critique paper

Stability and Change in Personality Disorder Lee Anna Clark

University of Iowa

ABSTRACT—The standard view of personality disorder is

that it is a maladaptive expression of personality traits,

which are relatively stable and unchanging. Thus, person-

ality disorder has been considered to have its roots in child-

hood and adolescence, to persist in adulthood, and to be

difficult to change. However, recent research has challenged

this view, revealing that personality continues to change,

albeit more slowly, well into adulthood, and that the

maladaptive manifestations of personality disorder are

much less stable than previously believed. These research

findings are described, and factors that influence stability

and change in personality disorder are discussed. The

emerging view of personality disorder has important impli-

cations for diagnosis, assessment, and treatment of person-

ality pathology.

KEYWORDS—stability; change; personality disorder

The common view of personality—at least in Western cultures—

is that it is relatively unchanging. For example, meeting friends

at a class reunion with whom we lost touch 10, 20, 30 years ago,

we more typically think, ‘‘She hasn’t changed a bit,’’ than think,

‘‘It feels like I’m talking to a stranger; I can’t believe this is the

same person I knew in high school.’’ This view of personality

stability has long been shared by many personality researchers,

who were apt to cite William James’ (1890/1950) famous state-

ment, ‘‘by the age of 30, the character has set like plaster, and

will never soften again’’ (p. 121). Building on this view, per-

sonality disorder (PD) in the official Diagnostic and Statistical

Manual of Mental Disorders (DSM-IV; American Psychiatric

Association, 2000, p. 689) is defined, in part, as personality

traits that are ‘‘stable and of long duration,’’ as well as ‘‘inflexible

and pervasive.’’ Consequently, PD is characterized by psycho-

social dysfunction, including, for example, having conflict-rid-

den or unstable social and marital relationships, or a notable

absence of such relationships; showing poor judgment and de-

cision making or a marked inability to make decisions; and

having legal problems and employment difficulties.

Although PDs are defined as maladaptive personality traits,

officially they are assessed and diagnosed categorically; that is,

individuals either meet or do not meet criteria for one or more

DSM-IV PDs. Each PD diagnosis represents a pervasive pattern

of one or more maladaptive traits, which are illustrated by seven

to nine specific criteria. For example, ‘‘distrust and suspi-

ciousness’’ are the defining traits of paranoid PD, and ‘‘persis-

tently bears grudges’’ (p. 694) is a specific criterion reflecting

one or both of these traits. For a PD diagnosis, one must manifest

a certain number (typically four or five) of a disorder’s criteria

over time and across situations. (A list of the 10 DSM-IV PDs and

the pervasive patterns defining them is presented in Table 1.)

Measurement and conceptual difficulties with the current

categorical diagnostic system are widely acknowledged (Clark,

2007). Nonetheless, although maladaptive personality traits

have proved to have substantial reliability and validity (Widiger

& Frances, 2002), there is considerable reluctance to replace

the current criterion-based, categorical PD diagnostic system

with one that is based directly on trait dimensions. There are

various reasons for this reluctance, including the fact that in-

formation on the trait set that constitutes personality pathology is

incomplete, that most trait assessments inadequately reflect

extreme psychosocial dysfunction, and that how such a system

would inform treatment remains unclear. Thus, PD researchers

are exploring integrating personality trait and PD research, and

it is in this context that recent research findings have challenged

a simplistic view of personality and PD stability.

NORMAL-RANGE PERSONALITY STABILITY AND

CHANGE

There are multiple ways to consider personality stability and

change (Roberts & Mroczek, 2008). This article focuses on

three: (a) Mean-level/developmental stability/change indexes

the extent to which average trait levels change within a popu-

lation cohort; (b) rank-order stability/change reflects the extent

Address correspondence to Lee Anna Clark, Department of Psy- chology, University of Iowa, E11 SSH, Iowa City, IA 52242-1407; e-mail: [email protected].

C U R R E N T D I R E C T I O N S I N P S Y C H O L O G I C A L S C I E N C E

Volume 18—Number 1 27Copyright r 2009 Association for Psychological Science

to which individuals maintain their relative trait levels; (c) and

diagnostic stability/change refers to whether or not individuals

meet criteria for the same diagnostic categories across time. If

diagnostic stability/change is assessed dimensionally (e.g., the

number of criteria met), then diagnostic stability/change is the

same as rank-order stability/change.

