in symptoms between ages 3–5 and 6–8.
between bottom-up and top-down approaches. We However, using the top-down approach 3- to 5-year-olds
found no considerable difference between boys and girls were more likely to often lose temper, and 6- to 8-year-
in the different ODD criteria or the 14 categories from olds were touchy or easily annoyed significantly more
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 12 of 14

often than children 3- to 5-year-olds. Interestingly, nega- asked to fill in various standardized questionnaires, often
tive emotional traits were described in the bottom-up based on diagnostic criteria. This kind of information is
approach as grudging (unpleasant, jealous, negative, thought to be valid and reliable. The bottom-up approach
grumpy) or displeased (frustrated, never satisfied, whiny), also gives important information when clinicians ask
while no mother spontaneously described their children questions that are more open-ended about the child´s
as touchy, easily annoyed or resentful. This might indi- history and context, which are not captured in struc-
cate that the diagnostic criteria that describe irritabil- tured questionnaires. In clinical settings assessments and
ity (touchy or easily annoyed and angry and resentful) treatment decisions usually are based on combining both
are more appropriate to describe older children, while sources of information.
grudging and displeased are a quality of descriptions
more appropriate for younger children. This might be Limitations
something to explore further in future studies. This study had several limitations. Present study has
There is also frequency criterion of the diagnosis in the a cross-sectional design and highlights convergences
DSM-5, providing guidance on minimum symptom fre- and divergences between top-down and bottom-up
quencies for different age groups. For children younger approaches concerning symptoms of ODD diagnosis. To
than 5 years, the behaviors should occur on most days for really know if the children who exhibited more severe
a period of at least 6 months, and for those aged 5 years conduct problem in this study later develop conduct
or older, the behavior should occur at least once peer disorder (CD), a longitudinal study would be needed.
week [7]. Our results from top-down and bottom-up Another important limitation was the small sample,
approaches raise questions about the frequency criteria regarding comparisons between boys and girls. In the
in DSM-5. Are there a risk that we over-diagnose chil- comparison of themes and categories, our calculation
dren from 5 years with ODD? According to Wakschlag of statistical significance using qualitative data might be
and colleagues, temper tantrums seem to be more com- considered problematic. However, although the qualita-
mon in preschool-aged children [38]. Our findings is in tive questions were open-ended, they structured to limit
line with this. The frequency of other symptoms appears the frames of interpretation. Another limitation was the
to be more constant between the age groups in the pre- use of K-SADS as measure for children younger than
sent study. It may be, that differences in frequency of dis- 6 years (35% of the sample). The diagnostic questions in
ruptive behavior symptoms between younger and older K-SADS are designed for children aged 6 to 17 years, but
children are greater in community samples than in clini- the probes and scoring criteria were adapted for children
cal samples, where the rates of externalizing behavior 3–5 years as necessary. This adaptation relied on the three
continue to be high for this group of children. psychologists’ experiences in developmental psychology.
Participants in the study were mothers only; multiple
Mixed methods design participants for each child would have provided a fuller
An important purpose with mixed methods is to investi- description of the children’s functioning. However, by
gate convergences and divergences in the interface of the solely interviewing mothers, we got a more homogene-
results of the two methods. Through the QUAN method ous group. Even if interesting, it was beyond the scope
we got information of the frequency distribution of the of present study to examine effects of comorbidity on
various criteria in the ODD diagnosis, and differences mothers´ description of the children. Furthermore, in an
between gender and age groups. Through the QUAL analysis of the most common form of comorbidity in our
method, we got mothers’ stories of the most difficult sample (ODD + ADHD) it did not seem to influence the
problems they experienced with their children. By mixing bottom-up descriptions of the children.
