Psychiatria Danubina, 2010; Vol. 22, No.

2, pp 153–166 Original paper

© Medicinska naklada - Zagreb, Croatia

EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT,

PERSONALITY, AND PERSONALITY DISORDERS
Dragan M. Svrakic1,2 & Robert C. Cloninger1
1
Washington University School of Medicine, Department of Psychiatry, St Louis, MO, USA
2
VA Medical Center in St Louis, St Louis, MO, USA

received: 10.1.2010; revised: 20.3.2010; accepted: 15.4.2010

SUMMARY
After 30 years of clinical work and research based on categorical criteria for personality disorders (Diagnostic and Statistical
manual of Mental Disorders – DSM IV TR) and (International Classification of Diseases - ICD 10th revision), a solid conceptual
understanding and treatment of these disorders have not been established. For the field to move forward, it is imperative that future
classifications introduce major revisions of the concept, diagnosis, and classification of personality disorders. This paper proposes
one such revision.
Based on recent advances in molecular biology and epigenetics, we define personality disorders as maladaptive syndromes
developed trough person-environment interaction. We conceptualize maladaptation as a failure of integrative functions of
personality (i.e., those that carry out adaptive processes) caused by strong biogenetic dispositions or by pathological environmental
effects, or both. Hence, accurate diagnosis of personality disorder depends upon neurobiological (innate) and adaptive (interactive)
etiological factors. We propose a 2-step diagnostic algorithm for personality disorders: adaptive processes (i.e., character) are used
to diagnose maladaptation, whereas biological aspects (i.e., temperament) are used to specify dominant clinical presentation and for
differential diagnosis. We suggest that the term “Personality Disorder” be replaced by a more appropriate term "Adaptation
Disorder” as the latter reflects more accurately the real nature of the disorder and distributes the causality of maladaptive
syndromes more evenly, between the person and the environment.
Diagnostic, research, and treatment advantages of the proposed solution are discussed in some detail.

Key words: epigenetics – psychiatric classification – diagnosis – personality disorders- adaptation

* * * * *
INTRODUCTION solely on the basis of extreme behavior dimensions for
the following reasons:
After 30 years of clinical work and research based One of the most robust findings in personality
on categorical criteria for personality disorders research is that four broad dimensions underlie
(Diagnostic and Statistical manual of Mental Disorders normal temperament and symptoms of personality
– DSM IV TR (APA 2000)) and (International disorders (Cloninger et al. 1993, Ignjatovic &
Classification of Diseases - ICD 10th revision (ICD Svrakic 2003, Trull & Durrett 2005, Livesly 2007,
1992)), our conceptual understanding and treatments of Cloninger & Svrakic 2009). With respect to normal
these disorders have not substantially advanced. Recent temperament, these are (a) Harm Avoidance, (b)
survey of 400 experts demonstrates a major disapproval Novelty Seeking, (c) Reward Dependence, and (d)
of the categorical concept of personality disorders Persistence (Cloninger & Svrakic 2009). With
(Bernstein et al. 2007). Clearly, the field is stagnating, respect to personality disorders, the dimensions have
with no clear directions. For the field to move forward, been variously labeled in the literature: (a) Anxiety /
it is imperative that future classifications revise both the Neuroticism (b) Antisocial / Dissocial (c) Social
concept and the diagnosis of personality disorders. Withdrawal / Asocial, and (d) Conscientiousness /
Compulsivity (Trull & Durrett 2005, Livesly 2007).
Future classifications: imperative for change Note that the dimensions identified for personality
disorders are in fact extreme variants of normal
There is a growing pressure to “dimensionalize” temperament traits (specifically, the continuous
personality disorders, i.e., to describe them as variants are: high Harm Avoidance with Anxiety /
quantitative variants along a number of behavior Neuroticism, high Novelty Seeking with Antisocial /
dimensions or traits (Livesly 2007, Cloninger & Svrakic Dissocial, low Reward Dependence with Social
2009). Some (Livesly 2007) are advocating that the Withdrawal / Asocial, and high Persistence with
focus ought to be on extreme ("pathological") behavior Conscientiousness / Compulsivity). In conclusion,
traits, which is intended to separate syndromes of no traits, extreme or otherwise, have been identified
personality disorder from normal personality variants. as discriminative of personality disorders. These
However, it is of critical importance to recognize that disorders are not de novo personality manifestations,
personality disorders cannot be accurately defined but continuous (although extreme) variants of

153
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

normal temperaments. As discussed later, the differences in both normal and extreme behavior
defining aspect of personality disorder is poor variants, ii) for a conceptual revision that includes
adaptation to the environment, not extreme maladaptation to the environment as an integral
dimensions or traits. component in the diagnosis, and iii) for the distinction
The robustness of the 4-factor structure of between biogenetic and adaptive personality processes –
temperament throughout history, across normative which is critical to guide more productive research,
studies, clinical and nonclinical samples, and more reliable diagnosis, and more efficient treatment of
measuring instruments, indicates that it reflects personality disorders (discussed later in the text).
natural building units of personality that “carve the
nature at its joints”: these four dimensions are Personality and behavior development:
general determinants of fundamentally different beyond the Nature-Nurture dichotomy
behaviors patterns (see (Cloninger & Svrakic 2009)
for review) – stable over time, across situations, and Recent breakthroughs in the field of epigenetics
cross culturally (Cloninger et al. 1994, Miettunen et have advanced our understanding of molecular
al 2006). Of note, the Five Factor Model (FFM) mechanisms underlying gene – environment (GxE)
(Costa & McCrae 2008) defines four factors (i.e., interaction. It has become increasingly clear that genetic
Neuroticism, Extraversion, Conscientiousness, and influences on the phenotype are not direct and
Openness) which correspond to the four dimensions invariable. DNA is no more considered to be the master
described above. A closer review of the fifth factor blueprint for physical and behavioral features that
in the FFM (Agreeableness) reveals that it actually operates in an ecological vacuum. Rather, DNA outlines
corresponds to adaptive behaviors referred to as the overall adaptive potential of an organism through
character traits in this work. broadly outlined ("uncommitted”) physical and behavior
dispositions which serve as building material for the
a) No rational (i.e., non-arbitrary) cut off points on the
final phenotypic outcome in response to specific
continuum between normal and extreme positions on
behavior dimensions have been found. For environmental stimuli (Templeton 2006).
illustration, behavior traits typical of personality We refer the reader to (Sato et al. 1998, Impey et al.
disorders (e.g., compulsivity) are continuous with 2004, Pruessner et al. 2004, Fraga et al. 2005, Colvis et
normal traits (e.g., persistence) with no clear cut-off al. 2005, Champagne 2008, Champagne & Curley 2008,
points to mark the transition from normal persistence Svrakic et al. 2009) for details. Here, we provide a
to pathological compulsivity. Hence, diagnostic review of epigenetic influences on behavior develop-
models defined only by extreme traits must use ment that are most relevant to this paper. These epige-
arbitrary cutoff decisions to break a continuous netic findings are of fundamental importance for future
dimension and extract only its extreme aspects. This research, conceptualization, and treatment of personality
presents a strong argument against artificially disorders and psychiatric disorders in general.
separating the study of normal personality from that The field of epigenetics studies changes in gene
of maladapted personality. expression that are mitotically and/or meiotically
b) Personality disorder phenotypes defined solely by heritable and do not involve changes in the DNA
extreme traits are diagnostically imprecise as normal sequence. As shown by a number of authors (Impey et
individuals may be included as well. For example, a al. 2004, Fraga et al. 2005), the substrate for GxE
very shy person, who is otherwise well adapted and interaction is not directly DNA, but rather the activity
functional, would qualify for the diagnosis based status of a gene – which is regulated (“marked” or "pro-
only on extreme traits, but would not qualify if other
grammed") to be either active or silent depending on the
aspects of their personality (e.g., quality of
nature of the modulating stimulus. Such regulation of
adaptation) were considered. Indeed, a number of
authors agree that an independent evaluation of gene expression is referred to as "epigenetic".
impairment is required to diagnose personality Recall that DNA is wrapped around a complex of
disorders because high or low levels of traits are not histone proteins forming the chromatin globe, where
necessarily indicative of pathology (Cloninger et al. DNA is either accessible ("euchromatin") or
1993, Ignjatovic & Svrakic 2003, Trull & Durett inaccessible ("heterochromatin"). In order for DNA to
2005, Cloninger & Svrakic 2009). The core deficit is be expressed, chromatin must be in its active state, with
frequently described as a "failure to adapt" (Trull & nucleic acid sequences exposed to transcription factors.
Dutrrett 2005, Livesly 2007, Cloninger & Svrakic Recently, genome wide analysis of transcription factor
2009) which is created through person environment binding sites in vivo has been reported (Impey et al.
interaction and cannot be captured by diagnostic 2004). Transcription factors bind in a coordinated way
models based on extreme temperament traits alone. (“transcriptome”) to networks of genomic targets
To summarize, we agree with the expected involved in a particular adaptive response (“regulons”).
dimensionalization of personality disorders but argue i) The targets for transcription factors are not always gene
against using only extreme (“pathological’) traits to promoters, but also other transcription factors and/or
define the personality disorder phenotype but rather non-coding RNA, adding to the complexity and
continuous dimensions that account for individual multiplicity of possible GxE interaction outcomes

154
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

(Colvis et al. 2005). The long-term pattern of disappear (Champagne & Meaney 2007). Therefore,
epigenetically marked genes creates the epigenome (or, social environment in adulthood can alter phenotypic
as it were, a “programmed" genome) which ultimately features developed earlier in life through GxE
determines physical and behavioral phenotypic interaction. These results provide evidence for environ-
outcomes. ment x environment (ExE) interaction (Champagne
Chromatin configuration is considered to be the link 2008) in which already developed epigenetic features
between external environment and cellular DNA. change under the modifying influence of different
Extracelluar signals can reach the intracellular nucleus environmental conditions. Although the precise
(one example is the MAPK/ERK pathway) and change mechanism is still unclear, there is evidence that these
chromatin configuration, mostly through histone tail ExE effects are also mediated through DNA
(de) acetylation and (de) methylation. This in turn methylation (Weaver et al. 2004, 2005, 2006). Evidence
recruits DNA modulating enzymes and proteins (e.g., for ExE interaction provides important guidelines for
DNA methyl transferase – DNMT and methyl CpG psychotherapy theory and practice: psychotherapy
binding protein 2 - MeCP2) and regulates gene expression. should be rehabilitative, or, as it were, a human
In other words, environmental effects can influence the equivalent of “care” in animal models in order to be
transition between active and inactive chromatin and thus able to trigger epigenetic processes and lasting,
control gene expression (Colvis et al. 2005). neurobiological mechanisms for behavior changes.
Of note, epigenetic processes are involved in normal
development (e.g., cell differentiation, silencing of the Epigenetic modulation of personality traits
X chromosome. etc), pathological processes (e.g.,
cancer, schizophrenia, etc) and adaptive response to the There is a growing consensus that most, if not all,
environment (e.g., resistance to stress, personality traits, behavior traits develop epigenetically, through GxE
etc). Here we focus on adaptive environmental effects interaction (Tempelton 2006). If observations of
only. children are any indication, behavior development starts
out with few broad dispositions (e.g., distress, content)
which branch into successively more specific traits in
Epigenetic regulation of gene
response to specific environmental requirements and
expression is stable, but reversible
tasks (Tempelton 2006). This developmental branching
Early epigenetic DNA markings tend stabilize into follows a self-organizing pattern to satisfy multiple,
adulthood (Champagne 2008), but are reversible, even complex, and changing internal and external demands
in adulthood, through sustained environmental effects, (Cloninger & Svrakic 2009). In other words, behavior
which can be either chemical (Fraga et al. 2005) or development reflects not only GxE interactions with the
social / situational (Nithianantharajah & Hannan 2006, external environment (across different gradients of
Champagne 2008) or both. In rodents, for example, environmental specificity and complexity) but also
natural variations in maternal behavior are associated internal modifications by growing cognitive capacities
with high anxiety in novel situations, increased stress and self-awareness. Hence, manifest behavior involves
reactivity, and increased corticosterone response to numerous layers of epigenetic modulations, both
stress in the offspring (Colvis et al. 2005, Champagne vertical (i.e., successively more specific variants of a
2008). At the molecular level, this involves altered single trait) and horizontal (i.e., alternative adaptive
histone acetylation, increased DNA methylation, and solutions). These specific, socialized personality traits
reduced Nerve Growth Factor binding (NGFI-A is a are increasingly less guided by innate biogenetic
transcription factor), which results in decreased determinants and more by rational adaptive solutions,
expression of the glucocorticoid receptor (GR17) in the not nearly as stable over time or as predictable as their
hippocampus (Sato et al. 1998). These behavioral and antecedent biogenetic dispositions.
molecular effects are reversed by early postnatal cross- However, even complex social traits (e.g., character
fostering and/or by pharmacological manipulations in traits), expected to be modulated mostly by the
adulthood (Colvis et al. 2005, Champagne 2008). environment, show heritability estimates similar to
those observed for temperament (Gillespie et al. 2003).
Environment – Environment interaction: As we discuss later in the text, temperament and
guidelines for psychotherapy character traits either develop independently, but from
Epigenetic regulation of DNA expression is not the same underlying biogenetic dispositions, or
limited to early infancy or parenting. Postweaning character crystallizes from temperament, via conceptual
environments have also been shown to modulate transformation of temperament traits. Either way,
phenotypic features via epigenetic mechanisms biogenetic roots of temperament and character are
(Champagne 2008). As noted above, variations in strong and their heritability estimates are expected to be
maternal behavior influence vulnerability to stress in the similar.
offspring (Colvis et al. 2005, Champagne 2008). If these DNA susceptibility to environmental influences
offspring are placed in socially isolated or enriched continues throughout lifetime: MZ twins are roughly
postweaning housing conditions, the group differences concordant for the degree of DNA methylation and

155
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

histone H3 and H4 acetylation of genes in peripheral Maladaptation: the distinguishing feature

blood lymphocyte and other non-neural tissues at 3 of personality disorder
years of age, but at the age of 50 years there is a four-
fold difference (Fraga et al. 2005). It is unlikely that this Official classifications (DSM, ICD) do not define
magnitude of difference in gene expression between 50- personality disorders based on extreme traits, but
year-old MZ twins compared to 3-year-old MZ twins is require that the diagnosis be based on behaviors that
entirely accounted for by random stochastic effects and “deviate markedly from the expectations of the society”.
thus may also be related to the degree of discordance in In other words, deviation from the expected normative
environmental variables between the twins (Champagne standards (which is another way of saying poor
& Curley 2008), i.e., to environmental effects that alter adaptation) is a decisive diagnostic point. Accordingly, a
the expression of their otherwise identical genetic number of authors argue that an independent evaluation
makeup. Clearly, MZ twins may be genetically of impairment is needed because high or low standings
on behavior traits are not ipso facto pathological (Trull &
identical, but are very different epigenetically
Durrett 2005, Livesly 2007, Wakefield 2008, Cloninger
(especially with aging) which has major ramifications
& Svrakic 2009). For example, Livesley (2007) describes
for research.
personality disorders as a … "failure to solve adaptive life
Epigenetic modulations and GxE interaction have
tasks relating to identity or self, intimacy and attachment,
been implicated in adult psychiatric disorders in
and prosocial behavior”. (p. 203). This and similar
humans. For example, poor maternal care coupled with
suggestions recognize the fundamental aspect of the
over-protection in childhood (the co called
personality disorder diagnosis: the core deficit is
“affectionless control”) increase the risk of antisocial
maladaptation, i.e., deviant or poor adaptation to the
traits, addictions, attention deficit, OCD, anxiety
environment, not extreme behavior traits (although
disorders, and depression (Sato et al. 1998) in
extreme traits may have interfered with successful
adulthood. In contrast, good maternal care correlates
adaptation).
with high self-esteem, decreased trait anxiety, and
Maladaptation is frequently defined as a failure of
decreased salivary cortisol in response to stress
integrative functions of personality, i.e., those that carry
(Pruessner et al. 2004). Clearly, early care in humans
out adaptive processes (Wakefield 2008). These
sets the stage for a wide spectrum of disorders or mental
integrative, adaptive functions (referred to as character
health, via epigenetically modulated expression of
in this work) are not developmentally simultaneous with
initial genetic susceptibilities, much more so than
biogenetic dispositions to temperament, but emerge
previously believed. This has major implications for the
later, through person-environment interaction and
etiological understanding and prophylaxis of psychiatric
conceptual learning, as an adaptive "interface" to
disorders in general, as well as for the promotion of
optimize the fit between the internal needs and external
well-being and mental health since early childhood.
normative pressures (Cloninger et al. 1993, Svrakic et
al. 2009). In other words, these integrative, adaptive
Non-genomic (Lamarckian) inheritance functions determine whether an individual with very
of phenotypic features high Novelty Seeking develops into a well adapted
In addition to traditional genetic transmission, there researcher, explorer, or hobbyist, or into a poorly
is growing evidence of nongenomic inheritance of adapted antisocial personality.
adaptive phenotypic features. Acquired behaviors are Although several experts have recognized the central
incorporated into the epigenome, primarily via DNA importance of maladaptation for the concept of
methylation, and are subsequently transmitted onto personality disorder, only a few proposals to measure it
multiple generations trough incomplete removal of have been reported. The two noteworthy examples are
epigenetic marks during meiosis – all without changes in described below.
the DNA sequence (Fraga et al. 2005, Champagne 2008). The psychoanalytic concept of defense mechanisms
The above findings challenge some of the has been used for this purpose (Mulder et al. 1996,
methodologies traditionally used in personality research. 1999). Specifically, primitive defense mechanisms
Specifically, the non genomic, Lamarckian transmission centered around splitting, but not mature defenses
of acquired features makes it impossible to determine centered around sublimation, correlate with immaturity
the exact genetic roots of behavior traits by methods of and symptoms of personality disorders (Mulder et al.
quantitative genetics. If, as it certainly seems to be the 1996, 1999). The feasibility of assessing defense
case, an environmentally induced, acquired trait can be mechanisms for widespread, routine clinical use to
transmitted to the next generation by incomplete diagnose personality disorders is open to debate.
removal of chromatin remodeling without changes in Another alternative is to use operationalized
the DNA sequence - this trait will appear to be genetic behavior traits reflective of poor adaptation such as
in twin or family studies even though it is character traits of Self-deirectendness and
environmentally caused - because quantitative genetics Cooperativenss (Cloninger et al. 1993) as discussed
can not discriminate between non-genomic and genetic later in detail. Suffice here to say that low scores on
inheritance. Self-Directedness and Cooperativeness account for

156
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

fragmented self-concept, reduced ability to work, and typically chronic and unchanged over many years. At
reduced ability to get along with people typical of the peak of their deviant development, maladapted
personality disorders. In clinical research, these two personality eventually reaches the point of relative
character traits correlate consistently with symptoms of stability (Svrakic et al. 1996, Cloninger et al. 1997).
personality disorders (Svrakic et al. 1993). To the best Spontaneous change towards better adaptation is rare,
of our knowledge, no study has repudiated this finding because efforts to achieve a new developmental peak (a
out of cca. 350 cited replication reports in peer- new point of stability) are discouraged by a period of
reviewed journals. Of note, high (low) character scores initial instability associated with change. In dynamical
also correlate with mature (immature) defenses (Mulder systems’ parlance, maladapted patients go through a
et al. 1996) supporting the concurrent validity of both period of temporary decrease in fitness before reaching
concepts. a better adapted fitness peak in a wider epigenetic
landscape. This "U" shaped developmental curve
Maladapted personality: explains the treatment resistance seen with many
a self-organizing pathological system individuals with personality disorders and more
generally resistance to change observed in many normal
As described elsewhere (Svrakic et al. 1996, individuals.
Cloninger et al. 1997, Cloninger & Svrakic 2009),
personality development reflects a complex, self-
Disorders of personality or
organizing, adaptive effort to maximize the fit between
internal needs (defined by emotional temperament
disorders of adaptation?
traits) and external demands (defined by social norms). Adaptation to the environment is a process that
Even a maladapted personality reflects this self- begins very early in life, as early as in utero. As noted,
organizing effort which, in these cases, results in one’s failure to adapt can reflect either extreme,
suboptimal personal and social adaptive solutions. inflexible behavior dispositions or pathogenic environ-
Specifically, this adaptation in the "wrong direction” mental effects, or both. So far, however, only the person
can be caused by extreme, inflexible early behavior took the "blame" for maladaption - as implied in the
dispositions or by pathogenic environmental effects, or term "Personality Disorder".
both. On one hand, strong genetic dispositions influence With the above in mind, we suggest that the
one’s perception of the environment and one’s choice of diagnosis “Personality Disorder” be replaced by a more
activities and relationships (the so called "gene- appropriate diagnostic term "Adaptation Disorder”.
environment correlation") (Saudino et al. 1997). For There are several reasons to suggest this change in
example, individuals with constitutionally high nomenclature.
aggression have a tendency to engage in aggressive, First, the term “Adaptation Disorders” reflects more
antisocial behaviors and to surround themselves with accurately the etiological reality of the problem as it
antisocial, aggressive individuals. On the other hand, distributes the causality more fairly, between the person
pathogenic environmental factors modulate maladaptive
and the environment.
behavioral and cognitive styles (e.g., antisocial
Second, the term “Adaptation Disorder” reflects the
behaviors are frequently learned in one’s family –
mostly from the father). As shown in humans (Sato et essential feature of the disorder, i.e., chronic mal-
al. 1998) and in cross-fostering studies of animals adaptation to the environment. Most other psychiatric
(Maestripieri 2005, Champagne & Curley 2008), disorders are termed after dominant symptoms (e.g.,
environmental effects are powerful enough to modify Anxiety Disorders), or impaired function (e.g., Sleep
behavior traits in any direction: "good" environments Disorders), or after historically assumed etiology (i.e.,
(i.e., affectionate care) can ameliorate "bad" epigenomes Schizophrenia = "split mind"). No other disorder
(i.e., high disposition to aggression) and "bad" involves terminology that singles out a higher order
environments (i.e., those provoking fear) can entity, a general denominator of psychic life (i.e.,
pathologize "good" epigenomes (e.g., can turn personality), but not its affected component, i.e.,
confidence into fearfulness). In animal models, these impaired adaptation.
effects have been shown to be independent of parental Finally, it is clearly less stigmatizing to convey to
care and possible in adulthood as well, indicating that the patient the diagnosis of adaptation disorder vs.
the critical period for shaping lasting behavior features personality disorder. It is also more positive and
can be extended beyond early experiences and parenting motivating to direct treatment towards “adaptation
(Nithianantharajah 2006, Champagne 2008). With this problems” than towards “personality disorder".
in mind, adult social environments and rehabilitative An analogous solution to the one suggested here
psychotherapy may have significant corrective power in already exists for the group of Adjustment Disorders,
psychiatric treatments and correctional facilities trough which are defined as an acute failure to adapt
both GxE and ExE interaction. (Ignjatovic et al. 2003). In contrast, personality dis-
Personality disorder does not mature with time orders (or as we suggest Adaptation Disorders) reflect a
(except for some of the most drastic symptoms such as chronic failure to adapt (APA 2000). Incidentally,
violence or severe narcissism); other symptoms are Adjustment Disorders are not called "Acute Personality

157
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

Disorders” because the diagnosis here appropriately categorical system: traditional categories of personality
focuses on the failed function (i.e., failure to adjust) not disorder can be retrieved from dimensional data as
on the higher order entity of personality, although one’s unique configurations of temperament dimensions
acute response to stress may reflect either underlying (Cloninger & Svrakic 2009) (see below).
personality factors (e.g., one’s emotionality) or
objective aspects of stress, or both (much like the Proposed solution: The Psychobiological Model
etiology of chronic maladaptation). of Personality ("The Seven Factor Model")
A specifier to differentiate between four clinical
subtypes of mal-adapted behavior is based on the At present, the most frequently used dimensional
dominant emotional / temperament symptoms (Svrakic models of personality and personality disorder are the
et al. 1993). As advocated in prior work (Svrakic et al. Five Factor Model – FFM (Costa & McCrae 2006), the
2009), Adaptation Disorder can be sub-classified as: a) Psychobiological Model of Temperament and Character
anxious type, b) impulsive type, c) inhibited type, d) (Cloninger et al. 1994, Cloninger & Svrakic 2009) and
obsessive / anankastic type (each subtype reflects Dimensional Assessment of Personality Pathology –
relative dominance of one temperament trait in the DAPP (Livesley 2007). Out of these 3 models, the
overall temperament profile of the patient), and e) mixed Psychobiological model is most congruent with the
type (in cases when more than one of the four current epigenetic understanding of behavior
underlying traits is prominent). Analogously, in DSM development and ethiopathogenesis of personality
IV, Adjustment Disorders are divided into subtypes disorder. By no means do we suggest here that this
based on dominant clinical symptoms (depressed model is final and flawless. Further revisions, even
subtype, anxious subtype, disorder of conduct subtype, major ones, may be necessary to keep up with scientific
etc) (APA 2000). advances. At the time of its introduction in 1987 and
We suggest that the diagnosis of Adaptation 1993, it was certainly an avant-garde model appealing
Disorder be graded into “mild”, “moderate”, and to many for its visionary hypotheses. In the meantime,
“severe”. These quantifiers of severity are needed to many of these hypotheses have been tested and
reduce the problem of diagnosing only severe empirically supported in the US and worldwide (see
prototypical cases which is inherent in categorical (Cloninger & Svrakic 2009) and (Cloninger et al. 1994)
decisions. Most clinicians find it easier to choose from a for review).
scaled, more flexible, four-point graded choice (no case, Other two available dimensional models, the FFM
mild, moderate, severe) than between two categorical and the DAPP, conceptualize personality disorders
choices (“case” or “no case”). Character scores are based on extreme behavior dimensions only (DAPP),
applied to quantify the degree of severity of confound biological and adaptive personality processes
maladaptation (Cloninger et al. 1993, 2009, Svrakic et (FFM), do not consider maladaptation as an integral
al. 1993). diagnostic component of personality disorders (FFM)
and do not provide tools / ways to assess maladaptation
Traditional categories of personality in clinical work (DAPP and FFM).
disorders are not completely lost… In what follows, we briefly describe the
The above proposal does include traditional Psychobiological Model of Temperament and Character
categories of personality disorders (e.g., Narcissistic, (also known as the Seven Factor Model). For details on
Borderline, Schizoid, etc). These traditional personality see Cloninger and Svrakic (Cloninger & Svrakic 2009).
disorder categories convey vivid, but not always According to the Psychobiological Model, perso-
accurate clinical information about patients because nality structure consists of two distinct but interacting
they are not clearly separated (due to the overlapping components of temperament and character, each
DSM criteria). Indeed, most clinicians prefer to use reflecting different forms of learning and memory and
more general DSM Clusters (e.g., Dramatic, Eccentric, each providing unique developmental and functional
Fearful) to describe their patients. Moreover, individual significance in human life (Cloninger et al. 1993, 1997,
categories of personality disorder are neither discrete Svrakic et al. 1996). The model describes four
taxons nor permanent configurations of dimensions, but temperament dimensions (i.e., the biological "core" of
rather meta-stable, quasi-discrete combinations of personality) and three character dimensions (i.e., the
component processes that interact as expressions of a “adaptive interface” of personality), but allows for their
nonlinear dynamical system (Miettunen et al. 2006). behavioral variability and alternative activation based
Consequently, categorical diagnoses may alternate in on reciprocal causality between biogenetic factors,
time and across situations. For example, Narcissistic experience, and adaptation. Each dimension is defined
personality may at times alternatively present with a as a bipolar continuum from low to high expression,
dominant Antisocial or Histrionic façade (Svrakic & capturing both normalcy and extreme presentations.
McCallum 1991) and vice versa, making these Temperament and character dimensions and their facet
categorical diagnoses variable or, as it were, "a moving scales are presented in Table 1.
target". Our model provides for the continuity with the

158
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

Table 1. TCI Temperament and Character scales, subscales, and descriptors of high and low scores
I TEMPERAMENT High scorers Low scorers
HARM AVOIDANCE (HA)
HA1: worry and pessimism vs. uninhibited optimism pessimistic optimistic
HA2: fear of uncertainty vs boldness fearful daring
HA3: shyness with strangers vs outgoing shy outgoing
HA4: fatigability and asthenia vs energetic fatigable energetic
NOVELTY SEEKING (NS)
NS1: exploratory excitability vs. stoic rigidity exploratory reserved
NS2: impulsiveness vs. reflection impulsive deliberate
NS3: extravagance vs. reserve extravagant thrifty
NS4: disorderliness vs. orderliness irritable stoical
REWARD DEPENDENCE (RD)
RD1: sentimentality vs aloof sentimental cold
RD2: sociability vs reclusive open isolative
RD3: attachment vs. detachment warm detached
RD4: dependence vs. independence affectionate independent
PERSISTENCE (PS)
PS1: eagerness of effort vs procrastination industrious lazy
PS2: work hardened vs spoiled determined spoiled
PS3: ambitiousness vs underachieving enthusiastic underachiever
PS4: perfectionism vs pragmatism perfectionist pragmatist
II CHARACTER
SELF-DIRECTEDNESS (SD)
SD1: responsibility vs. blaming responsible blaming
SD2: purposefulness vs. lack of goal direction purposeful goalless
SD3: resourcefulness vs. helplessness resourceful passive
SD4: self-acceptance vs. self-striving confident self-conscious
SD5: spontaneity vs. lack of congruent second nature spontaneous conflicted
COOPERATIVENESS (CO)
C1: social acceptance vs. social intolerance tolerant intolerant
C2: empathy vs. social disinterest compassionate revengeful
C3: helpfulness vs. unhelpfulness empathic insensitive
C4: compassion vs. revengefulness helpful selfish
C5: fairness/principles vs. self-serving opportunism principled opportunistic
SELF-TRANSCENDENCE (ST)
ST1: self-forgetful vs. self-conscious acquiescent alienated
ST2: transpersonal identification vs Self-differentiation insightful concrete
ST3: spiritual acceptance vs rational materialism transpersonal avoiding
ST4: enlightened vs objective (optional subscale) creative conventional
ST5: idealistic vs practical (optional subscale) spiritual skeptical

Temperament traits and related behavior habits temperament traits correspond to the concept of
develop early in life, via associative learning and "endophenotypes” which are easier to study given their
synaptic strengthening which creates stable percepts, state-independence and temporal and phenomenological
affects, and procedural memory. Specifically, tempera- stability.
ment traits (Harm Avoidance, Novelty Seeking, Reward Character traits (Self-Directedness, Cooperativeness,
Dependence, and Persistence) are conceptualized as and Self-Transcendence) develop later in life, through
behavioral derivatives of primary emotions (i.e., fear, person-environment interaction, and involve conceptual
aggression, joy, persistence) and develop through and insight learning and higher cognitive processes of
associative learning (e.g., inhibition, activation, symbolic representation, logic, propositional memory,
continuation, perseveration of behavior) in response to etc. These traits are much more influenced by adaptive
specific environmental stimuli (e.g., danger, novelty, epigenetic mechanisms, and thus less stable in time,
reward, and frustrative non-reward, respectively) more culturally molded, and more prone to change with
(Cloninger & Svrakic 2009). Temperament dimensions the changing environment. Exact etiological origins of
have a number of dissociable biogenetic correlates (see character are difficult to identify with currently avai-
(Cloninger & Svrakic 2009) for review). If these lable methodologies. The two most likely scenarios are:
correlates are assumed to be genetic in nature, then i) character and temperament develop independently,

159
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

from the same underlying biogenetic dispositions, but From an adaptive perspective, the developing
via different learning styles, i.e., conceptual learning character traits serve as a conceptual “interface” which
(character) vs. associative learning (temperament); or, optimize adaptation of temperament (i.e., early
ii) character crystallizes from temperament, through emotionality) to the environment by reducing
cognitive processing and conceptual transformation of discrepancies between one's emotional needs and norm-
antecedent temperament traits during person - favoring social pressures (Cloninger & Svrakic 2009).
environment interaction. Each of these scenarios Some of the crucial differences between temperament
accounts for the observed high heritability of character and character are presented in Table 2.
traits (Gillespie et al. 2003).

Table 2. Key differences between temperament (associative or procedural learning) and character (conceptual or semantic
learning)
Temperament Character
Learning type Procedural Propositional
Awareness level Automatic Intentional
Memory form Percepts, concepts Procedures, propositions
Learning principles Associative conditioning Conceptual, Insight
Key Brain System Limbic system, Striatum Frontal/Temporal Neocortex, Hippocampus
Manifesting traits Habits, emotional Adaptive concepts, socialized
Primary emotions Secondary emotions
Associated emotions
(anger, fear, perseverance, attachment) (pride, compassion, ethics, altruism)
Role of subject
Passive, reproductive Active, constructive
in mental activity
Form of mental stimulus-response interactive networks (conceptual schema)
representation sequences varying additively in strength varying qualitatively in configuration

The Psychobiological Model is a quantitative model maladaptation, temperament to distinguish the dominant
of normal and maladaptive personality development as a symptomatic presentation. This is achieved as a two
self-organizing multidimensional dynamical system step process:
created through GxE interaction. Such systems are Step 1.
defined as complex adaptive system in biology Maladaptation is assessed by two essential features
(Cloninger et al. 1997). Based on a sophisticated of character - low Self-Directedness and low Coopera-
mathematical model (Svrakic et al. 1996, Cloninger et tiveness – which indicate a fragmented (immature) self-
al. 1997) we have shown that character traits develop as concept, problems with identity, reduced ability to
a nonlinear function of underlying temperament traits, work, and reduced ability to get along with people
socio-cultural factors, and random life events. (Svrakic et al. 1993). This is highly congruent with
In summary, character and temperament are proposals by other leading authors in the field (e.g.,
etiologically related (either share the same biogenetic Livesely) who conceptualizes the core deficit in
roots, or character develops from temperament), develop- personality disorder as a "three-level", multifaceted
mentally related (final character outcomes are limited by adaptive failure: at the individual level, personality
antecedent temperament traits) and functionally related disorder involves poorly developed or fragmented
(through bidirectional interaction, mature internalized representations of self and others, at the interpersonal
concepts regulated by character modify the significance level, it involves difficulties resolving attachment
and the salience of sensory percepts and affects regulated problems and developing the capacity for sustained
by temperament and vice versa). intimacy, and at the group level, it involves problems
The Temperament and Character Inventory (TCI) with prosocial behavior, altruism, and maintaining the
cooperativeness needed for effective social functioning
(Cloninger et al. 1994) is a family of tests, self-reports
(Livesley 2007). Livesley’s concepts of adaptive failure
and interviews, designed to measure temperament and
are captured by low scores on the TCI character traits of
character traits described above (Table 1). The TCI has
Self-Directedness and Cooperativeness (for descriptors
been validated across normative, clinical, and non
see Table 1).
clinical samples, cross culturally, and internationally
In a study of 136 patients (Svrakic et al. 1993) we
(Cloninger et al. 1994, Miettunen et al. 2006).
used logistic regression to predict the likelihood of
personality disorder diagnoses, based on the subject’s
Diagnosing maladaptive syndromes: Self- Directedness and Cooperativeness scores. With
a two step process lower character scores, the predictions became
Diagnosis of personality disorder includes biological markedly improved, reaching >85% for the lowest
and adaptive aspects: character is essential to assess scores. This finding suggests that some compound

160
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

variable, encompassing both Self-Directedness and Step 2.

Cooperativeness, may be a useful measure of After the diagnosis of personality disorder is
adaptation. One obvious such measure is a weighted established based on character, temperament is used for
sum of the two character scores, with weights assigned sub-classification and differential diagnosis. As shown
according to their observed correlations with personality in prior work (Svrakic et al. 1993), temperament traits
disorders. Since the two correlations are nearly equal efficiently discriminate DSM Clusters of personality
(0.10 and 0.11, respectively, as calculated from logistic disorders: Harm Avoidance is dominant in Cluster C
regression), the measure of adaptation in this case (“fearful” personalities), Novelty Seeking in Cluster B
reduces to a simple sum of the two (“impulsive” personalities), low Reward Dependence in
scores. With this in mind, and without introducing Cluster A (“asocial” personalities), and high Persistence
arbitrary categorical cutoffs, one can then speak of mild, in Obsessive Compulsive personality disorder (no DSM
moderate, and severe maladaptation within the existing Cluster). Moreover, individual categories of personality
measures of character, by dividing this compound score disorder are described by composite configurations
into three equal intervals. This makes the concept of involving all four temperament dimensions, i.e., by a
adaptation proposed in this work readily available for profile composed of high and/or low scores on Novelty
research and clinical use. Seeking, Harm Avoidance, Reward Dependence, and
An alternative way is to recognize that the Persistence (Svrakic et al. 1993). This dimensional
compound measure of adaptation is normally distributed approach preserves most traditional, categorical
in populations, as confirmed by our (Cloninger et al. diagnoses described in DSM and ICD which are still
1993, Svrakic et al. 1993) and other studies. One can occasionally used by clinicians. Namely, unique
then use the standard statistical procedure to extract combinations of high and/or low scores on temperament
regions of mild, moderate, and severe maladaptation dimensions create eight composite configurations which
from the character scores. This can also be implemented correspond to eight categorical DSM diagnoses of
in research and clinical practice in a straightforward personality disorders. For instance, Histrionic
manner. personality is characterized by high Novelty Seeking,
Whichever method one opts for, it is clear that a low Harm Avoidance, and high Reward Dependence
simple, practical measure of adaptation, both specific (Svrakic et al. 1993). Antisocial personality has the
and quantitative, can be construed from readily same profile except that Reward Dependence is low (see
available personality scores. Table 3 for more detail).

Table 3. Traditional Categories of Personality Disorders and TCI Dimensions

Categorical
TCI TEMPERAMENT DIMENSIONS
Diagnosis
Harm Novelty Reward
Persistence
Avoidance Seeking Dependence
Antisocial Low High Low High
Histrionic Low High High Low
Borderline High High Low Low
Narcissistic High High High High
Avoidant High Low High Low
Dependent Low Low High High
Schizoid Low Low Low Low
Obsessive/Anxious High Low Low High

"Composite" 2-step diagnosis: clinical and selfishness, lack of purpose, hostility, etc) is what
research advantages makes extreme behavior traits maladaptive.
This two-step, composite definition of the perso-
Consistent (or defining) and variable (or discri-
nality disorder phenotype has several advantages over
minating) features of personality disorder incorporated
other dimensional and categorical approaches. First, it
in the 2-step diagnostic algorithm are summarized in provides specific guidelines for research on both
Table 4. As already noted, extreme temperament biogenetic processes (procedural and associative
dimensions are associated with long term personal, learning and habit forming associated with tempera-
social, and/or occupational impairments described as ment) and adaptive processes (insight learning,
personality disorder only when accompanied by low conceptualization of self and environment associated
character traits. In other words, poorly developed with character) in the etiopathogenesis of maladap-
character (i.e., low responsibility, low resourcefulness, tation. Second, it provides treatment guidelines in

161
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

clinical work: affective instability, dysphoria, impulsi- disorder is defined as "the failure to develop a sense of
vity, and other symptoms associated with temperament self-identity and the capacity for interpersonal
are primarily treated with pharmacotherapy, whereas functioning that are adaptive in the context of the
maladaptive behavior styles and concepts (i.e., individual’s cultural norms and expectations" (37). This
character) are more amenable to psychotherapy. In other adaptive failure is manifested as an impaired sense of
words, medication is used to tone-down extreme self-identity (e.g., poor identity integration, poor
temperament traits which set the stage for a more integrity of self-concept, and low self-directedness)
efficient psychotherapy to facilitate character change. and/or as failure in interpersonal functioning (e.g., lack
Lastly, the 2-step diagnostic approach increases the of empathy, intimacy, low cooperativeness, and
specificity of diagnosis by reducing the likelihood of incomplete integration of representations of others).
diagnosing as personality disorder individuals with Clearly, this is highly in accord with our proposal that
extreme traits only or individuals with anxiety disorders, personality disorder be defined based on character traits
depression, or bipolar disorder, who are otherwise well of Self-Directedness (i.e., one’s Self concept) and
adapted and functional. Cooperativeness (i.e., one’s capacity for interpersonal
functioning) (Cloninger & Svrakic 2009, Svrakic et al.
Table 4. 2-Step Diagnosis Quantifiable (Dimensional) 1993, 2009).
Features of Personality Disorders (Adaptation Disorders) Specifically, the DSM-5 defines personality disorder
CONSISTENT FEATURES as the adaptive failure and deficits in one or both of the
low Self-directedness following two areas: A. Impaired sense of self-identity,
irresponsible, blaming and B. Failure to develop effective interpersonal
no mature goals functioning.
resourceless, helpless
poor self-esteem A. Impaired sense of self-identity evidenced by one
undisciplined or more of the following:
low Cooperativeness i Identity integration: poorly integrated sense of self
intolerant of others or identity, such as limited sense of personal unity
lack of empathy and continuity, shifting self-states, beliefs that the
unhelpful self presented to the world is a façade;
revengeful ii Integrity of self-concept: impoverished and poorly
unprincipled differentiated sense of self or identity - such as
VARIABLE FEATURES difficulty identifying and describing self attributes,
high persistence sense of inner emptiness, poorly delineated
(obsessive-compulsive symptoms only) interpersonal boundaries, definition of the self
low reward dependence (odd cluster only) changes with social context;
high novelty seeking (erratic cluster only iii Self-directedness: low self-directedness, such as
high harm avoidance (anxious cluster only) inability to set and attain satisfying and rewarding
personal goals, lacks direction, meaning, and
Addendum: DSM-5 draft purpose to life;
Since the original first version of this paper was All of the above aspects are adequately captured by
formulated, the DSM-5 draft has become available on low Self-Directedness in the TCI (see Table 1),
line, for comments and suggestions (APA 2010). We are including the psychodynamic concepts of identity and
pleased to report that the proposal to revise DSM-5 self-integration (as noted earlier, low scores on the TCI
(APA 2010) largely corresponds to the proposals in this Self-Directedness correlate highly with primitive
paper. Specifically, that the Work Group has recom- defense mechanisms, which typically underlie confused
mended a significant reformulation of the approach to identity and fragmented self concept) (Mulder et al.
the assessment and diagnosis of personality disorders, 1996, 1999).
most importantly the revised definition of personality
B. Failure to develop effective interpersonal
disorder, the provision for clinicians to rate dimensions
functioning as manifested by one or more of the
of personality traits, a limited set of personality types,
following:
and the ratings of the overall severity of personality
dysfunction. i Empathy: impaired empathic and reflective
capacity such as difficulty to understand the mental
DSM-5 definition of personality disorder. We are
states of others;
especially pleased that the Work Group provides a new
definition of personality disorder which is focused on ii Intimacy: impaired capacity for close relationships
the failure to adapt, rather than on extreme traits, such as inability to establish or maintain closeness
consistent with our proposals in earlier work (20, 31) and intimacy, to function as an effective attachment
and in this paper. In the DSM-5 draft, personality figure or to establish and maintain friendships;

162
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

iii Cooperativeness: failure to develop the capacity introversion (conceptually similar to low Reward
for prosocial behavior, e.g., failure to develop the Dependence and Social Withdrawal), and compulsivity
capacity for socially typical moral behavior, lack (conceptually similar to high Persistence and
of altruism; Obsessionality). Therefore, the decision to use six (and
iv Complexity and integration of representations not four) traits is somewhat puzzling. Nevertheless, the
of others: poorly integrated representations of selected traits may well be useful for describing
others, such as poorly related images of significant prominent features in people who do not qualify for a
others. diagnosis of personality disorder. For example, it is
Again, these interpersonal aspects are adequately often useful to describe someone as introverted,
captured by low Cooperativeness in the TCI (see Table compulsive, or disinhibited, etc, even if they do not
1), including the psychodynamic concepts of object- qualify for a personality disorder diagnosis. These terms
are all familiar to clinicians, which will make their
represenations (as noted earlier, low scores on
application easy. Unfortunately, there is no consensus in
Cooperativeness and Self-directedness correlate with
the field about the number or the content of traits
primitive defense mechanisms and partial object
describing personality. As a result, the proposed six
relations) (Mulder et al. 1996, 1999).
domains are tentative, labeled by the DSM-5 Work
Level of personality functioning. The Work Group
Group as "pending empirical validation" which means
provides an independent assessment for severity of the
that the number could be revised in the final version.
above adaptive impairments. Each of the two
The most questionable of the proposed traits is the
impairments is rated on a 5 point scale (0=absent,
domain of schizotypy. In accord with ICD 10, we
4=severe) and specific criteria are provided for these
advocate that Shizotypy with its cognitive, perceptual,
ratings. In our prior work, we have used scores on the
and behavioral impairments etiologically belongs to
TCI character dimensions to assess the severity of Axis I, where it could be studied more productively,
maladaptation (Svrakic et al. 1993, Cloninger & Svrakic together with other schizophrenia spectrum disorders.
2009).
Some of the popular models of personality (e.g.,
Livesley, Cloninger) can be used without major
Personality trait domains revisions to study personality disorders as defined by
In an attempt to dimensionalize personality DSM-5. This is even more the case should the six
disorders, the Work Group identifies 6 personality trait proposed domains / traits be reduced to four in the final
domains each comprised of several lower order, more version of the manual.
specific trait facets. The six dimensions are: 1. Negative
Emotionality (e.g., anxiety, depression, guilt/ shame, Personality disorder types
worry, etc.), 2. Introversion (e.g., withdrawal from
In order to keep continuity with the traditional
other people, restricted affective experience and expres-
categories of personality disorder, the Work Group
sion, limited hedonic capacity, etc), 3. Antagonism
recommends five personality disorder types (Antisocial
(antipathy toward others and an exaggerated sense of
/Psychopathic, Avoidant, Borderline, Obsessive-Com-
self-importance including narcissism, histrionism,
pulsive, and Schizotypal) and provides symptomatic
antisociality, etc), 4. Disinhibition (behavior is driven
by current internal and external stimuli, rather than by description for each type. As a novel feature, each type
past learning and consideration of future consequences is rated for its typicality on a dimension of graded
including impulsivity, recklessness, irresponsibility, membership ranging from 5=TYPICAL to 1=NO
etc), 5. Compulsivity (tendency to think and act MATCH. In addition, these personality types are also
according to a narrowly defined and unchanging ideal, defined as trait composites using the above described 6
and the expectation that this ideal should be adhered to personality domains / traits. A list specifying component
by everyone – includes perfectionism, perseveration, domains / traits for each type (e.g., component traits of
rigidity, orderliness), and 6. Schizotypy (odd or unusual Borderline Type are negative emotionality, schizotypy,
behaviors and cognitions, including both process (e.g., disinhibition, and antagonism). Finally, the extent to
perception) and content (e.g., beliefs) such as unusual which the domains / traits are descriptive of the
perceptions, beliefs, cognitive dysregulation, dissocia- particular type of personality disorder is rated on a 4
tion proneness, etc. For the four of the proposed six point scale (from not descriptive to extremely
personality domains empirical evidence is robust across descriptive). Of note, these five personality types are
different personality models (Livesley, Costa and also tentative, pending empirical validation.
McCrae, Cloninger) and they correspond to the four
fundamental dimensions of normal temperaments and CONCLUSION
personality disorders described earlier: negative emotio-
nality (conceptually similar to high Harm Avoidance The reader may get confused with all the different
and Neuroticism), disinhibition (conceptually names, contents, and descriptions of numerous
corresponds to high Novelty Seeking and Antisociality), personality traits in the literature. As noted, there is no

163
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

consensus in the field about the number or the content measure individual differences in learning and
of traits describing personality. Statistically, no factor personality development and structure. Statistical
analytical solution for personality traits is better than analyses, including factor analyses, were used to fine
others. However, the central problems in personality tune the psychometric properties of the TCI, not to
research is the question of natural units (“natural formulate the model or to determine its contents.
joints”) to define individual differences in personality, Psychometrically, the TCI provides a coherent body of
and this question is not answered by linear factor data to account for alternative approaches to personality
analysis. Linear statistical methods are inadequate to assessment, including the proposal for DSM-5 and has
study complex non linear dynamical systems such as predictive validity as good as or better than other
personality. Such methods take only a snap shot of available tests (Grucza & Goldberg 2007). The TCI
underlying interacting traits, can not discriminate measures are compared to those of Eysenck and
between distinct and sequential developmental strata in Zuckerman (Zuckerman & Cloninger 1996) in Table 5
personality structure (e.g., emotional vs. adaptive), can and to those of Costa and McCrae in Table 6 (Cloninger
not detect their dynamic interaction or their coordinated 2006).What is important for a clinician to know is that
activation by specific environmental stimuli. Rather, the same name does not always mean the same thing:
linear approach flattens out all behaviors into an different tests with the same name measure different
artificial plane and lumps all correlated behaviors things, as seen by examining the TCI correlates of the
(based on their average proximity to the axis) into Neuroticism measures of Eysenck, Zuckerman, and
higher order “factors”– regardless of their motivational Costa and McCrae. In addition, the constructs in the
priority or underlying learning processes, specific TCI provide a way to evaluate the personality disorder
eliciting stimuli, and unique motivational power. If as proposed for DSM-5 (illustrated in Table 7). The TCI
factor analysis is used, it should be used separately for character traits directly measure the features pivotal for
biogenetic and adaptive personality processes to avoid the definition and the diagnosis common to all
misleading results. TCI measures were developed personality disorders (Self-directedness and Cooperati-
rationally, based on a number of complementary veness). The temperament measures allow distingui-
perspectives – such as ethology, evolution, genetics, shing the types and traits proposed for DSM-5 as well.
neuroscience, behavior science, and biology to better

Table 5. Correlations (r x 100) between the Temperament and Character Inventory (TCI) scales and those of the
Eysenck Personality Questionnaire (EPQ-revised) and the Zuckerman-Kuhlman Personality Questionnaire ZKPQ
(correlations over .4 in bold, significant correlations only shown, n=207, adapted from Zuckerman & Cloninger, 1996)
HA NS RD P SD CO ST
EPQ Neuroticism 59 - 45
EPQ Extraversion -53 44 23 18
EPQ Psychoticism 41 -45 -29 -31 -42
EPQ Lie -21 25 34
ZKPQ Neuroticism 66 -49
ZKPQ Impulsive Sensation -39 68 -20 28
ZKPQ Hostility -27 -32 -60
ZKPQ Sociability -38 37 31
ZKPQ Activity -29 46 36
Legend: HA- Harm Avoidance; NS – Novelty Seeking; P – Persistence; RD – Reward Dependence
SD – Self-directedness; CO – Cooperativeness; ST – Self-Transcendence

Table 6. Correlations between the scales of Temperament and Character Inventory-Revised (TCI-R) and the NEO-PI-
Revised (correlations over .4 in bold, significant correlations only shown, multiple correlation also shown, n=662,
adults in USA)
HA NS RD P SD CO ST mR
NEO Neuroticism 63 -20 -62 -28 75
NEO Extraversion -55 40 52 40 25 22 77
NEO Openness -25 43 25 37 54
NEO Conscience -26 -34 51 41 70
NEO Agreeability -23 40 31 61 20 66
mR 76 65 68 60 67 65 45
Legend: HA- Harm Avoidance; NS – Novelty Seeking; P – Persistence; RD – Reward Dependence
SD – Self-directedness; CO – Cooperativeness; ST – Self-Transcendence/ NEO – Neuroticism, Extraversion, Openness

164
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

Table 7. Relationship of TCI dimensions to DSM-V proposal of Personality Trait Domains

DSM-V Proposal SD CO ST HA NS RD PS
General Features
Self-identity impaired Low
Interpersonal function poor Low
Types
Antisocial Low Low Low High Low
Avoidant Low Low High Low High
Borderline Low Low High High Low
Obsessive Low Low High Low Low High
Schizotypal Low Low High High Low Low
Traits
Negative Emotionality Low Low High Low
Introversion Low High Low Low
Antagonism Low Low Low High Low
Disinhibition Low Low High
Compulsivity High Low High
Schizotypy Low High
Legend: HA- Harm Avoidance; NS – Novelty Seeking; P – Persistence; RD – Reward Dependence
SD – Self-directedness; CO – Cooperativeness; ST – Self-Transcendence

REFERENCES psychiatry. 9th ed. London: Lippincott, Williams and

Wilkins; 2009 (in press)
1. APA. Diagnostic and statistical manual of mental 11. Colvis CM, Pollock JD, Goodman RH, et al. Epigenetic
disorders. 4th ed revised. DSM IV –TR APA: Washington, mechanisms and gene networks in the nervous system. J
DC; 2000. Neurosci 2005; 25:10379–10389.
2. APA. DSM 5 draft. 2010. www.dsm5.org 12. Costa P, McCrae R. NEO PI R. Psychological Assessment
3. Bernstein DP, Iscan C, Maser J, and the Boards of Resources Inc. 2008.
Directors of the Association for Research in Personality 13. Fraga MF, Ballestar E, Paz MF., et al. Epigenetic
Disorders and the International Society for the Study of differences arise during the lifetime of monozygotic twins.
Personality Disorders. Opinions of personality disorder Proc. Natl. Acad. Sci. 2005;102: 10604–10609.
experts regarding the DSM IV personality disorders 14. Gillespie AN, Cloninger CR, Hath CA, Martin GN. The
classification system, J Pers Dis. 2007; 21:536–551. genetic and environmental relationship between
4. Champagne AF, Meaney JM. Transgenerational effects of Cloninger's dimensions of temperament and character.
social environment on variations in maternal care and Person Ind Diff. 2003; 35:1931-46
behavioral response to novelty, Behav Neurosci. 2007; 15. Grucza RA, Goldberg LR. The comparative validity of 11
121:1353–63. modern personality inventories: predictions of behavioral
5. Champagne, F.A., Curley, J.P. Epigenetic mechanisms acts, informant reports, and clinical indicators. J Pers
mediating the long-term effects of maternal care on Assess. 2007; 89:167-187.
development, Neurosci Biobehav Rev 2008. Jan 18 (Epub 16. Ignjatovic TD, Svrakic, D. Western Personality Models
ahead of print) Applied in Eastern Europe: Yugoslav data. Compr
6. Cloninger CR, Svrakic D, Przybeck T. A psychobiological Psychiatry, 2003 (44):51-59.
model of temperament and character. Arch Gen 17. Impey S, McCorkle RS, Cha-Molstad H,et al. Defining the
Psychiatry 1993; 50:975–990. CREB Regulon: A Resource Genome-Wide Analysis of
7. Cloninger, C.R., Przybeck, T.R., Svrakic, D., Wetzel, R: Transcription Factor Regulatory Regions Cell, 2004;119:
The Temperament and Character Inventory (TCI): A guide 1041–1054.
to its development and use, Washington University School 18. International Classification of Diseases, 10th revision,
of Medicine, Department of Psychiatry, St. Louis, MO, WHO, Geneva, 1992
1994 19. Johnson-Pynn J, Fragaszy DM, Cummins-Sebree S.
8. Cloninger, CR, Svrakic, N. M., Svrakic, D. Role of perso- Common Territories in Comparative and Developmental
nality in the development of mental order and disorder. Psychology: Quest for Shared Means and Meaning in
Development and Psychopathology. 1997;9:881-906. Behavioral Investigations. Int J Comparative Psychology.
9. Cloninger CR, Personality as a Dynamic Psycho- 2003;16:1-27.
biological System. In: (ed. Widiger TA, Simonsen E, 20. Livesley J. A framework for integrating dimensional and
Sirovatka PJ, et al.) Dimensional Models of Personality categorical classifications of personality disorders. J
Disorders: Refining the Research Agenda for DSM-V. Personal Disord 2007; 21:199–224.
Washington, D.C., American Psychiatric Press, 2006, S. 21. Maestripieri D. Early experience affects the inter-
73-76 generational transmission of infant abuse in rhesus
10. Cloninger CR, Svrakic D. Personality Disorders. In: (eds. monkeys. Proc. Natl. Acad. Sci. USA 2005;102:9726–
Sadock BJ, Sadock VA) Comprehensive textbook of 9729.

165
 Dragan M. Svrakic & Robert C. Cloninger: EPIGENETIC PERSPECTIVE ON BEHAVIOR DEVELOPMENT, PERSONALITY,
AND PERSONALITY DISORDERS Psychiatria Danubina, 2010; Vol. 22, No. 2, pp 153–166

22. Miettunen J, Kantoja¨rvi L, Veijola J, Ja¨rvelin M-R, 31. Svrakic NM, Svrakic DM, Cloninger CR. A General
Joukamaa M. International comparison of Cloninger’s Quantitative Theory of Personality Development:
temperament dimensions. Pers Indiv Diff. 2006; 41:1515– Fundamentals of a Self-Organizing Psychobiological
1526. Complex, Development and Psychopathology.
23. Mulder R T, Joyce PR, Sullivan PF, Bulik CM, Carter FA. 1996;8:247-272.
The relationship among three models of personality 32. Svrakic D, Lecic-Tosevski D, Divac Jovanovic M. DSM
psychopathology: DSM-III R personality disorder, TCI Axis II: personality disorders or adaptation disorders?
scores and DSQ defences. Psycholl Med. 1999; 29:943- Current Opinion Psych 2009;29:111-117
951. 33. Templeton A: Population Genetics and Microevolutionary
24. Mulder RT, Joyce PR, Sellman JD, et al. Towards an Theory, Wiley, New Jersey, 2006
understanding of defense style in terms of temperament 34. Trull T, Durrett CA: Categorical and dimensional models
and character. Acta Psychiatr Scand 1996; 93:99–104. of personality disorder. Annu. Rev Clin Psychol. 2005;
25. Nithianantharajah J, Hannan A.J. Enriched environments, 1:355–80.
experience –dependent plasticity and disorders of the 35. Wakefield JC. The perils of dimensionalization:
nervous system. Nat Rev Neurosci 2006;7:697–709. challenges in distinguishing negative traits from
26. Pruessner JC, Champagne F, Meaney MJ, Dagher A. personality disorders. Psychiatr Clin North Am. 2008;
Dopamine release in response to a psychological stress in 31:371–393.
humans and its relationship to early life maternal care: a 36. Weaver CI, Cervoni N, Champagne FA, D’Alessio CA,
positron emission tomography study using [11C] Sharma S, Seckl RJ, Dymov S, Szyf M, Meaney JM.
raclopride, J. Neurosci. 2004; 24: 2825–2831. Epigenetic programming by maternal behavior. Nat
27. Sato T, Sakado K, Uehara T, Narita T, Hirano S, Nishioka Neurosci, 2004;7:847–54.
K, Kasahara Y. Dysfunctional parenting as a risk factor to 37. Weaver CI, Champagne FA, Brown ES, Dymov S, Sharma
lifetime depression in a sample of employed Japanese S, Meaney JM, Szyf M. Reversal of maternal programming
adults: evidence for the ‘affectionless control’ hypothesis, of stress responses in adult offspring through methyl
Psychol. Med. 1998; 28: 737–742. supplementation: altering epigenetic marking later in life,
28. Saudino KJ, Pedersen NL, Lichtenstein P, et al. Can J Neurosci 2005;25:11045–54.
personality explain genetic influences on life events? J 38. Weaver CI, Meaney JM, Szyf M. Maternal care effects on
Pers Soc Psychol. 1997;72:196–206. the hippocampal transcriptome and anxiety-mediated
29. Svrakic D, McCallum K. Antisocial Behavior and Perso- behaviors in the offspring that are reversible in adulthood.
nality Disorders. Am J Psychother. 1991;45:181-197. Proc Natl Acad Sci USA 2006;103:3480–85.
30. Svrakic D, Whitehead C, Przybeck TR, Cloninger CR. 39. Zuckerman M, Cloninger CR. Relationship between
Differential diagnosis of personality disorders by the Cloninger's, Zuckerman's and Eysenck's dimensions of
seven-factor model of temperament and character. Arch personality. Personality and Individual Differences. 1996;
Gen Psychiatry. 1993; 50:991–999. 21:283-285.

Correspondence:
Dragan M. Svrakic, MD PhD
Washington University School of Medicine, Department of Psychiatry,
St Louis, MO, USA
E-mail: [email protected]

166