Recent meta-analyses of normal personality (Roberts &

DelVecchio, 2000; Roberts, Walton, & Viechtbauer, 2006) re-

vealed, surprisingly, that there is moderate mean-level change

through most of the life span. However, regarding rank-order

stability, change is modest to moderate throughout childhood

and adolescence, whereas as early as the 20s, trait rank-order

largely stabilizes, and this stability continues to increase slowly

with each successive decade until age 50. From the latter per-

spective, James’ ‘‘set like plaster’’—and our commonsense

feeling that personality is largely stable—appears valid. That is,

although the frame of reference (i.e., mean level) is changing

moderately, within that framework, adult personalities are

fairly—and increasingly—stable, beginning in the early 20s. It

is against this backdrop that I examine PD stability and change.

STABILITY AND CHANGE IN PD

Reports of PD stability vary depending on whether they examine

stability categorically (i.e., whether or not individuals were di-

agnosed with the same PD or PDs across time) or dimensionally

(i.e., rank-order stability measured using either the number of

PD diagnostic criteria met or personality-pathology-scale

scores). Studies examining categories find considerable diagnostic

change regardless of whether the retest interval is short (1 week),

moderate (up to 6 months), or somewhat longer (3 years), whereas

studies examining dimensions yield results similar to those for

normal-range personality: Short-term stability is quite high,

whereas more change is seen over longer intervals (Clark, 2007).

This pattern implicates categorical measurement error; that is, there

being little more diagnostic change in 3 years than in 1 week likely

reflects simply that many patients were diagnosed with (for example)

five criteria on one occasion and four on the next, which would

represent a diagnostic change (from meeting to not meeting criteria

for the disorder). However, when measured dimensionally, this

amount of change (i.e., from five criteria to four) is relatively minor.

More recent longitudinal PD studies provide assurance that these

results are not due simply to interrater unreliability (i.e., scoring

differences between interviewers; Clark, 2007; Grilo et al., 2004).

Further analyses indicate that PD instability is not due simply

to changes across the diagnostic boundary as just described.

Rather, it appears to represent actual decreases in PD mani-

festations. For example, in the Collaborative Longitudinal

Personality Study (CLPS), a large sample of patients diagnosed

with PD met on average 71% of the criteria of their primary PD at

baseline (e.g., 6.4 of 9, or 5.7 of 8 criteria), and the proportion

dropped steadily to 55%, 48%, and 42% at 6-month, 1-year, and

2-year follow-ups, respectively (Grilo et al., 2004). Moreover,

32% of the sample met two or fewer criteria the entire second

year. Similar findings of decreasing PD pathology over time have

been reported in other longitudinal studies, including a 6-year

study of borderline PD patients, a 4-year college-student study,

and a 20-year community epidemiological study, which found a

linear decline in PD manifestations from age 9 to 27 (see Clark,

2005, 2007, for reviews).

Given that age-related decreases in negative traits (e.g.,

neuroticism) also are found in normal personality (Roberts et al.,

2006; Roberts & Mroczeck, 2008) and given overlap between PD

criteria and normal personality traits, the question arises whe-

ther change in PD manifestations represents simply the same

phenomenon as change in normal personality traits or whether

additional processes are involved. If change in personality and

PD are due to the same processes, we might expect the degree

and age-ranges of change to be similar for PD criteria and

personality traits, and that PD criteria would show the same

rank-order stability as personality traits. Regarding the former

expectation, too few data have been published to examine the

TABLE 1

The 10 DSM-IV Personality Disorders and the Pervasive Patterns That Define Them (American Psychiatric Association, 2000)

Personality disorder Pervasive patterns

Paranoid Distrust and suspiciousness of others such that their motives are interpreted as malevolent

Schizoid Detachment from social relationships and a restricted range of expression of emotions in interpersonal settings

Schizotypal Social and interpersonal deficits marked by acute discomfort with, and reduced capacity for, close relationships, as

well as by cognitive or perceptual distortions and eccentricities of behavior

Antisocial Disregard for and violation of the rights of others occurring since age 15 years

Borderline Instability of interpersonal relationships, self-image, and affects, and marked impulsivity