the QUAL and QUAN methods and exploring the associ- A strength of the study was the interviewer-based
ations between the two datasets we got a more vivid pic- diagnostics, in which the clinician decides whether a
ture of the children´s behavior behind the criteria of the symptom is present or absent. Many studies use rating
ODD diagnosis. Thus, it was possible to trace different scales instead, and respondents (especially parents) usu-
expressions of externalizing behavior even in children as ally have limited exposure to the full range of normative
young as 3–8 years. The difficulties with mixed methods behaviors. Another strength was the problem-formulated
studies are often to find a way to integrate the findings of scales, in which mothers described the most troublesome
QUAL and QUAN [39]. The mixed methods (convergent problems they had with their children, giving an entirely
parallel) design used in the present study is similar to different weight to the problems described. A further
the method of collecting information and doing clinical strength was the vivid picture of the different behaviors
assessment at a children’s psychiatric clinic. Top-down and personality traits hidden behind the diagnostic crite-
information is gathered when children and parents are ria that were revealed in the qualitative content analysis.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 13 of 14

Conclusions Author details

1
Department of Psychology, University of Gothenburg, Box 500, 405
In the top-down approach, the ODD criteria helped 30 Göteborg, Sweden. 2 Institute of Health and Care Sciences, Sahlgrenska
to identify and separate commonly occurring oppo- Academy at the University of Gothenburg, Box 400, 405 30 Göteborg, Sweden.
3
sitional behavior from conduct problems, but in the 4
Research and Development, Skaraborg Hospital, 541 85 Skövde, Sweden.
Fakultet for Sosialfag/Faculty of Social Studies, Familieterapi Og Systemisk
bottom-up approach, the accepted diagnostic criteria Praksis/Family Therapy and Systemic Pratice, VID Vitenskapelige høgskole/VID
did not capture the entire range of problematic behav- Specialized University, Oslo, Norway. 5 Department of Psychology, University
iors-especially those behaviors that constitute a risk of Gothenburg, Box 500, 405 30 Göteborg, Sweden.
for antisocial behavior. The present study shows a gap Received: 17 September 2019 Accepted: 26 August 2020
between the diagnoses of ODD and conduct disorder
(CD) in younger children. Antisocial behaviors mani-
fest in preschool and early school years are not always
sufficiently alarming to meet the diagnosis of CD, nor References
are they caught in their entirety by the ODD diagnos- 1. Lindhiem O, Bennett C, Hipwell A, Pardini D. Beyond symptom counts
for diagnosing oppositional defiant disorder and conduct disorder? J
tic tool. One way to verify suspicion of early antisocial
Abnorm Child Psychol. 2015;43(7):1379–87.
behavior in preschool children would be to specify in 2. Wertz J, Agnew-Blais J, Caspi A, Danese A, Fisher HL, Goldman-Mellor S,
the ODD diagnosis if there is also subclinical CD. et al. From childhood conduct problems to poor functioning at age 18
years: examining explanations in a longitudinal cohort study. J Am Acad
Child Adolesc Psychiatry. 2018;57(1):54–60.
3. Gardner F, Leijten P, Harris V, Mann J, Hutchings J, Beecham J, et al. Equity
Abbreviations
effects of parenting interventions for child conduct problems: a pan-
ODD: Oppositional defiant disorder; CD: Conduct disorder; CU-traits: Callous-
European individual participant data meta-analysis. Lancet Psychiatry.
unemotional traits; LPE: Limited prosocial emotions; DBP: Disruptive behavior
2019;6(6):518–27.
problems; QUAL + QUAN: Qualitative + quantitative.
4. Moffitt TE. Life-course-persistent versus adolescence-limited antisocial
behavior. New York: Wiley; 2015. p. 570–598.
Acknowledgements
5. Tremblay RE. Developmental origins of disruptive behaviour problems:
We would like to thank all the participating parents, the Child and adolescent
the ‘original sin’ hypothesis, epigenetics and their consequences for
mental health services involved and research secretary Anna-Lena Emanuels‑
prevention. J Child Psychol Psychiatry. 2010;51(4):341–67.
son Loft for skillful support preparing the manuscript.
6. Keenan K, Boeldt D, Chen D, Coyne C, Donald R, Duax J, et al. Predictive
validity of DSM-IV oppositional defiant and conduct disorders in clinically
Authors’ contributions
referred preschoolers. J Child Psychol Psychiatry. 2011;52(1):47–55.
BML was responsible for the interviews, analysed and interpreted the data,
mental disorders: DSM-5™. 5th ed. Arlington: American Psychiatric Pub‑
7. American Psychiatric Association. Diagnostic and statistical manual of
provided a first draft of the manuscript. EKS analysed and interpreted the
qualitative data together with BML and UA, and UA was responsible for the
lishing, Inc.; 2013.
analysis of the quantitative data and the interrater reliability calculations
8. Rowe R, Costello EJ, Angold A, Copeland WE, Maughan B. Developmen‑
together with BML. UA made substantial contributions by discussing the
tal pathways in oppositional defiant disorder and conduct disorder. J
introduction, discussions and implications parts and EKS made significant con‑
Abnorm Psychol. 2010;119(4):726–38.
tributions to the study by discussing the study design and results. UA was one
9. Wakschlag LS, Tolan PH, Leventhal BL. Research review: ’Ain’t misbehavin’:
of the research leaders and responsible for the RCT, from which the present
Towards a developmentally-specified nosology for preschool disruptive
study is a part. UA and EKS continuously revised and provide feedback of the
behavior. J Child Psychol Psychiatry. 2010;51(1):3–22.
manuscript. All authors read and approved the final manuscript.
10. Demmer DH, Hooley M, Sheen J, McGillivray JA, Lum JAG. Sex differences
in the prevalence of oppositional defiant disorder during middle child‑
Funding
hood: a meta-analysis. J Abnorm Child Psychol. 2017;45(2):313–25.
Open access funding provided by University of Gothenburg. The project
11. Pardini DA, Frick PJ, Moffitt TE. Building an evidence base for DSM–5 con‑
received funding by the Bank of Sweden Tercentenary Foundation (Riksbank‑
ceptualizations of oppositional defiant disorder and conduct disorder:
ens Jubileumsfond), the Research Fund at Skaraborg Hospital, Skövde, Sweden
introduction to the special section. J Abnorm Psychol. 2010;119(4):683–8.
and from the Skaraborg Institute for Research and Development, Skövde,
12. Biederman I, Glass AL, Stacy EW. Searching for objects in real-world
Sweden.
scenes. J Exp Psychol. 1973;97(1):22–7.
13. Achenbach TM. Assessmenet of Psychopathology. In: Sameroff AJ,
Availability of data and materials
Lewis M, Miller SM, editors. Developmental psychopathology. New York:
The datasets used and analyzed during the current study are available from
Springer; 2000.
the corresponding author on reasonable request.
14. Rutter M. Research review: child psychiatric diagnosis and classification:
concepts, findings, challenges and potential. J Child Psychol Psychiatry.
Ethical approval and consent to participate
2011;52(6):647–60.
All procedures performed in the study involving human participants were
15. Crittenden PM. Gifts from Mary Ainsworth and John Bowlby. Clin Child
in accordance with the ethical standard of the institutional and national
Psychol Psychiatry. 2017;22(3):436–42.
research committee and with the 1964 Helsinki declaration and its later
16. Wakschlag LS, Henry DB, Tolan PH, Carter AS, Burns JL, Briggs-Gowan MJ.
amendments or comparable ethical standards. The study was approved by
Putting theory to the test: Modeling a multidimensional, developmen‑
Gothenburg University and the Sahlgrenska Academy Ethics committee (D:nr
tally-based approach to preschool disruptive behavior. J Am Acad Child
Ö 669–03).Informed consent was obtained from all individual participants
Adolesc Psychiatry. 2012;51(6):593–604.
included in the study.
17. Stringaris A, Goodman R. Longitudinal outcome of youth oppositionality:
Irritable, headstrong, and hurtful behaviors have distinctive predictions. J
Consent for publication
Am Acad Child Adolesc Psychiatry. 2009;48(4):404–12.
Not applicable.
18. Aebi M, Plattner B, Metzke CW, Bessler C, Steinhausen HC. Parent- and
self-reported dimensions of oppositionality in youth: Construct validity,
Competing interests
concurrent validity, and the prediction of criminal outcomes in adult‑
The authors declare that they have no competing interest.
hood. J Child Psychol Psychiatry. 2013;54(9):941–9.