Histrionic Excessive emotionality and attention seeking

Narcissistic Grandiosity (in fantasy or behavior), need for admiration, and lack of empathy

Avoidant Social inhibition, feelings of inadequacy, and hypersensitivity to negative evaluation

Dependent Excessive need to be taken care of that leads to submissive and clinging behavior and fears of separation

Obsessive-compulsive Preoccupation with orderliness, perfectionism, and mental and interpersonal control, at the expense of flexibility,

openness, and efficiency

28 Volume 18—Number 1

Stability and Change in Personality Disorder

issue closely (Harpur & Hare, 1994; Johnson et al., 2000), so it is

an important topic for future research.

Regarding rank-order stability, however, some preliminary

data do exist. Specifically, Grilo et al. (2004) reported that rank-

order stability over 2 years for the dimensional measure ‘‘number

of criteria met,’’ averaged across four PDs, was moderately high,

whereas by 4 years it had dropped notably. In the community

study (Johnson et al., 2000), the average retest correlation for 11

specific types of PD manifestation from age 14 to 16 was quite

similar to that found in normal-range traits for ages 12 to 18.

However, from age 16 to 22, the average stability coefficient

dropped rather than increased as normal personality traits do.

Thus, in both samples, dimensionally assessed PD manifesta-

tions clearly showed greater change compared to normal-range

personality traits, raising the question of whether this is because

personality is more unstable in individuals with PD than in those

without or because PD manifestations are more unstable than

personality traits per se.

Morey et al. (2007) provide an initial answer to this question:

The stability of personality traits in the same patient sample

examined by Grilo et al. (2004), over the same 2 years, was high

regardless of whether the measures were designed to assess per-

sonality pathology or normal-range personality. Further, trait

stability remained at the same high level in the 4-year data, so it

appears that PD manifestations are less stable than personality

traits in both patient and community samples, and not that indi-

viduals with PD have less stable personalities than those without.

This, in turn, raises questions regarding the factors involved in

personality and PD stability and change, particularly in the

differential stability of personality traits versus PD manifesta-

tions. My focus being on stability and change in PD, readers

interested in the stability and change of normal personality are

referred elsewhere (Caspi, Roberts, & Shiner, 2005; Fraley &

Roberts, 2005; Roberts, Wood, & Caspi, 2008).

FACTORS INFLUENCING PD STABILITY AND CHANGE

Personality disorder is notoriously difficult to treat (Critchfield &

Benjamin, 2006), therapy dropout rates are high (e.g., McFar-

land & Klein, 2005), and typical response to pharmacotherapy

is modest and limited in scope (Morana & Camara, 2006).

Although extended, newly developed psychotherapies have

documented efficacy (Gabbard, 2000), little change is found

with outpatient ‘‘treatment as usual’’ (e.g., Bohus et al., 2004),

the context of most PD-stability studies. Thus, in contrast to the

relatively sharp picture that has emerged for normal personality,

our understanding of PD change is still fuzzy.

Relations With Other Disorders

Stability and change in PD manifestations must be considered in

the context of co-occurring mental disorders. Multiple studies

document that a PD diagnosis typically co-occurs with one or

more other PD diagnoses, as well as with a wide array of de-

pressive, anxiety, eating, and disruptive disorders (Clark, 2007),

so PD stability might be reduced by changes in these other

disorders. However, research into interrelations between the

course of PD and that of other disorders has found that im-

provement in PD generally is more likely to lead to improvement

in other disorders than vice versa (Clark, 2005) and that the co-

occurrence of another disorder tends to increase PD stability. For

example, in the community study cited earlier, when adolescent

personality pathology co-occurred with another mental/behavioral

disorder, the likelihood of the personality pathology continuing or

worsening in young adulthood increased by up to 19 times.

Stability of Functioning

Several recent studies have found that PD-based dysfunction is

quite stable, clearly more so than diagnostic criteria are (e.g.,

Clark, 2007; Skodol et al., 2005). Moreover, although baseline

PD manifestations (scored dimensionally) and personality trait

scores related roughly equally to functioning at both baseline

and over 2 years, over 4 years personality traits predicted

functioning more strongly than did PD manifestations (Morey et

al., 2007). Thus, the stability of dysfunction appears linked more

closely to personality traits than to PD manifestations, which are

somewhat less stable. This may be in part because adaptive

functioning takes time to develop. That is, a decrease in indi-

viduals’ PD manifestations does not translate immediately into

adaptive outcomes, such as getting and holding a job or estab-

lishing and maintaining strong interpersonal relations with

friends and family.

Variability in PD Criteria

As described earlier, PD diagnosis currently requires manifes-

tation of a certain number of diagnostic criteria. Importantly, PD

criteria vary in the degree to which they tap acute, dysfunctional

behaviors (e.g., recurrent suicidal behavior) that resolve in

shorter time periods versus more long-standing maladaptive

characteristics (e.g., preference for solitary activities), and

several studies have found significant variability in criterion

stability. For example, less stable criteria include odd behavior

and constricted affect (schizotypal PD), self-injury, and behav-

iors to avoid abandonment (borderline PD), avoiding interper-

sonal jobs and potentially embarrassing situations (avoidant

PD), and miserly and strict moral behaviors (obsessive-com-

pulsive PD), whereas paranoid ideation (schizotypal PD),

affective instability and anger (borderline PD), feeling inade-

quate and socially inept (avoidant PD), and rigidity and diffi-

culty delegating (obsessive-compulsive PD) are more stable

(Clark, 2007). Thus, the lower stability of PD manifestations

(compared to traits) may be due to changes in those criteria that

represent acute symptoms, which respond to treatment (e.g.,

suicidal behavior) or diminish with maturation or stress reso-

lution (e.g., impulsivity or unassertiveness, which also diminish

Volume 18—Number 1 29

Lee Anna Clark

with age in normal samples), whereas the observed degree of PD

stability is based in criteria that reflect more basic temperament

traits (e.g., chronic anger, stress reactivity). This intriguing

possibility clearly warrants further research.

Another way PD criteria vary is in the extent to which they

directly reflect the particular pervasive maladaptive trait pattern

that defines the diagnosis. For example, the criterion ‘‘has a

grandiose sense of self-importance’’ (p. 717) is directly related to

the defining pattern of narcissistic PD, ‘‘grandiosity in fantasy or

behavior’’ (p. 717), whereas the criterion ‘‘unwilling to get in-

volved with people unless certain of being liked’’ (p. 721) relates

less directly to one of the defining patterns of avoidant PD,

‘‘social inhibition’’ (p. 721). Studies that have assessed per-

sonality traits in addition to—and independently of—PD cri-

teria and examined relations between them have shown that

change in personality traits predicts PD change but not vice

versa (Warner et al., 2004). Thus, for example, individuals could

become more willing to get involved with people regardless of

their certainty of being liked without this change affecting their

overall level of social dominance, but if their level of social

dominance changed, they most likely would become more

willing to get involved with people.

A NEW, EMERGING VIEW OF PD

Due to these recent findings, researchers are beginning to think

differently about the stability of PD, and to view PD not as ‘‘set in

plaster’’ but as comprised of traits—some maladaptively ex-

treme—that show change within relative stability, together with

related, but less stable, dysfunctional behaviors. The latter—

which likely account for the lower stability of PD compared to

trait assessments—may be either less-direct trait manifestations

(as in the above example of interpersonal avoidance and social

inhibition), or they may develop as defensive or compensatory

coping behaviors in response to stress—both external events

and stress that is self-created by one’s own dysfunction. Relat-

edly, the persistent dysfunction of individuals with PD may be

manifestations of more stable personality traits that—being

extreme—take longer to normalize, even if they change in the

same way and at the same rate as do normal-range traits.

This emerging view of PD suggests that it would be fruitful to

revise PD diagnosis to distinguish more acute manifestations

from personality traits, and several strikingly similar proposals

have been made (Clark, 2007). Although they differ in partic-

ulars, they all suggest that PD diagnosis has two components:

(a) diagnosing disorder, that is, assessing the level of psycho-

logical and social/interpersonal dysfunction (e.g., integrated

sense of self; family and occupational stability) and its acute

manifestations (e.g., ideas of reference, aggression, hyperper-

fectionism, suicidality); and (b) describing individuals’ per-

sonality traits. Over time, disorders might come and go, much as

in depression, whereas personality traits would be expected to

change less, and less rapidly.

This approach also suggests that initial treatment should focus

on more acute, changeable PD manifestations, whereas broader,

long-lasting outcomes may need to be effected through personal-

ity-trait change. Interestingly, one of the most well-established PD

treatments, dialectical behavior therapy, may succeed because it

is congruent with this emerging view of PD. Specifically, it first

targets specific behaviors that are life threatening, interfere with

treatment, and/or lower one’s quality of life, and later shifts its

focus to developing adaptive life skills (e.g., anger management)

and to resolving longstanding, problematic interpersonal dynam-

ics (likely based, at least partly, on personality traits).

CHALLENGES AND FUTURE RESEARCH DIRECTIONS

The convergence of this emerging view of PD and the impending

revision of the DSM presents significance challenges and great

opportunities. To facilitate a major revision in PD diagnosis,

assessment researchers need to delineate the components of

PD—underlying personality traits, acute dysfunctional behav-

iors, and persistent psychological and social/interpersonal

dysfunction. Psychopathology researchers need to learn how

these components are interrelated structurally and, together

with personality researchers, they need to understand better the

processes that underlie both the relative stability of personality

traits and the greater changeability of acute dysfunction.

Applied research also is needed, to learn how this reconceptu-

alization can be utilized in clinical settings, perhaps by focusing

first on changing the least stable PD elements. Finally, we need

to learn more about normative personality-change processes and

consider how they may inform treatment of patients with PD.

Recommended Reading Caspi, A., Roberts, B.W., & Shiner, R.L. (2005). (See References). A

comprehensive review of current knowledge about the develop-

ment of personality structure; behavioral genetics of personality;

the normative timeline of life-span personality development;

and the interplay of personality with social relations, social status,

and health.

Robins, C.J., Schmidt, H., & Linehan, M.M. (2004). Dialectical be-

havior therapy: Synthesizing radical acceptance with skillful

means. In S.C. Hayes, V.M. Follette, & M.M. Linehan (Eds.),

Mindfulness and acceptance: Expanding the cognitive-behavioral tradition (pp. 30–44). New York: Guilford. An overview of the five core elements of dialectical behavior therapy focused on the

characteristics that most distinguish this approach from more

standard behavioral treatments.

Shea, M., Stout, R.L., Yen, S., Pagano, M.E., Skodol, A.E., Morey, L.C.,

et al. (2004). Associations in the course of personality disorders

and Axis I disorders over time. Journal of Abnormal Psychology, 113, 499–508. An exploration of the longitudinal interrelations of four PDs (borderline, schizotypal, avoidant, and obsessive-com-

pulsive) with the mood and anxiety disorders, including consid-

eration of causation; also provides multiple references to other

publications of the Collaborative Longitudinal Personality Study.

30 Volume 18—Number 1

Stability and Change in Personality Disorder

Widiger, T.A., Trull, T.J., Clarkin, J.F., Sanderson, C., & Costa, P.T., Jr.

(2002). A description of the DSM-IV personality disorders with the five-factor model of personality. In P.T. Costa, Jr., & T.A. Widiger

(Eds.), Personality disorders and the five-factor model of personality (2nd ed., pp. 89–99). Washington, DC: American Psychological

Association. A thorough yet accessible treatment of how the DSM- IV personality disorders can be understood from the perspective of the five-factor model of personality.

Zanarini. M.C., Frankenburg, F.R., Hennen, J., Reich, B., & Silk, K.R.

(2005). The McLean Study of Adult Development (MSAD):

Overview and implications of the first six years of prospective

follow-up. Journal of Personality Disorders, 19, 505–523. An overview of the first 6 years of an intensive, longitudinal study,

begun in the early to mid-1990s, focused on examining the course

and outcome of serious personality pathology, specifically bor-

derline PD.

REFERENCES

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC: Author.

Bohus, M., Haaf, B., Simms, T., Limberger, M.F., Schmahl, C., Unckel,

C., et al. (2004). Effectiveness of inpatient dialectical behavioral

therapy for borderline personality disorder: A controlled trial.

Behaviour Research and Therapy, 42, 487–499.

Caspi, A., Roberts, B.W., & Shiner, R.L. (2005). Personality develop-

ment: Stability and change. Annual Review of Psychology, 56, 453–484.

Clark, L.A. (2005). Stability and change in personality pathology:

Revelations of three longitudinal studies. Journal of Personality Disorders, 19, 524–532.

Clark, L.A. (2007). Assessment and diagnosis of personality disorder:

Perennial issues and emerging conceptualization. Annual Review of Psychology, 58, 227–257.

Critchfield, K.L., & Benjamin, L.S. (2006). Principles for psychosocial

treatment of personality disorder: Summary of the APA Division 12

Task Force/NASPR Review. Journal of Clinical Psychology, 62, 661–674.

Fraley, R.C., & Roberts, B.W. (2005). Patterns of continuity: A dynamic

model for conceptualizing the stability of individual differences in

psychological constructs across the life course. Psychological Review, 112, 60–74.

Gabbard, G.O. (2000). Psychotherapy of personality disorders. Journal of Psychotherapy Practice & Research, 9, 1–6.

Grilo, C.M., Sanislow, C.A., Gunderson, J.G., Pagano, M.E., Yen, S.,

Zanarini, M., et al. (2004). Two-year stability and change of

schizotypal, borderline, avoidant, and obsessive-compulsive per-

sonality disorders. Journal of Consulting & Clinical Psychology, 72, 767–775.

Harpur, T.J., & Hare, R.D. (1994). Assessment of psychopathy as a

function of age. Journal of Abnormal Psychology, 103, 604–609.

James, W. (1950). The principles of psychology. New York: Dover. (Work originally published 1890)

Johnson, J.G., Cohen, P., Kasen, S., Skodol, A.E., Hamagami, F., &

Brook, J.S. (2000). Age-related change in personality disorder trait

levels between early adolescence and adulthood: A community-

based longitudinal investigation. Acta Psychiatrica Scandinavica, 102, 265–275.

McFarland, B.R., & Klein, D.N. (2005). Mental health service use by

patients with dysthymic disorder: Treatment use and dropout in a 7

1/2-year naturalistic follow-up study. Comprehensive Psychiatry, 46, 246–253.

Morana, H.C.P., & Camara, F.P. (2006). International guidelines for the

management of personality disorders. Current Opinion in Psychi- atry, 19, 539–543.

Morey, L.C., Hopwood, C.J., Gunderson, J.G., Zanarini, M., Skodol,

A.E., Shea, M.T., et al. (2007). Comparison of diagnostic models

for personality disorders. Psychological Medicine, 37, 983– 994.

Roberts, B.W., & DelVecchio, W.F. (2000). The rank-order consistency

of personality traits from childhood to old age: A quantitative re-

view of longitudinal studies. Psychological Bulletin, 126, 3–25.

Roberts, B.W., & Mroczek, D. (2008). Personality trait change in

adulthood. Current Directions in Psychological Science, 17, 31–35.

Roberts, B.W., Walton, K.E., & Viechtbauer, W. (2006). Patterns of

mean-level change in personality traits across the life course: A

meta-analysis of longitudinal studies. Psychological Bulletin, 132, 1–25.

Roberts, B.W., Wood, D., & Caspi, A. (2008). The development of

personality traits in adulthood. In O.P. John, R.W. Robins, & L.A.

Pervin (Eds.), Handbook of personality: Theory and research (3rd ed., pp. 375–398). New York: Guilford.

Skodol, A.E., Pagano, M.A., Bender, D.S., Shea, M.T., Gunderson, J.G.,

Yen, S., et al. (2005). Stability of functional impairment in patients

with schizotypal, borderline, avoidant, or obsessive-compulsive

personality disorder over two years. Psychological Medicine, 35, 443–451.

Warner, M.B., Morey, L.C., Finch, J.F., Gunderson, J.G., Skodol, A.E.,

Sanislow, C.A., et al. (2004). The longitudinal relationship of

personality traits and disorders. Journal of Abnormal Personality, 113, 217–227.

Widiger, T.A., & Frances, A.J. (2002). Personality disorders and the

five-factor model of personality. In P.T. Costa, Jr., & T.A. Widiger

(Eds.), Personality disorders and the five-factor model of personality (2nd ed., pp. 23–44). Washington, DC: American Psychological

Association.

Volume 18—Number 1 31

Lee Anna Clark