Ljungström et al. Child Adolesc Psychiatry Ment Health (2020) 14:34 Page 14 of 14

19. Axberg U, Broberg AG. Evaluation of “The Incredible Years” in Sweden: the 31. Longman T, Hawes DJ, Kohlhoff J. Callous-unemotional traits as markers
transferability of an American parent-training program to Sweden. Scand for conduct problem severity in early childhood: a meta-analysis. Child
J Psychol. 2012;53(3):224–32. Psychiatry Hum Dev. 2016;47(2):326–34.
20. Kaufman J, Birmaher B, Brent D, Rao UMA, Flynn C, Moreci P, et al. Sched‑ 32. Loeber R, Burke JD. Developmental pathways in juvenile externalizing
ule for affective disorders and schizophrenia for school-age children-pre‑ and internalizing problems. J Res Adolesc. 2011;21(1):34–46.
sent and lifetime version (K-SADS-PL): initial reliability and validity data. J 33. Okado Y, Bierman K. Differential risk for late adolescent conduct problems
Am Acad Child Adolesc Psychiatry. 1997;36(7):980–8. and mood dysregulation among children with early externalizing behav‑
21. Hollingshead AB. The four factor index of social position. Yale Journal of ior problems. J Abnorm Child Psychol. 2015;43(4):735–47.
Sociology. 2011;8:21–522. 34. Tremblay RE. Prevention of youth violence: why not start at the begin‑
22. Creswell JW, Plano Clark VL. Designing and conducting mixed methods ning? J Abnorm Child Psychol. 2006;34(4):481–7.
research. Los Angeles: SAGE Publications; 2011. 35. Frogner L, Andershed A-K, Andershed H. Psychopathic personality works
23. Graneheim UH, Lundman B. Qualitative content analysis in nursing better than CU traits for predicting fearlessness and ADHD symptoms
research: concepts, procedures and measures to achieve trustworthiness. among children with conduct problems. J Psychopathol Behav Assess.
Nurse Educ Today. 2004;24(2):105–12. 2018;40(1):26–39.
24. Elo S, Kyngas H. The qualitative content analysis process. J Adv Nurs. 36. Wright AW, Parent J, Forehand R, Edwards MC, Conners-Burrow NA, Long
2008;62(1):107–15. N. The relation of parent and child gender to parental tolerance of child
25. Cheek J, Lipschitz DL, Abrams EM, Vago DR, Nakamura Y. Dynamic reflex‑ disruptive behaviors. J Child Fam Stud. 2013;22(6):779–85.
ivity in action. Qual Health Res. 2015;25(6):751–62. 37. Ostrov JM, Kamper KE, Hart EJ, Godleski SA, Blakely-McClure SJ. A gender-
26. Morse JM. Critical analysis of strategies for determining rigor in qualitative balanced approach to the study of peer victimization and aggression
inquiry. Qual Health Res. 2015;25(9):1212–22. subtypes in early childhood. Dev Psychopathol. 2014;26(3):575–87.
27. Crick NR, Zahn-Waxler C. The development of psychopathology in 38. Wakschlag LS, Choi SW, Carter AS, Hullsiek H, Burns J, McCarthy K, et al.
females and males: current progress and future challenges. Dev Psycho‑ Defining the developmental parameters of temper loss in early child‑
pathol. 2003;15(03):719–42. hood: implications for developmental psychopathology. J Child Psychol
28. Martel MM, Gremillion ML, Roberts B. Temperament and common Psychiatry. 2012;53(11):1099–108.
disruptive behavior problems in preschool. Personal Individ Differ. 39. O’Cathain A, Murphy E, Nicholl J. Integration and publications as
2012;53(7):874–9. indicators of "Yield" from mixed methods studies. J Mixed Methods Res.
29. Chang H, Olson S, Sameroff A, Sexton H. Child effortful control as a 2007;1(2):147–63.
mediator of parenting practices on externalizing behavior: evidence
for a sex-differentiated pathway across the transition from preschool to
school. J Abnorm Child Psychol. 2011;39(1):71–81. Publisher’s Note
30. Klingzell I, Fanti KA, Colins OF, Frogner L, Andershed A-K, Andershed H. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
Early childhood trajectories of conduct problems and callous-unemo‑ lished maps and institutional affiliations.
tional traits: the role of fearlessness and psychopathic personality dimen‑
sions. Child Psychiatry Hum Dev. 2016;47(2):236–47.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions