Shorter Oxford Textbook of Psychiatry, 6th Ed.

CHAPTER 7. Personality and personality disorder


Personality types

The origins of personality

The assessment of personality

Personality disorder

The historical development of ideas about abnormal personality

The classification of abnormal personalities

Descriptions and diagnostic criteria

Rates of personality disorder in the clinic and the general population


The course of personality disorder


The management of personality disorders

The management of specific personality disorders

Ethical problems


The term personality refers to those enduring qualities of an individual that are shown in their ways of behaving in a wide variety of circumstances, and which we use to distinguish between people. Personality therefore differs from mental disorder in that the behaviours which define it have been present throughout adult life, whereas the behaviours that define mental disorder differ from the person’s previous behaviour. When we say that a mentally ill person is ‘not their normal self’, we are drawing on our understanding of their personality and usual behaviour. The distinction is easy to make when behaviour changes markedly over a short period of time (as in a manic disorder), but can be difficult when the changes occur very gradually (as in some cases of schizophrenia).

The importance of personality

Gaining an understanding of and familiarity with a patient’s personality is important in psychiatry. Different personalities predispose to some psychiatric disorders, and they may account for unusual features or colour the presentation of a psychiatric disorder (‘pathoplastic’ factors). They may also explain how a patient approaches treatment, and dictate different strategies for establishing and maintaining a successful therapeutic relationship.

Personality as predisposition

Personality can predispose to psychiatric disorder by modifying an individual’s response to stressful events. For example, adverse circumstances are more likely to induce an anxiety disorder in a person who has always worried about minor problems.

Personality as a pathoplastic factor

Personality can contribute to unusual features of a disorder, particularly when personality traits become exaggerated with illness. For example, rumination and inhibition may be the presentation of depression in an individual with an obsessional personality. The underlying diagnosis can be obscured if the psychiatrist has not made an accurate assessment of personality.

Personality in relation to treatment

Personality is an important determinant of a person’s approach to treatment. For example, people with obsessional traits may become frustrated and resistant if treatment does not follow their expectations, and anxiety-prone people may discontinue medication prematurely because of concerns about minimal side-effects. Some people with a severe disorder of personality, particularly so-called Cluster B disorders (antisocial, borderline, impulsive, histrionic, and narcissistic personality disorders; see below) have often been effectively excluded from services because of the difficult relationships they form with clinicians (see Lewis and Appleby, 1988). This is a significant problem, as there is strong evidence that personality disorder is common in clinical populations, and that people with personality disorders have an increased risk for a range of mental illnesses (Moran, 2002). There is now a concerted move to remedy this (National Institute for Mental Health in England, 2003), which will be addressed later in this chapter and in Chapter 20.

Patients are often aware of their personalities—they know if they are ‘emotional’ or ‘conscientious’ or ‘anxious.’ Acknowledging this diplomatically (making sure to note the positive aspects as well as the negative ones) and discussing how it may interact with their treatment (and with their life generally) can be an effective tool in maintaining treatment.

Personality types

A first step in understanding personality is to identify some basic types. Clinicians have generally derived these from a mixture of clinical and common-sense collective experience with several generally recognizable categories, such as a sociable and outgoing type, a solitary and self-conscious type, and an anxious and timid type. Psychologists have adopted a more rigorous scientific approach using personality tests to measure aspects of personality (‘traits’) such as anxiety, energy, flexibility, hostility, impulsiveness, moodiness, orderliness, and self-reliance. These are then subject to statistical methods to discover which of them cluster into ‘factors.’

Although these statistical procedures appear very scientific, it is important to remember that their results are determined by the original hypotheses of the investigators (which traits they considered to be important and included in their analyses, and the order in which they included them, etc.). Like diagnoses, these are essentially working hypotheses, which are continuously evolving and should not be treated with too much reverence—their value lies in their utility.

Different investigators have derived different personality factors from such traits. Cattell (1963) identified five factors, whereas Eysenck (1970a) originally proposed only two ‘dimensions’ (high-level factors), which he labelled extraversion–introversion and neuroticism. Subsequently he added a third dimension, ‘psychoticism ‘(Eysenck and Eysenck, 1976), but his use of the term is rather misleading, denoting coldness, aggressivity, cruelty, and antisocial behaviour.

A five-factor formulation for personality has persisted, although the terms used for the factors have varied. They can be identified as openness to experience (or novelty seeking), conscientiousness, extraversion–introversion, agreeableness (or affiliation), and neuroticism. Widiger and Costa (1994) proposed that each of these five factors is composed of six traits, and developed assessment inventories based on self-report and report from informants (Costa and McCrae, 1992), as well as a semi-structured interview (Trull and Widiger, 1997).

Cloninger (1986) and Cloninger et al. (1993) developed an alternative scheme with three ‘basic behavioural dispositions’ expressed as four basic temperaments (a seven-factor model). The behavioural dispositions are behavioural activation, behavioural inhibition, and behavioural maintenance. Cloninger considered behavioural activation to be associated with the basic temperament of novelty seeking, behavioural inhibition with harm avoidance, and behavioural maintenance with reward dependence. The fourth basic temperament is persistence. Cloninger’s scheme also includes three character traits, namely self-directedness, cooperativeness, and transcendence. The various types of personality and personality disorder are constructed from these four basic temperaments and three character traits. Cloninger’s scheme is noteworthy for its inclusion of both inherited differences in brain function and the effects of experience.

These factor schemes of personality have survived despite sustained criticism that they are largely dependent on what was included in the original factor analyses. Zuckerman (2005) has also criticized these theories because they assume an alignment between personality traits and brain systems (Zuckerman, 2005). In recent years there has been some refinement, with greater emphasis on cognitive factors. Matthews has proposed a cognitive–adaptive theory (Matthews and Dorn, 1995; Penke and Denissen, 2007), and Mischel has emphasized the dynamic nature of personality in his cognitive–affective system (Mischel et al., 2004). Current texts on personality theory (for example, Matthews et al., 2003) are weighty tomes, and those interested in deepening their understanding might be best advised to start by consulting a clinical psychology colleague.

Despite these scientific approaches, clinicians continue to use everyday words to describe the positive and negative features of normal personality. Positive attributes include ‘outgoing’, ‘self-confident’, ‘stable’, and ‘adaptable.’ Negative attributes include ‘sensitive’, ‘jealous’, ‘irritable’, ‘impulsive’, ‘self-centred’, ‘rigid’, and ‘aggressive.’

The origins of personality

The biological basis of personality types

Genetic influences

Everyday observation suggests that children often resemble their parents in personality. Such similarities could be either inherited or acquired through social learning. Three kinds of scientific study have been used to study the inheritance of personality.

Studies of body shape and personality. Different personalities have been linked to body shape (‘beware of Brutus, he has a lean and hungry look’). If this were true, the link could be genetic. Kretschmer (1936) described three types of body build—pyknic (stocky and rounded), athletic (muscular), and asthenic (lean and narrow). He suggested that the pyknic body build was linked to the cyclothymic personality type (sociable with variable moods), whereas the asthenic build was related to the ‘schizotypal’ personality type (cold, aloof, and self-sufficient). Kretschmer’s ideas were based on subjective judgements, and were influenced by experience of the ‘associated’ psychotic disorders (manic-depressive disorder and schizophrenia). Sheldon used quantitative methods to assess physique and more objective ratings of personality, and failed to support the link (Sheldon et al., 1940).

Studies of twins. More direct evidence has been obtained from personality tests of identical twins reared together or reared apart. These suggest that the heritability for traits of extraversion and neuroticism is 35–50% (McGuffin and Thapar, 1992). The heritability of other traits is broadly similar.

Linkage studies. Molecular genetic studies have sought linkages for measures of novelty seeking and of neuroticism. Several quantitative trait loci have been identified that influence variations in neuroticism (Fullerton et al., 2003). Linkage between harm avoidance and a region on 8p21 has been reported in one study (Zohar et al., 2003). Such studies are difficult and require confirmation, but they may clarify the biological basis of some aspects of personality.

Childhood temperament and adult personality

Young infants differ in patterns of sleeping and waking, approach or withdrawal from new situations, intensity of emotional responses, and span of attention. These differences, which are described in more detail in Chapter 22, could form the basis for personality development. Although they do persist into later childhood, they have not been shown to be related to adult personality (Berger, 1985).

Childhood experience and personality development

Everyday experience suggests that childhood experience shapes personality (society is built on this premise), but it is not easy to demonstrate it. Experiences that seem relevant are difficult to quantify or even to record reliably, and it is extremely difficult and expensive to conduct prospective studies that span the time period from childhood events to adult personality. Retrospective studies are easier to arrange, but recall of childhood experiences is unreliable. Although scientifically collected information is sparse (see the section on aetiology of personality disorder below), the psychodynamic theories of Freud retain considerable influence.

The Freudian theory of personality development emphasizes events during the first five years of life. Freud proposed that crucial stages of libido development (which he rather unhelpfully referred to as ‘oral’, ‘anal’, and ‘genital’) must be accomplished successfully for healthy personality development. Failure or fixation at particular stages explained certain features of adult personality—for example, difficulties at the anal stage would lead to obsessional personality traits. Some personality growth was possible later, through identification with people other than the parents, but this influence was less strong. Freud’s explanation is excessively comprehensive and flexible. It can be made to explain almost all personality variation in terms of infantile experience, but its very flexibility makes it impossible to test scientifically. However, it retains enormous intuitive appeal and is widely used both within and outside of psychiatry.

Jung’s theory of personality development also recognizes the importance of psychic events in early life, but unlike Freud, Jung considered personality development to be a lifelong process. He considered individuation to be the aim of personal development, and his theories are particularly valued in relation to the disorders of later life. He proposed a structured theory of personality and introduced the terms ‘introvert’ and ‘extrovert.’

Adler and the neo-Freudians rejected Freud’s exclusive focus on libido development. Adler emphasized a struggle for mastery (overcoming the ‘inferiority complex’) as the driving force in personality development. The ‘neo-Freudians’ (Fromm, Horney, and Sullivan) increasingly emphasized social and peer-group factors in personality development.

Erik Erikson proposed an essentially similar process to that of Freud, although his nomenclature is less offputting, and is framed in terms of the individual developmental challenges. The oral stage is trust versus mistrust when feelings of security develop. The anal stage is autonomy versus doubt, when the child learns self-control, social rules, and self-confidence. The genital stage is initiative versus guilt, when children develop their image of themselves as people, leading to confidence and initiative. Erikson extended Freud’s idea of a latency period into adolescence, and called this last stage the period of industry versus inferiority. Erikson’s scheme recognizes the importance of adolescence for personality development, whereas Freud largely ignored it.

The assessment of personality

The assessment of personality is discussed in Chapter 3, but two points need to be emphasized. The first is that the assessment of personality used in everyday life cannot be applied reliably in clinical practice. Normally we assume that current behaviour reflects the person’s habitual ways of behaving (their personality), and in general this assumption is correct. This is not the case when we assess patients, because their current behaviour reflects the effects of their illness as well as their personality. A patient’s personality can only be judged confidently from reliable accounts of past behaviour, which have been obtained wherever possible from informants as well as from the patient.

Secondly, the assessment instruments developed for personality mentioned earlier (see p. 130), although more reliable in healthy individuals, can be misleading in the presence of mental disorder. In addition, they rarely measure the traits that are most relevant to clinical practice. Personality tests, although useful in research, are seldom used in clinical practice. For a review of personality assessment methods, see Westen (1997) and Clark and Harrison (2001).

The importance of personality assessment

The assessment of personality is important when making decisions about aetiology, diagnosis, and treatment. In aetiology, knowledge of personality helps to explain why certain events are stressful to that patient. In diagnosis, an understanding of personality may explain the presence of unusual features in a disorder which might otherwise cause uncertainty. In treatment, an assessment of personality helps to explain the patient’s reaction to their illness and its treatment, and aids the establishment of an effective therapeutic relationship. Personality assessment should be an integral part of every formulation, and not just reserved for those where a personality disorder is suspected.

It is best to record a series of descriptive terms chosen from the features of accepted personality disorders, because the more theoretical personality factors are too general to help the clinician. Examples would be ‘sensitive’, ‘lacking in self-confidence’, and ‘prone to worry.’ Such descriptions help to construct a picture of the unique features of each patient, which is a fundamental element of good clinical practice.

Personality disorder

The concept of abnormal personality

Some personalities are obviously abnormal—for example, paranoid personalities characterized by extreme suspiciousness, sensitivity, and mistrust. However, it is impossible to draw a sharp dividing line between normal and abnormal personalities. Abnormal personalities are in practice recognized because of the pattern of their characteristics, but our current classificatory processes in psychiatry demand that we identify criteria for inclusion. However, precisely which criteria should be used to make this distinction remains controversial. Two types of criteria have been suggested, namely statistical and social.

According to the statistical criterion, abnormal personalities are quantitative variations from the normal, and the dividing line is decided by a cut-off score on an appropriate measure. This approach is attractive, as it parallels that used successfully when defining abnormalities of intelligence, it appears non-judgemental, and it has obvious value in research. However, its usefulness in clinical practice is uncertain.

According to the social criterion, abnormal personalities are those that cause the individual to suffer, or to cause suffering to other people. For example, an abnormally sensitive and gloomy personality causes suffering for the individual who has it, and an emotionally cold and aggres sive personality causes suffering for others. These criteria are subjective and lack the precision of the first approach, but they serve the needs of clinical practice better and they have been widely adopted.

It is not surprising that it is difficult to frame a satisfactory definition of abnormal personality. ICD-9 describes personality disorders as follows:

Severe disturbances in the personality and behavioural tendencies of the individual, not directly resulting from disease, damage or other insult to the brain, or from another psychiatric disorder. They usually involve several areas of the personality and are nearly always associated with considerable personal distress and social disruption. They are usually manifest since childhood or adolescence and continue throughout adulthood.

ICD-10 emphasizes enduring patterns of behaviour, but the ICD-9 definition is more concise, and still valuable.

The ‘personal distress’ referred to in ICD-9 may sometimes only become apparent late in life (e.g. when a longstanding supportive relationship is lost). There are usually, although not always, significant problems in occupational and social performance.

It is important to recognize that people with abnormal personalities generally also have favourable traits, which the clinician should always assess. For example, those with obsessional traits are often dependable and trustworthy. Management plans that play to an individual’s strengths are more likely to be helpful for them.

Personality change

In some circumstances during adult life there may be a profound and enduring change in personality that is distinct from the temporary changes that may accompany stressful events or illness. This lasting change may result from:

• injury to or organic disease of the brain

• severe mental disorder, especially schizophrenia

• exceptionally severe stressful experiences (e.g. those experienced by hostages or by prisoners undergoing torture).

ICD-10 contains categories for each type of change. Change in personality due to organic disease of the brain is classified with the organic mental disorders in section F00, and includes the changes that occur following encephalitis and head injury. In DSM-IV this condition is diagnosed as personality change due to a general medical condition.

In ICD-10 the other two forms of personality change listed above are classified in section F60, disorders of adult personality and behaviour. To diagnose enduring personality change after psychiatric illness, the change of personality must have lasted for at least 2 years, be clearly related to the experience of the illness, and not have been present before it. Enduring personality change after a catastrophic experience must also have lasted for at least 2 years and have followed a stressful experience that was extreme (e.g. prolonged kidnapping, a terrorist attack, or torture). Victims are commonly hostile, irritable, distrustful, socially withdrawn, and on edge. The condition may follow post-traumatic stress disorder, but is distinct from it.

The historical development of ideas about abnormal personality

The concept of abnormal personality can be traced back to the beginning of psychiatry as a discipline at the turn of the nineteenth century, when the French psychiatrist Philippe Pinel described manie sans délire. By this he meant patients who were prone to outbursts of rage and violence, but who were not deluded. Delusions were then regarded as the hallmark of mental illness, and délire is the French term for delusion. J. C. Prichard, senior physician to the Bristol Infirmary, wrote about abnormal personality in 1835 in A Treatise on Insanity and Other Disorders of the Mind. After referring to Pinel’s manie sans délire, he suggested a new term, moral insanity, which he defined as a:

morbid perversion of the natural feelings, affections, inclinations, temper, habits, moral dispositions and natural impulses, without any remarkable disorder or defect of the intellect or knowing or reasoning faculties, and in particular without any insane delusion or hallucination.

(Prichard, 1835, p. 6)

Prichard’s moral insanity included conditions which we would now diagnose as personality disorder. It is important to remember that the term ‘moral’ had a wider meaning at this time, much akin to our term ‘social.’ The ‘moral treatment’ that was introduced by Pinel and developed by the Tuke family in the York Retreat in the 1790s was focused not on ethical judgements but on socialization. Prichard did not confine the term to people who had always behaved in these ways:

When, however, such phenomena are observed in connection with a wayward and intractable temper, with a decay of social affections, an aversion to the nearest relatives and friends formerly beloved—in short, with a change in the moral character of the individual—the case becomes tolerably well marked.

Later, the term ‘moral insanity’ was used by Henry Maudsley more as we would understand an antisocial personality disorder, to describe someone as having:

no capacity for true moral feeling—all his impulses and desires, to which he yields without check, are egoistic, his conduct appears to be governed by immoral motives, which are cherished and obeyed without any evident desire to resist them.

(Maudsley, 1885, p. 171)

Maudsley expressed a growing dissatisfaction with the term, which he referred to as ‘a form of mental alienation which has so much the look of vice or crime that many people regard it is an unfounded medical invention’ (Maudsley, 1885, p. 170).

Julius Koch introduced the term ‘psychopathic inferiority’ (Koch, 1891) for this group with marked abnormalities of behaviour in the absence of mental illness or intellectual impairment. Inferiority was subsequently replaced by personality. Emil Kraepelin was at first uncertain how to classify these people, and it was not until the eighth edition of his textbook was published that he finally adopted the term ‘psychopathic personality.’ He devoted a whole chapter to it, including not only the antisocial type but also six others (excitable, unstable, quarrelsome, and eccentric, together with liars and swindlers).

Kurt Schneider broadened the concept of psychopathic personality. Whereas Kraepelin’s seven types of psychopathic personality consisted of people who caused inconvenience, annoyance, or suffering to others, Schneider included people with markedly depressive or insecure characters, using the term ‘psychopathic’ to cover the whole range of abnormal personality, not just antisocial personality. The term came to have two meanings—a wider meaning of abnormal personality of all kinds, and a narrower meaning of antisocial personality—which has led to subsequent confusion.

In the 1959 Mental Health Act for England and Wales, psychopathic disorder was interpreted narrowly, and in the 1983 Act it was defined as:

a persistent disorder or disability of mind (whether or not including significant impairment of intelligence) which results in abnormally aggressive or seriously irresponsible conduct on the part of the person concerned.

This narrow concept, which emphasized aggressive or irresponsible behaviour, made its inclusion in the Act controversial. Consequently, the proviso of ‘treatability’ as a condition of detention was added exclusively to it, out of all the mental disorders. This distinction was abandoned in the 2007 amendment to the Mental Health Act (see Chapter 4) because it was perceived to present a barrier to effective treatment.

A recent development is the administrative category of dangerous and severe personality disorder (DSPD). This refers mainly to men with antisocial personality disorder and a history of serious violent or sexual offending. Its definition is based on three criteria:

1. a history of serious violent offending and the risk of at least 50% that a similar offence is likely to occur again

2. a severe personality disorder defined either clinically or using the Hare Psychopathy Checklist (Hare, 1991)

3. a demonstrable link between the offending and the personality disorder (Home Office, 1999).

This category (and four pilot units to treat 300 such men) was introduced in order to reduce risk. However, the recent change in the law to allow indeterminate detention of prisoners who still pose a serious risk may render the initiative redundant.

Because the term ‘psychopathic personality’ is ambiguous, the preferred terms are personality disorder and antisocial personality disorder to denote the wide and narrow senses, respectively.

The classification of abnormal personalities

General issues

The use of categories

Personality traits are continuously distributed, but the classification of psychiatric disorders uses categories and require definitions and boundaries. However, the criterion for inclusion (distress to the person or to others) is very imprecise, so cases that fall just short of it (subthreshold cases) are frequent. These sub-threshold cases often present clinical problems that are similar to those of definite cases.


It is not only the boundary between normal and abnormal personality that is imprecise and arbitrary; the boundaries between the different types of personality are also ill defined. This leads to an important divergence of practice between clinical psychiatrists and researchers and some psychologists. Many patients have features contained in the criteria for more than one personality disorder (Fyer et al., 1988). Structured instruments for measuring personality disorder, such as the International Personality Disorder Examination (Loranger et al., 1997), are often used by researchers and psychologists. They permit more than one personality diagnosis to be made in a single patient, and the term ‘comorbidity’ is used. From a clinical perspective it seems more likely that the patient has a single personality disorder that has features which overlap two of the arbitrary sets of criteria used in the current systems of diagnosis. We would usually diagnose the single personality disorder that best fits the mixed picture and is most useful in understanding and helping the patient in front of us. The concept of comorbidity is used when a patient has both a mental illness and a personality disorder (or more often a mental illness and substance abuse), but makes little clinical sense for personality disorder alone.

Conditions related to personality disorder and classified elsewhere

Cyclothymia and schizotypal disorder were previously classified as personality disorders. In ICD-10 both of these conditions have been removed from the personality disorders and classified instead with the mental disorders (cyclothymia with affective disorders, and schizotypal disorder with schizophrenia). This reflects the fact that these two conditions may begin in adult life, and in the case of schizotypal disorder, evidence from family studies that links it genetically to schizophrenia (see p. 145). In DSM-IV, cyclothymic disorder is classified with mood disorders but schizotypal disorder is retained as a personality disorder. In both classifications, multiple personality disorder is classified with dissociative disorders.

Classification of personality disorders in ICD-10 and DSM-IV

In Table 7.1 the classification of personality disorders in ICD-10 is compared with that in DSM-IV. The two schemes are broadly similar, but there are some differences.

Table 7.1 Classification of personality disorders


The use of Axis II

In DSM-IV, personality disorders are classified on a different ‘axis’ (Axis II) from mental disorders (which are classified on Axis I). This emphasizes the different nature of the two types of diagnosis, and encourages an assessment for personality disorder in every case. This convention is not adopted in ICD-10. The personality should still be assessed, and if no personality disorder is present this should be recorded in the formulation.

Grouping into clusters

In DSM-IV, but not in ICD-10, personality disorders are grouped into three ‘clusters’:

1Cluster A: paranoid, schizoid, and schizotypal

2Cluster B: antisocial, borderline, histrionic, and narcissistic

3Cluster C: avoidant, dependent, and obsessive–compulsive.

This convention is adopted later in this chapter.

Different names for the same personality disorder

The individual names for personality disorders differ some what between the two classifications. This reflects the real uncertainty about their boundaries and also an attempt through renaming to de-stigmatize some disorders. In ICD-10 the term ‘dissocial’ is used, whereas in DSM-IV the term ‘antisocial’ (the term used in this book) is used. In ICD-10, the term ‘anankastic’ is used, whereas ‘obsessive–compulsive’ is used in DSM-IV. In ICD-10, the term ‘anxious’ is used, whereas ‘avoidant’ is used in DSM-IV.

Conditions that are present in one classification but not the other

Emotionally unstable impulsive personality disorder and enduring personality change not attributable to brain damage or disease are found in ICD-10 but not in DSM-IV. Narcissistic personality disorder and passive–aggressive personality disorder are included in DSM-IV but not in ICD-10 (passive–aggressive personality disorder is listed as a category for further study). As noted above, schizotypal personality disorder is classified with schizophrenia in ICD-10 (and named schizotypal disorder), whereas in DSM-IV it is classified as a personality disorder.

Descriptions and diagnostic criteria

This section contains an account of the abnormal personalities listed in ICD-10 and DSM-IV (see Table 7.1). The criteria for diagnosis are lengthy, and differ somewhat in wording and emphasis in the two systems. The main common features from the two sets of definitions are given, simplified or paraphrased where appropriate to give a more general description.

All personality disorder diagnoses must meet the basic criteria for personality disorder summarized from the ICD-10 criteria for research as follows:

• The characteristic and enduring patterns of behaviour differ markedly from the cultural norm and in more than one of the following areas: cognition, affectivity, control of impulses and gratification, and ways of relating to others.

• The behaviour is inflexible, and maladaptive or dysfunctional in a broad range of situations.

• Personal distress is caused to others and/or to self.

• The presentation is stable and long-lasting, usually beginning by late childhood or adolescence.

• The behaviour is not caused by another mental disorder, or by brain injury, disease, or dysfunction.

There are several diagnostic instruments for personality disorder (see Box 7.1), but they are mainly of value in research.

Cluster A personality disorders

Paranoid personality disorder

Such people are suspicious and sensitive (see Table 7.2). They have a marked sense of self-importance, but easily feel shame and humiliation. They are suspicious and constantly on the lookout for attempts by others to deceive or exploit them, which makes them difficult for other people to get on with. They are usually mistrustful and often jealous. They have difficulty making friends and avoid involvement in groups. They may appear secretive, even devious, and self-sufficient to a fault. They take offence easily and see rebuffs where none was intended.

They are sensitive to rebuff, prickly, and can be argumentative, often reading demeaning or threatening meanings into innocent remarks. Ernst Kretschmer described how their suspicious ideas can be so intense that they are mistaken for persecutory delusions. These sensitive ideas of reference are considered further in Chapter 12. Paranoid individuals can be resentful and bear grudges, with a strong sense of their rights. They may engage in litigation, often persisting with this against all advice.

Paranoid personalities sometimes have a sense of self-importance, and may consider that others have prevented them from fulfilling their potential.

Box 7.1 Some instruments for diagnosing personality disorders

International Personality Disorders Examination (Loranger et al, 1994)

• Assesses DSM-IV and/or ICD-10 personality disorders categorically and/or dimensionally.

• Semi-structured clinical interview in two versions: a DSM-IV version with 99 sets of questions, and an ICD-10 version with 67 sets of questions. There is a screening questionnaire for each version which can be used to identify people who are unlikely to have a personality disorder.

Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II) (First et al, 1995)

• Assesses DSM-IV personality disorders either categorically or dimensionally.

• Semi-structured clinical interview with 119 sets of questions. There is a screening interview which can be used to identify people who are unlikely to have a personality disorder.

Structured Interview for DSM-IV Personality (SIDP-IV) (Pfohl et al, 1997)

• Assesses DSM-IV and ICD-10 personality disorders either categorically or dimensionally.

• Semi-structured clinical interview with 101 questions in a thematic version and 107 questions in a disorder-by-disorder version.

For further information on these and other instruments for personality assessment, see Rush (2000), chapter 32.

Table 7.2 Features of paranoid personality disorder





Resentful Bears grudges


Schizoid personality disorder

The term ‘schizoid’ was suggested by Kretschmer (1936), who believed that this type of personality is related to schizophrenia, although this idea has not been confirmed (see Chapter 11). People with this disorder are emotionally cold, detached and aloof, introspective, and prone to engage in excessive fantasy (see Table 7.3). They are unable to express either tender feelings or anger, and show little interest in sexual relationships. When the disorder is extreme they appear callous. They are generally ill at ease in company, do not form intimate relationships, and show little family feeling. They are seclusive, following a solitary course through life, and often remain unmarried. They show little sense of enjoyment or pleasure in the activities that most people enjoy, which contributes to their separation from others. Introspective and prone to fantasy, they are interested in intellectual matters, and have a complex inner world of fantasy, albeit without emotional content.

Schizotypal personality disorder

Schizotypal individuals are socially anxious and behave eccentrically. They experience cognitive and perceptual distortions, and have oddities of speech and inappropriate affective responses. This personality disorder appears to be related to schizophrenia, so it is not classified as a personality disorder in ICD-10 (as it is in DSM-IV), but placed with schizophrenia and called schizotypal disorder.

Table 7.3 Features of schizoid personality disorder

Emotionally cold



Lacking enjoyment


Schizotypal individuals experience social anxiety in company, so have difficulty in forming relationships and lack friends and confidants. They feel different from other people and do not fit in. (This is in contrast to schizoid individuals, who are quite content with this situation.) Their cognitive and perceptual distortions include ideas of reference (but not delusions), suspicious ideas, odd beliefs, and magical thinking (e.g. belief in clairvoyance, mind reading, and telepathy), and unusual perceptual experiences (e.g. awareness of a ‘presence’, or experiences bordering on hallucinations). They also show oddities of speech, such as unusual constructions, words, and phrasing, as well as vagueness, and a tendency to digression. Their affective responses are unusual, and they appear stiff, odd, and constricted in their emotions. Their behaviour is eccentric, with odd mannerisms, unusual choices of clothing, disregard of conventions, and awkward social behaviour.

Cluster B personality disorders

This is the group of personality disorders that looms largest in current psychiatric practice. Whereas most personality disorders ‘colour’ the presentation and management of patients, individuals with cluster B disorders are often referred specifically with a request for treatment of the personality disorder itself.

Antisocial (dissocial) personality disorder

The term ‘antisocial’ is used in DSM-IV and in this book, whereas ‘dissocial’ is the term used in ICD-10. Antisocial personality disorder is not simply another term for delinquency. People with this disorder show a callous lack of concern for the feelings of others. They disregard the rights of others, act impulsively, lack guilt, and fail to learn from adverse experiences (see Table 7.4). Often their abnormal behaviour is made worse by the abuse of alcohol or drugs. A lively and influential description of this type of personality disorder was written in 1941 by Hervey Cleckley in his book The Mask of Sanity(Cleckley, 1964).

The criteria for diagnosis differ slightly in the two classifications; the following criteria are taken from ICD-10. The DSM-IV criteria include the requirement of conduct disorder before the age of 15 years.

A central feature is a callous lack of concern for others. People with antisocial personality disorder may be exploitative and even violent, and their sexual activity lacks tenderness. They may inflict cruel or degrading acts on other people, and their partners may be physically or sexually abused. They may have a superficial charm, but their relationships are shallow and short-lived. They are irresponsible and depart from social norms. They do not obey rules and may repeatedly break the law, often committing violent offences. Their offending typically begins in adolescence.

Table 7.4 Features of antisocial personality disorder


Transient relationships


Impulsive and irritable

Lacking guilt and remorse

Failure to accept responsibility

Such individuals are impulsive, rarely plan ahead, and typically have an unstable work record. They take risks, disregarding their own safety and that of other people. They are irritable, and when angry sometimes assault others in a violent way. These features of personality are accompanied by a striking lack of guilt or remorse and a failure to change their behaviour in response to punishment or other adverse outcomes. They avoid responsibility, transferring blame on to other people and rationalizing their own failures. They are often deceitful and irresponsible about finances.

The concept of ‘psychopathy’ overlaps with antisocial personality disorder. Although it has been dropped from the diagnostic classification, it is still in use in psychology (particularly forensic psychology) and the prison service, and has a very specific meaning. It is generally considered to be a character trait, and is usually assessed using a structured instrument such as the Hare Psychopathy Checklist-Revised (PCL-R) (Hare, 1991). This distinguishes two factors—violent and impulsive behaviour, and charm and exploitation. The terms ‘psychopathy’ and ‘antisocial personality disorder’ are often unhelpfully interchanged, and psychopathy is probably best avoided by general psychiatrists.

Borderline personality disorder

The term ‘borderline personality disorder’ is a confusing and unsatisfactory one, which reflects its complex history. It was originally used in psychodynamic circles to describe people who showed marked ‘instability’ of presentation, as opposed to a steady neurotic state. The proposed ‘border’ at that time was with psychosis. Kernberg (1975) described it as involving:

1. ego weakness with poor control of impulses

2. ‘primary process’ (i.e. irrational) thinking despite intact reality testing

3. use of less ‘mature’ defence mechanisms, such as projection and denial

4. diffuse personal identity.

People with borderline personality disorder experience their lives as being dominated by strong and fluctuating emotions that often overwhelm them. They strive for affection and intimacy but are regularly disappointed, and may exhaust their partners with the intensity of their emotional demands. They are themselves confused by the strength and unpredictability of their moods as they plunge into anger or despair. They are often insecure in their personal identity and need reassurance and stability, which they may then find constricting and irritating. Self-harm is common (either suicidal attempts, or cutting to release tension). Such self-destructive behaviour can be extreme and may dominate the relationship with care services. Alcohol and drug abuse are also common, as the person attempts to self-medicate their distressing emotions. Clinical populations are dominated by relatively young women, although the epidemiological studies challenge this distribution.

When this type of personality disorder was introduced into the classification systems, more objective criteria were developed, but it remains unsatisfactory. The diagnosis was (and still is) almost exclusively applied to young women. This raised the question of whether it was a more culturally determined expression of mood disorder (i.e. depression) filtered through a histrionic (at that time referred to as ‘hysterical’) personality. The suggestion that it may respond (at least partially) to antidepressants has fuelled this controversy, as has the greater degree of variation over time than in other personality disorders, and the observation that in most patients it seems to resolve.

Not surprisingly, with such uncertainty different names have been adopted in the two classifications—borderline personality disorder in DSM-IV, and emotionally unstable personality disorder in ICD-10. The latter is divided into a ‘borderline type’ and an ‘impulsive type.’ In DSM-IV there are nine features, of which five are required for a diagnosis. In ICD-10 there are five criteria for each sub-type, and four are required for diagnosis. The criteria (slightly abbreviated) are shown in Table 7.5. Several features of the ICD-10 impulsive type are among the criteria for borderline personality disorder in DSM-IV.

Borderline personality disorder remains a controversial and contested diagnosis. It was originally entered into DSM-III only after a protracted debate and on the grounds that its utility would be researched. Many people who meet the criteria for borderline personality disorder in DSM-IV also meet the criteria for histrionic, narcissistic, and antisocial personality disorder (Skodol et al., 2002a).

Impulsive personality disorder

This disorder is recognized in ICD-10 as a subtype of emotionally labile personality disorder. It is not included separately in DSM-IV, although several of its features match the criteria for borderline personality disorder. The diagnostic criteria are listed in Table 7.5, and three criteria must be met for the diagnosis. People with impulsive personality disorder cannot control their emotions adequately, and are liable to outbursts of sudden unrestrained anger which they subsequently regret. They may use physical violence occasionally, causing serious harm. Unlike people with antisocial personality disorder, they do not generally have other difficulties in relationships.

Table 7.5 Abbreviated criteria for emotionally unstable and borderline personality disorders


Histrionic personality disorder

Histrionic personality disorder is included in both ICD-10 and DSM-IV, but the criteria are somewhat different. Table 7.6 lists the features that are diagnostic criteria in ICD-10, and also notes the criteria that differ between the two systems.

Self-dramatization is a striking feature, and may include ‘emotional blackmail’, angry scenes, and demonstrative suicide attempts. These individuals are suggestible and easily influenced by others, especially by figures of authority. They seek attention and excitement, are easily bored, and have short-lived enthusiasms. They have a shallow labile affect. Their emotional life is dramatic and may exhaust others. However, they recover quickly and often cannot understand why others are still upset. They seek intimacy and can be inappropriately seductive.

Table 7.6 Features of histrionic personality disorder



Shallow labile affect

Seeks attention and excitement

Inappropriately seductive

Over-concern with physical attractiveness

Note: DSM-IV has two additional criteria:

• Speech excessively impressionistic

• Considers relationships to be more intimate than they are

Histrionic individuals are often insecure about their value, and consequently may be overconcerned with physical attractiveness. They can appear self-centred and vain. They may also have a marked capacity for self-deception, being convinced by their own fabrications—however elaborate and improbable these may be—even when other people have seen through them. In its most extreme form, histrionic personality disorder is observed in ‘pathological liars’ and swindlers. Blanche DuBois in Tennessee Williams’ A Streetcar Named Desire poignantly portrays the disastrous consequences that such a personality can have both for the individual and for those around them.

Narcissistic personality disorder

This personality disorder is listed in DSM-IV but not in ICD-10, where it is one of the disorders coded in the residual category ‘Other specific personality disorder.’ The DSM-IV criteria are summarized and paraphrased in Table 7.7.

People with this disorder have a grandiose sense of self-importance and are boastful and pretentious. They are preoccupied with fantasies of unlimited success, power, beauty, or intellectual brilliance. They consider themselves special, and expect others to admire them and offer them special services and favours. They feel entitled to the best, and seek to associate with people of high status. They exploit others and do not empathize with or show concern for their feelings. They envy the possessions and achievements of others, and expect that those individuals will envy them in the same way. They appear arrogant, disdainful, and haughty, and behave in a patronizing or condescending way.

Table 7.7 Features of narcissistic personality disorder

Grandiose sense of self-importance

Fantasizes about unlimited success, power, etc.

Believes him- or herself to be special

Requires excessive admiration

Sense of entitlement to favours and compliance

Exploits others

Lacks empathy

Envious of others, and believes that others envy him or her

Arrogant and haughty

Narcissistic personality disorder carries probably the most pejorative overtones of any of the personality disorder diagnoses. It should be reserved for only the most obvious examples, and only after you have made sure that the diagnosis is not coloured by personal dislike. On balance it is probably a term to avoid if at all possible, like some older personality terms such as ‘immature personality’ or ‘inadequate personality’, which have no place in modern mental healthcare.

Cluster C personality disorders

Avoidant (anxious) personality disorder

DSM-IV uses the term ‘avoidant’ and in ICD-10 the term ‘anxious’ is preferred, with ‘avoidant’ as an accepted alternative. The diagnostic criteria in the two classifications are listed in Table 7.8. Liability to tension is a criterion only in ICD-10.

People with this disorder are persistently tense. They feel insecure and lack self-esteem. They feel socially inferior, unappealing, and socially inept. They are preoccupied with the possibility of rejection, disapproval, or criticism, and worry that they will be embarrassed or ridiculed. They are cautious about new experiences, avoid risk, and avoid social activity. They have few close friends, but they are not emotionally cold, and indeed crave the social relationships that they cannot manage to attain.

Table 7.8 Features of avoidant (anxious) personality disorder

Feelings of tension*

Feels socially inferior

Preoccupied with rejection

Avoids involvement

Avoids risk

Avoids social activity

Note: DSM-IV has two additional criteria:

• Restraint in intimate relationships, due to fear of being shamed or ridiculed

• Inhibited in new personal situations, due to feelings of inadequacy

* Not a DSM-IV criterion.

Dependent personality disorder

Table 7.9 lists the diagnostic criteria in ICD-10 and DSM-IV. People with this disorder allow others to take responsibility for important decisions in their lives. They are unduly compliant, but are unwilling to make direct demands on other people. They lack vigour and feel unable to care for themselves. Lacking self-reliance, they avoid responsibility, and may need excessive help to make decisions.

They are often protected by support from a more energetic and determined partner, and may only come to medical attention when the partner leaves or dies.

Table 7.9 Features of dependent personality disorder

Allows others to take responsibility

Unduly compliant

Unwilling to make reasonable demands*

Feels unable to care for him- or herself

Fear of being left to care for him- or herself

Needs excessive help to make decisions

Note: Three additional criteria are used in DSM-IV. They can be summarized as follows:

• Experiences difficulty in initiating projects

• Goes to excessive lengths to obtain support

• Urgently seeks a supportive relationship

* Not a criterion in DSM-IV.

Obsessive–compulsive personality disorder

DSM-IV uses the term ‘obsessive–compulsive personality disorder’, whereas ICD-10 uses the term ‘anankastic’ (originated by Kahn, 1928). Kahn used this term to avoid the implication that this personality disorder is linked to obsessive–compulsive disorder (see p. 199). Table 7.10 lists the diagnostic criteria in ICD-10 and DSM-IV.

These individuals are preoccupied with details and rules, order and schedules. They have an inhibiting perfectionism that makes ordinary work a burden as they are immersed in endless detail. They may lack imagination and fail to take advantage of opportunities. However, they usually have high moral standards, are often excessively conscientious, and may be rather judgemental. They can appear humourless and ill at ease when others are enjoying themselves.

These people lack adaptability to new situations. They are generally rigid and inflexible, avoiding change and preferring a familiar routine. They can be stubborn and controlling and generally thrifty, sometimes to the point of being miserly. They may hoard objects and money. They may be pedantic and unduly concerned with social conventions. They can also be particularly troubled by excessive doubt and caution, leading to indecision.

Table 7.10 Features of obsessive–compulsive (anankastic) personality disorder

Preoccupied with details, rules, etc.

Inhibited by perfectionism

Over-conscientious and scrupulous

Excessively concerned with productivity

Rigid and stubborn

Expects others to submit to their ways

Excessively pedantic and bound by convention*

Excessively doubting and cautious*

Note: DSM-IV has two additional criteria:

• Cannot discard worthless or worn out objects

• Miserly; hoards money

* Not a criterion in DSM-IV.

Passive–aggressive personality disorder

This term is applied to a person who is reluctant to express their hostility or even disagreement in situations, but instead relies on passive resistance. This can take the form of procrastination, dawdling, stubbornness, deliberate inefficiency, or feigned forgetfulness. This category does not appear among the personality disorders in ICD-10 or DSM-IV.

Affective personality disorders

Some people have lifelong disorders of mood regulation. They may be persistently gloomy (depressive personality disorder) or habitually in a state of inappropriate elation (hyperthymic personality disorder). A third group alternates between these two extremes (cycloid or cyclothymic personality disorder). These types of personality have been described for many years, and are readily recognized in clinical practice. In both ICD-10 and DSM-IV they are classified under disorders of mood and not under disorders of personality. In ICD-10 they are classified under ‘persistent mood (affective) states’ (cyclothymia or dysthymia), and in DSM-IV under cyclothymia or dysthymia. Nevertheless, it is useful to describe them briefly here.

People with depressive personality disorder always seem to be in low spirits. They have a persistently gloomy and pessimistic view of life, and may brood about their misfortunes and worry unduly. They have little capacity for enjoyment and are dissatisfied with their lives. Some are irritable and bad-tempered.

People with hyperthymic personality disorder are habitually cheerful and optimistic, and exhibit a striking zest for living. If they have these traits to a moderate degree, they are often highly effective and successful individuals whom we envy. However, if the traits are more pronounced they can show poor judgement and irritate those around them.

People with cyclothymic personality disorder alternate between the extremes of depressive and hyperthymic states described above. This instability of mood is much more disruptive than either of the persisting conditions. People with this disorder are periodically extremely cheerful, active, and productive, at which times they take on additional commitments in their work and social lives. When their mood changes they become gloomy and defeatist, and with reduced energy levels they now find their newly acquired obligations a burden.

Rates of personality disorder in the clinic and the general population

Clinical population rates

For a psychiatrist, the more important question is how frequently personality disorder will be encountered in their clinical practice. A series of studies have investigated this, and the 16 largest and most rigorous studies have been summarized by Zimmerman et al. (2008). All of these included over 500 subjects, and the largest consisted of 129 286 patients served by the New York State mental health services. The largest study using structured assessments is the MIDAS study, which included 859 consecutive outpatient attendees. Zimmerman’s review showed that much higher rates were found in structured assessments (generally around 40%) than in clinical assessments, which were much closer to the rates found in community samples, at 11–20%. In the UK, a recent study of 2528 community mental health team (CMHT) patients (Newton-Howes et al., 2010) using a simple structured assessment reported a rate of 40%. Among primary care consultations in the UK the rate is as high as 30% (Moran et al., 2000).

However, we know that rates in the clinic and rates in the general population are likely to be very different. Treatment seeking by individuals with personality disorder is powerfully affected by a number of factors (Narrow et al., 1993; Kessler et al., 1999), particularly the existence of an Axis I disorder.

General population rates

Epidemiological research into personality disorders in the general population began with the development of standardized instruments using DSM-III criteria. Studies require large samples because the prevalence rates of some personality disorders are low. It is also difficult to identify personality disorder reliably in community surveys without access to informants. Data come mainly from studies in the USA, the UK, and Germany, and from a recent World Health Organization (WHO) mental health survey review (Huang et al., 2009). The most extensive data come from 10 epidemiological studies in the USA and four in Europe cited by Huang et al. (2009). The prevalence rates for any personality disorder from these studies range from 3.9% to 15.7%. The WHO survey reported by Huang summarized 13 surveys in 13 countries involving 21 162 subjects and using a screening interview followed by the International Personality Disorder Examination (IPDE) (Loranger et al., 1997), administered by trained lay researchers. The total prevalence (i.e. the sum of all types) of personality disorder in this study was 6.1%, and rates ranged from 2.4% in Western Europe to 7.6% in the USA and 7.9% in Colombia. The latest and probably most reliable figure for the UK is 4.4% (Singleton et al., 2000). The WHO study found that personality disorder was more prevalent in males, younger people, and those with less education.

Estimates of the prevalence of the various types of personality disorder from the review by Guzzetta and de Girolamo (2009) are shown in Table 7.11. In these studies, sample sizes ranged from 200 to 1600. Some of these estimates have a substantial range, reflecting among other sources of variation the use of different assessment instruments. The rates of some disorders vary between men and women. Antisocial personality disorder is more common among men (estimated ratios range between 2.1 and 7.1). Although borderline and histrionic personality disorders are clinically reported more frequently in women, this was not consistently confirmed in epidemiological studies.

Structured assessment instruments have high reliability but may have more questionable validity. Certainly they seem to generate more positives than most clinicians would expect, and the absolute rates should be treated with some caution. However, the relatively primitive state of personality disorder theory and assessment may explain much of this variation. Similarly, the surprising findings of remarkably high rates of remission in personality disorder (see p. 147) when using these instruments indicate that they should be treated cautiously.

Table 7.11 Median prevalence rates of personality disorders in epidemiological surveys



General issues

The causes of personality disorder are uncertain, and this is an area in which hypotheses are many and research findings varied. The two major areas are genetic factors and various kinds of early life experience. Some personality disorders have been linked aetiologically with the psychiatric disorders that they resemble. The state of evidence for each of these three general issues will be considered with the individual personality disorders.

The study of the influence of early life experiences is made difficult by the long interval between these experiences and the diagnosis of personality disorder in adult life. Psychodynamic theories linking childhood experience and personality are hard to test, but will be mentioned where they have been influential.

Most clinicians accept that there are causal links between childhood experience, personality, and personality disorder. It is agreed good practice to assess childhood experiences and to use common-sense judgement to decide whether any of these may have influenced the personality. For example, extreme and repeated rejection by the parents may explain low self-esteem in adult life. Retrospective studies present obvious problems, but prospective studies are now becoming available. For example, independently documented gross physical neglect or abuse or sexual abuse in childhood has been shown to be associated with cluster B personality disorders (Johnson et al., 1999).

Since different causes have been suggested for the various personality disorders, they will be considered separately. Antisocial personality disorder is considered first, because more research findings have been reported for this than for the other disorders.

Antisocial personality disorder

Genetic causes

Twin studies. The much quoted early twin studies by Lange (1931) and Rosanoff et al. (1934) were concerned with probands with repeated conviction for criminal offences rather than antisocial personality disorder as such, so their relevance is uncertain. More recent twin studies have confirmed the heritability of antisocial behaviour in adults, and have shown that genetic factors are more important in adults than in antisocial children or adolescents where the shared environmental factors dominate (Lyons et al., 1995).

Adoption studies. Cadoret (1978) found that adoptees separated at birth from a parent who had displayed persistent antisocial behaviour had higher rates of antisocial personality disorder than did adoptees whose biological parents were not antisocial. This finding held whether the biological father or the biological mother had shown antisocial behaviour. However, among the offspring, antisocial personality disorder was diagnosed more often in men than in women, although the women had an increased rate of what was then diagnosed as hysteria (and is probably closer to the current diagnosis of borderline personality disorder). Cadoret suggested that hysteria (borderline personality disorder) is an expression in women of the same genetic endowment that causes antisocial personality disorder in men. Similar findings were reported in an earlier study by Crowe (1974). In addition, a small study of the biological parents of adoptees with antisocial behaviour found an excess of antisocial behaviour compared with the biological parents of children who were not antisocial (Schulsinger, 1982).

Cadoret et al. (1995) also studied the family environment of their adoptees and found that adverse factors in the adoptive environment (e.g. marital problems or substance abuse) independently predicted adult antisocial behaviours.

Linkage studies. These have proved difficult, and at the time of writing no confirmed linkage has been reported.

Cerebral pathology and cerebral maturation

The observation that some brain-injured patients show aggressive behaviour suggested that a minor degrees of brain injury might be a cause of antisocial personality. However, there is no convincing evidence to support this idea. Magnetic resonance imaging (MRI) studies have found reduced prefrontal grey matter in people with antisocial personality disorder (Raine et al., 2000). Deficits in the amygdala response to emotional stimuli and reduced amygdala volume in antisocial individuals may be linked to the characteristic lack of empathy (Blair, 2003).

5-Hydroxytryptamine and aggression

Abnormalities in brain 5-hydroxytryptamine (5-HT) neurotransmission have been reported in patients with impulsive and aggressive behaviour, although not specifically in relation to antisocial personality disorder. Low levels of the 5-HT metabolite 5-hydroxyindoleacetic acid (5-HIAA) have been found in the cerebrospinal fluid of subjects who have committed acts of unpremeditated violence (Linnoila and Virkkunen, 1992). In addition, more recent studies have found abnormalities in 5-HT-mediated neuroendocrine function and brain 5-HT2A-receptor availability in vivo in people with high trait aggressiveness (Coccaro et al., 2010; Rosell et al., 2010). It has been suggested that the same abnormalities may be relevant to personality disorders that are characterized by impulsive behaviour, particularly in the presence of prefrontal deficits (Dolan et al., 2002; New et al., 2004).

Developmental theories

Separation. In a much-quoted early study, Bowlby (1944, 1946) studied 44 ‘juvenile thieves’ and concluded that separation of a young child from their mother could lead to antisocial behaviour and failure to form close relationships. This work stimulated much research into the effects of separating children from their mothers (see p. 634), introduced radical changes into aspects of social and paediatric care, and has given rise to attachment theory. Not all separated children are affected adversely, and the effects of separation depend on many factors, including the child’s age, the previous relationship with the mother and father, and the reasons for separation. Rutter (1972) showed that parental marital disharmony partly accounts for the association between separation and antisocial disorder in sons.

Social learning in childhood. Scott (1960) proposed four ways in which antisocial behaviour could develop through social learning:

1. growing up in an antisocial family

2. lack of consistent rules in the family

3. overcoming other problems (e.g. aggressive behaviour to hide feelings of inferiority)

4. poor ability to sustain attention, and other impediments to learning.

There is evidence for some of these factors. Antisocial personality disorder is associated with physical abuse and with violent parenting without consistently applied rules (Pollock et al., 1990), and with low IQ and large family of origin (Farrington et al., 1988). The main value of Scott’s proposal is in providing a framework for the assessment.

Childhood behaviour problems and antisocial personality. A 30-year follow-up study of children attending a child guidance clinic found an association between behaviour problems in childhood and antisocial personality disorder in adult life (Robins, 1966). Only a minority of those even with serious antisocial behaviour in childhood became persistently antisocial in adult life, but most of the adults with antisocial personality disorder had behaviour problems in childhood. The association was stronger if in childhood there was more than one kind of antisocial behaviour, and if antisocial acts were repeated. Stealing among boys and sexual delinquency among girls were especially likely to be followed by antisocial behaviour in adult life. These early findings have been confirmed, and early conduct problems have been shown to predict antisocial personality independently of associated adverse family and social factors (Hill, 2003).

Paranoid personality disorder

Little is known about the causes of this disorder. Some investigators have reported that paranoid personality disorder is more frequent among first-degree relatives of probands with schizophrenia than among the general population (Kendler et al., 1985), but others have not confirmed this finding (Coryell and Zimmerman, 1989).

Schizoid personality disorder

The cause of this disorder is unknown. As noted above, schizoid personality does not appear to be closely related genetically to schizophrenia (Fulton and Winokur, 1993). The psychoanalyst Melanie Klein outlined what she called the ‘paranoid–schizoid position’ (Klein, 1952). This was an early developmental phase before infants could sustain a clear understanding of their mother as an individual who came and went (the ‘depressive position’, so called because the infant could then miss the mother). Klein’s views were not intended to explain either paranoid or schizoid disorders, although they are often quoted as such, but rather to emphasize the importance of the mother–child relationship more than she thought Freud had done.

Schizotypal personality disorder

This is more frequent among biological relatives of individuals with schizophrenia than among adopted relatives or controls (Kendler et al., 1981). A review of 17 structural imaging studies of people with this personality disorder found brain abnormalities that were similar in most ways to those in people with schizophrenia (Dickey et al., 2002). Similarly, there is strong evidence for a genetic contribution to this personality disorder, with heritability scores of 0.35–0.81 in twin studies (Ji et al., 2006; Kendler et al., 2006, 2007). These findings suggest that this personality disorder may be a milder form of schizophrenia, or that the two are related in some other way.

Borderline personality disorder

There is now an impressive volume of research into biological factors in borderline personality disorder (New et al., 2008), and some evidence for the importance of genetic factors. Twin studies show heritability scores of 0.65–0.76 (Torgersen et al., 2000; Ji et al., 2006). First-degree relatives of patients with borderline personality disorder are ten times more likely to be treated for borderline personality disorder or ‘borderline personality disorder-like personality disorder’ (Loranger et al., 1982). Genetic studies have focused on several candidate genes, particularly those associated with serotonin regulation. Significant associations have been found with the 5-HTT gene (Ni et al., 2006). Similarly, a wide range of imaging studies have implicated subtle grey matter changes in prefrontal and limbic regions, as well as functional dysregulation of the cortico-limbic circuitry involved in emotional regulation (Silbersweig et al., 2007). A recent ligand binding and PET imaging study has suggested that there are abnormalities in opioid regulation in the same brain regions (Prossin et al., 2010).

It is becoming accepted that it may be more fruitful to study the genetics of personality traits (e.g. dysregulation of affect and poor impulse control) than the personality disorders themselves (Siever et al., 2002). Psychodynamic theories propose a disturbed relationship with the mother at the stage of individuation of the child (Kern-berg, 1975), and people with a borderline personality are more likely to report that they experienced physical and sexual abuse in childhood (Berelowicz and Tarnopolsky, 1993).

Histrionic personality disorder

There have been few objective studies of the causes of this personality disorder. The genetics of histrionic personality disorder have not been studied with standardized methods of assessment, and the few reported investigations have yielded inconsistent findings (McGuffin and Thapar, 1992). Freud’s views of the role of the Oedipus complex and the failure of mature intimacy leading to over-sexualized relationships are impossible to test, but may help when dealing with histrionic individuals.

Obsessional personality disorder

Obsessional personality appears to have a substantial genetic aetiology (Murray and Reveley, 1981), although its nature is unknown. Psychoanalytic explanations were initially influential, but have now faded even within psychotherapy circles.

Anxious–avoidant, dependent, and passive–aggressive personality disorders

The genetics of these personality disorders have not been specifically studied, and their causes remain unknown at present.

The course of personality disorder

Personality disorders are defined as lifelong conditions, so little change would be expected with time. There is little reliable evidence about their outcome, other than for borderline personality disorder and to a lesser extent antisocial personality disorder. The results of those studies which have been conducted make surprising reading. Much of the influential work comes from the USA and uses repeated assessment according to DSM-IV criteria for personality disorder diagnosis or for personality disorder traits. Overall this work indicates that personality disorder is much less stable than had previously been believed. Skodol (2008) concludes that of studies conducted in the DSM-III era, over 50% of patients subsequently failed to meet personality disorder criteria over time. In this review of four large naturalistic studies, two reported changes in personality traits and two reported changes in diagnoses. All of them showed marked improvements. For example, a 48% reduction in personality disorder traits between 15 and 22 years of age (Cohen et al., 2005) and a reduction in the number of individuals continuing to meet the diagnostic criteria for borderline personality disorder are outlined below. In the face of such figures it is difficult not to question the validity of the measures. A judgement has to be made as to whether they are genuinely measuring what we understand by personality.

Borderline personality disorder. The outcome is very varied, suggesting that it may not be a single entity, or could in fact be a questionable one (Skodol et al., 2002b). In the most influential outcome study, the McLean Hospital follow-up (Zanarini et al., 2006), 290 borderline personality disorder patients were interviewed using the SCID II (First et al., 1997) every 2 years for 10 years. A total of 88% achieved ‘remission’ over this time period, 39% by 2 years, a further 22% by 4 years, a further 22% by 6 years, and so on. Those who still showed a borderline personality disorder more often had comorbid substance abuse or a history of childhood sexual abuse. A high rate of suicide has been found in some studies (8.5% in the study by Stone et al., 1987), raising the suspicion of an atypical affective disorder.

Aggressive and antisocial personality disorders. Clinical impressions indicate that minor improvement may take place slowly, especially in aggressive and antisocial behaviour. In the study by Robins and Regier (1991), about one-third of people with persistent antisocial behaviour in early adult life improved later as judged by the number of arrests and contacts with social agencies. However, they still had problems in relationships, as shown by hostility towards partners and neighbours, as well as an increased rate of suicide. There is a problem of circularity here in that criminality is one of the features of antisocial personality disorder in DSM-IV (although not in ICD-10), and not all antisocial personality disorder patients identified in community studies (Regier et al., 1990) have problems with the law, and not all criminals meet the criteria for antisocial personality disorder (Hare, 1983; Coid and Ullrich, 2010). Whether antisocial personality disorder improves significantly over time is the subject of conflicting evidence. Robins and Regier (1991) reported a high rate of remission in the third and fourth decades of life, whereas Rutter and Rutter (1993) emphasized its stability. The view that such patients ‘burn out’ may be as much a reflection of a reduction in help seeking as of late maturing.

Much less is known about the longer-term outcome in the other personality disorders. The overwhelming clinical impression is that borderline personality disorder is distinguished from the other disorders by its apparent variability. However, in the longitudinal studies using structured assessments there is also a high level of remission for other personality disorders. For instance, in a collaborative multi-site study of 668 DSM-IV-diagnosed personality disorder patients in the USA (Skodol et al., 2005), remission occurred in 38% of those with obsessive–compulsive personality and 23% of those with schizotypal personality disorder within 2 years.

For schizotypal personality disorder, Mehlum et al. (1991) reported a 3- to 5-year follow-up in which the outcome of this personality disorder was worse than that of borderline personality disorder. ‘Cluster C’ personality disorders (see p. 140) seem to have a better outcome than the other groups.


There is little evidence to guide the clinician in the choice of treatment for personality disorders. Few studies have met the basic requirements of randomization, blind assessment, and an appropriate control treatment. Tyrer and Davidson (2000) have identified several reasons for the lack of clinically relevant research.

Comorbidity is common between personality disorders and other disorders, especially mood disorders and substance abuse. Few trials have ruled out the possibility that changes following treatment are due to improvement in a comorbid condition rather than in the personality disorder. Virtually all studies have used outcome measures such as mood or self-harming rather than change in personality traits. Ideally, two kinds of trial are required—trials with pure personality disorder to show whether treatment has an effect independent of any comorbid condition, and pragmatic trials including comorbid disorders to show the treatment effect that can be expected in clinical practice. Duration of treatment and follow-up are too short when judged against the natural course of personality disorders. Tyrer and Davidson (2000) suggest that the minimum duration should be 2 years. Personality disorder is often associated with poor collaboration with treatment and high rates of drop-out from both treatments and studies.

Drug treatments

Apart from their effect on comorbid mood disorder, drug treatments could potentially affect the biological basis of a personality disorder, or might have a non-specific effect on anxiety, aggression, or other symptoms.

Antipsychotic drugs

Typical antipsychotics have been used to treat borderline and schizotypal personality disorders, but the clinical trials have had mixed results (see Tyrer and Bateman, 2004; Lieb et al., 2010). There is some evidence that antipsychotic drugs can ameliorate certain symptom clusters in borderline personality disorder patients (e.g. affective dysregulation, impulsive behaviour, and cognitive perceptual disturbances), but the current evidence is limited. In addition, there is some evidence that olanzapine may increase self-harming behaviour (Lieb et al., 2010). Some antipsychotic drugs are also reported to be of benefit in schizotypal personality disorder (Koenigsberg et al., 2003; Keshavan et al., 2004).

Antidepressant drugs

Amitriptyline has been tested against placebo as a treatment for borderline personality disorder (Soloff et al., 1986; Soloff, 1994). Some patients responded well, while others did not respond at all, possibly due to an associated depressive disorder. A recent meta-analysis suggested that SSRIs are not helpful in the depressive symptomatology that occurs as part of borderline personality disorder, but may be indicated for distinct episodes of comorbid depression (Lieb et al., 2010).

Mood stabilizers

Valproate has been reported to produce improvements in aggression, depression, and general symptomatology in borderline disorder in a small randomized trial (Hollander et al., 2003). Reports that lithium reduces anger and impulsiveness in antisocial personalities have not been convincingly replicated. There is an obvious practical difficulty in that individuals with impulsive behaviour may not comply with the safe use of lithium. A meta-analytical review concluded that anticonvulsant mood stabilizers such as topiramate, valproate, and lamotrigine may be beneficial in a number of aspects of borderline personality, including affective dysregulation and impulsive behaviour (Lieb et al., 2010).

Psychological treatments

One of the most important services that a psychiatrist can provide for a personality-disordered patient is an honest yet diplomatic sharing of their understanding of what is going on. Most patients have recognize that they have a problem of some kind and may be quite realistic about what psychiatrists have to offer. They may only have come to the psychiatric services because others have suggested it. An opportunity to talk through the strengths and weaknesses of their habitual responses can help an individual to put their characteristic behaviour patterns in perspective and perhaps find less damaging strategies for dealing with difficulties. It can also avoid the development of unrealistic expectations of what others can do, which could otherwise lead to unproductive dependency. Regular attendance at outpatient clinics when there is in reality nothing that can be done can seem benign and compassionate, but it carries a strong misleading message both for the patient and for others who relate to them.

There is growing confidence in the efficacy of psychotherapy for patients with personality disorders. Bateman and Tyrer (2004a) have concluded that ‘there is encouraging evidence that some patients with personality disorder are treatable … however, there is inadequate evidence to make specific recommendations for any particular therapy.’ A meta-analysis of 14 psychodynamic trials and 11 cognitive–behaviour therapy (CBT) trials found effect sizes of 1.46 for psychodynamic therapies and 1.00 for CBT (Leichsenring and Leibing, 2003). However, the effect sizes cannot be compared directly because of differences in patient populations and other variables. Similar encouraging but non-specific results have been reported for borderline personality disorder in two Cochrane Collaboration reviews for both psychological therapies and pharmacological treatments (Binks et al., 2009;Stoffers et al., 2010). For a review of psychological treatments for personality disorder, see Bateman and Tyrer (2004a).

Supportive therapy

Psychological support has been the mainstay of treatment of people with personality disorder. For some, modest but useful progress can be achieved by having a secure point of reference and a sense of being taken seriously over a period of months or sometimes years. Support may be provided by a member of the psychiatric team or, if the person has broken the law, by a probation officer. The dangers of this rather ‘paternalistic’ approach are increasingly being recognized, and supportive psychotherapy is probably best restricted to fairly senior staff, as it is difficult to do well and has a number of pitfalls.


Problem-solving counselling (see p. 576) can help patients to deal with stressful circumstances that provoke abnormal behaviour or painful feelings. The approach is less likely than other forms of counselling to lead to dependency and transference problems in patients with cluster C disorders. Its practical step by-step aspects are also appreciated by patients with cluster B disorders.

Psychodynamic counselling, which links past experiences to present difficulties, should be used selectively with personality disorders. It is likely to help young people who lack confidence, have difficulty in forming relationships, and are uncertain about the direction that their lives should take. However, for borderline and antisocial personalities, isolated psychodynamic counselling is problematic.

Dynamic psychotherapy

This is a much more ambitious treatment and requires that the person wishes to change and is able to collaborate. The technique often has to be modified, with less emphasis on the reconstruction of past events and more on ways of relating to others, coping with external difficulties, and dealing with feelings. The analysis and management of transference and counter-transference are considered crucial to its success. A more recent modification, known as ‘mentalization’, aims to help patients to increase their awareness of their own immediate feelings. Through this they can start to understand feelings, beliefs, and desires more generally, and eventually their reactions to others and others’ reactions to them (see Bateman and Fonagy, 2004). There have been too few controlled studies of this kind of treatment for personality disorder to allow conclusions to be drawn about its effectiveness (see Bateman and Tyrer, 2004a), but it has become enormously influential as a model for therapeutic day hospitals.

Cognitive therapy

Cognitive therapy was adapted for use with personality disorders by Beck and Freeman (1990). The usual cognitive therapy techniques are used (see p. 579), but the results are still preliminary. Manual-assisted cognitive therapy failed to reduce the repetition of self-harm in borderline personality disorder (Bateman and Tyrer, 2004a). However, a specialized form of very intensive cognitive therapy, known as dialetical behaviour therapy (Linehan et al., 2002), has been claimed to significantly reduce self-harm in patients with borderline personality disorder. It is extremely demanding of both the patient and an extensive therapy team, but is currently considered to be the treatment of choice.

Cognitive–analytic therapy

This approach (see p. 590) has been used with borderline personality disorder (Ryle, 1997), but its value has not been established in clinical trials. Young’s schema-focused approach is also used (Young, 1999), but this too awaits evaluation in controlled trials.

The management of personality disorders


As well as deciding the diagnosis, the strengths and weaknesses of the individual should be assessed. Strengths are important, because treatment should build on favourable features as well as attempting to modify unfavourable ones. The patient’s circumstances should be assessed next.

General aims of management

Although there has been some progress in effecting change in personality disorders, management still consists largely of helping people to find a way of life that conflicts less with their character. Whatever treatment is used, the aims should be modest and considerable time should be allowed to achieve them. A trusting and confiding relationship is the basis of treatment. However, care should be exercised to avoid excessive dependency. Group treatment can sometimes be a way of lessening dependency and increasing social learning.

Often more than one professional is involved in the care of these patients, helping with pharmacological, psychological, and social aspects of care. This can be a risky strategy, and close collaboration is needed to avoid inconsistencies of approach. Many of these patients react badly to changes in staff, which may re-enact painful losses, rejections, or separations in their earlier life. One consistent, albeit limited relationship may be a much better learning experience than a range of contacts, no matter how skilled they are.

The overall plan should include attempts to help the patient to have less contact with situations that provoke difficulties, and more opportunity to develop assets in their personality. They should be encouraged to take an active part in planning their care, and the reasons for decisions should be explained clearly and discussed fully. Helping them to avoid adding to their problems by misusing drugs or alcohol is essential. Even if no major improvement is achieved, this approach may stabilize the situation until some fortuitous change in the patient’s life brings about improvement.

Choice of psychotherapy. Psychodynamic or cognitive psychotherapy may be explored when the patient is well motivated and enough stability has been achieved to enable them to focus on the treatment. The approach will depend on the type of personality disorder (see below), on local availability, and on the patient’s own preference.

Choice of medication. Medication plays only a small part in the treatment of personality disorders. It should not be the first choice, and when it is prescribed it should be part of a wider plan that embraces psychological and social care. Anxiolytic medication should be used with special caution because of the risk of dependency. Anti-depressants are used mainly for associated mood disorder—a common reason for the worsening of the emotional and behavioural problems associated with the personality disorder. The use of antidepressants in borderline personality disorder is discussed below. Antipsychotics are occasionally useful to reduce aggressive behaviour.

Organization of services. People with personality disorder can be cared for by a single practitioner who integrates the psychological, pharmacological, and social aspects of treatment, calling on specialist help when needed. This makes it easier for the practitioner to understand and help with relationship problems, and it emphasizes the essential unity of the person’s diverse problems. However, it can be difficult to sustain when the personality disorder is severe. Alternatively, care can be provided by a community team with a key worker who coordinates the contributions of other team members. Since the recent publication of the Department of Health paper, Personality Disorder: No Longer a Diagnosis of Exclusion (National Institute for Mental Health in England, 2003) in the UK, specialist teams have been established specifically for the treatment of personality disorder. Their configuration varies enormously, but a form of therapeutic day programme based on therapeutic community principles and often incorporating principles of mentalization is commonly the central feature (Bateman and Fonagy, 2001). For a review of services for patients with personality disorder, see Bateman and Tyrer (2004b).


Progress is often achieved as a series of small steps by which the person gradually moves closer to a satisfactory adjustment. Setbacks are common, but they can be used constructively, as it is at these times that the patient is most likely to be willing to confront their problems. Although therapists should try long and hard to help patients with personality disorder, they should recognize that some people may not benefit.

The management of specific personality disorders

The type of personality disorder is not always a good guide to the choice of effective treatment. Nevertheless, some associations have been described, mainly on the basis of clinical experience, and these will be considered next.

Patients with paranoid personality disorder do not engage well with treatment because they are touchy and suspicious. Patients with schizoid personality disorder avoid close personal contact, and often drop out after a few sessions of treatment. If they can be persuaded to continue they tend to intellectualize their problems and question the value of their treatment. The therapist should try to help these people to become more aware of their problems and respond to them in ways that cause fewer difficulties. For both disorders progress is slow and the results are limited. Exploratory psychotherapy is unlikely to succeed, and medication is generally unhelpful.

Treatment of antisocial (dissocial) personality disorder usually consists of the general measures discussed above. Medication has little value, except that an antipsychotic drug may have the temporary effect of calming aggressive behaviour that has arisen in response to increased stress, and antidepressant medication may be needed for a comorbid depressive disorder. Lithium has been claimed to reduce aggressive behaviour in some patients (see p. 147). However, it should not be prescribed unless it is reasonably certain that the patient will collaborate closely with the dosage regimen and the other requirements for safety (see pp. 512 and 552). Anxiolytic drugs should be avoided because they may cause disinhibition and dependency.

Patients with obsessional personality disorder do not respond well to psychotherapy. Unskilled treatment can lead to excessive morbid introspection which leaves the person feeling worse rather than better. Treatment should be directed towards avoiding situations that increase the patient’s difficulties, and developing better ways of coping with stressful situations. Patients often seek help during an associated depressive disorder, and it is important to identify and treat this comorbid condition.

Patients with avoidant (anxious) personality disorder generally have low self-esteem, and fear disapproval and criticism. They can be helped by a therapeutic relationship in which they feel valued and able to reconsider their perception of themselves. As with obsessional personality disorder, the possibility of a comorbid depressive disorder that requires treatment should be considered.

Patients with dependent personality disorder are usually helped more by problem-solving counselling in which they are encouraged to take increasingly more responsibility for themselves. Medication should be avoided unless there is an associated depressive disorder.

Therapeutic communities and Dangerous Severe Personality Disorder units

Some of the earliest therapeutic communities, such as the Henderson Hospital in England, specialized in the treatment of antisocial personality disorder. (For a description of the principles of this form of treatment, see p. 595). No controlled evaluation of this treatment has been carried out, but it has been widely applied within prison services, where the results (measured pragmatically in terms of outbreaks of behavioural disturbance) seem to be relatively promising. The last decade has witnessed a resurgence of interest in the psychological treatment of antisocial personality disorder. This reflects society’s increasing intolerance of risk, and greater precision in the understanding and measurement of psychopathy.

Although psychopathy and antisocial personality disorder are not synonymous, the overlap is considerable. The Psychopathy Checklist-Revised (PCL-R) developed by Hare (1991) reliably identifies high-risk psychopaths. The administrative category of dangerous severe personality disorder (DSPD) was introduced in England in 2000, and pilot units were set up in high secure hospitals and maximum security prisons for their treatment. The determination of DSPD (which is not a diagnosis) is based on psychopathy scores, personality disorder diagnosis, and an assessment of high risk. The treatment is anticipated to take several years, and is mainly based on CBT and some specific offending programmes developed for prisons. It is unclear how effective this controversial and highly expensive treatment will prove to be.

Histrionic personality disorder

These patients make many demands on their carers. Frequent problems include attempts to impose impractical conditions on treatment, requests for inappropriate medication, and demands for help at unreasonable times. Other problems include seductive behaviour, threats of self-harm, and attempts to prolong interviews unreasonably. It is important to identify these problems early on, because once established they cannot be easily controlled. The behaviours should be discussed with the patient, and clear limits set by offering appropriate help while explaining which behaviours cannot be accepted. Treatment should focus on developing more adaptive ways of responding to stressful situations. Medication has little value unless there is a comorbid depressive disorder.

Borderline personality disorder

The general problems in managing borderline personality disorder are similar to those encountered with histrionic personality disorder (see above). This is not surprising, given the diagnostic overlap. However, borderline personality disorder has been more thoroughly researched, and more therapeutic options have been established. Problem-solving counselling (see p. 576) is sometimes helpful. Recent guidance from the National Institute for Health and Clinical Excellence (2009c) advised against the use of medication unless a clear comorbid disorder (e.g. major depression) is present. However, a meta-analytical review by Lieb et al. (2010) concluded that there was some evidence, albeit limited, that antipsychotic drugs and anticonvulsant mood stabilizers could have beneficial effects on aspects of borderline personality disorder, such as affective instability, impulsive behavioural dyscontrol, and cognitive perceptual symptoms. Nevertheless, it does not appear likely that current drug treatments are able to modify core borderline symptomatology such as fear of abandonment, feelings of emptiness, and identity disturbance.

Several kinds of psychological treatment can be tried if resources are available, with encouraging results reported for dialectical behaviour therapy (Linehan et al., 1991) and psychodynamic day units, although none has been demonstrated to have a substantial and reliable effect. The following points should be considered.

Dynamic psychotherapy

People with borderline personality disorder can be difficult to treat with the usual forms of dynamic psychotherapy, and over-optimistic treatment may result in reduced emotional control and increased impulsiveness. Kernberg (1993) described an approach called expressive psychotherapy, which has been claimed to give good results (Stone et al., 1987), but no randomized controlled trials have been reported. For a review of dynamic psychotherapy for borderline personality disorder, see Higgitt and Fonagy (1993).

Dialectical behaviour therapy (DBT)

This type of therapy is described on p. 588. As noted above, this treatment is one of the few that have been evaluated in a controlled trial—with female borderline patients who had made repeated suicide attempts. DBT was effective in reducing self-harm, but the effect on the personality disorder was not assessed directly. It is an extremely demanding therapy, and most of the benefits were lost on longer follow-up (Linehan et al., 1991).

Group psychotherapy

This treatment has the advantages that transference relationships are spread over the group instead of being focused on the therapist, and group members’ comments on acting out behaviour may be accepted more readily than those of the therapist.

Psychoanalytically oriented ‘mentalization’ day treatment

This approach has been subjected to two small controlled trials (Bateman and Fonagy, 1999, 2001), which showed significant sustained improvement. It is currently very much in favour. The treatment combines analytic groups and psychoanalytic individual treatment, and is designed to increase the patient’s ability to better understand their own mental state in the first instance, and through this the mental states of others. The day units emphasize a team approach and an accepting psychologically informed role for all staff. Nursing staff find that the approach can be learned relatively quickly, is rewarding to use, and can be generalized to other clinical settings.

Ethical problems

Stigma and patient involvement

From the early ideas of moral insanity to the present time, the concept of personality disorder has been linked to a moral judgement. People who are told that they have a disordered personality believe there is something inadequate about who they are. This can be entrenched by some professionals’ attitudes, and by the difficulty that people with personality disorder experience in accessing psychiatric and other services. Once the diagnosis has been made it tends to stick and to affect the person’s subsequent care, as well as the way that other people think of them over the longer term. For this reason, many clinicians are reluctant to make the diagnosis.

This situation is now changing, and the recent policy document, Personality Disorder: No Longer a Diagnosis of Exclusion (National Institute for Mental Health in England, 2003), has highlighted the need for all psychiatrists to engage with personality-disordered individuals. The existence of a personality disorder is associated with increased rates of other psychiatric disorders. Even if a psychiatrist does not believe that they can change the patient’s personality, that is no reason not to treat the associated problems. This renewed interest has also led to the establishment of specific psychotherapeutic teams and services for personality-disordered patients. These services are very varied in their form (reflecting the absence of strong evidence for specific treatments) but, as mentioned above, many of them are centred around a psychodynamic day service (Bateman and Tyrer, 2002).

Further reading

Gelder MG, López-Ibor JJ Jr and Andreasen NC (eds) (2009). Section 4.12: Personality disorders. In: The New Oxford Textbook of Psychiatry. Oxford University Press, Oxford. (The seven chapters in this section provide a comprehensive review of personality disorder.)

Gunderson JG and Hoffman PD (eds) (2005). Understanding and Treating Borderline Personality Disorder: a guide for professionals and families. American Psychiatric Press Inc., Washington, DC.

Schneider K (1950) Psychopathic Personalities (trans. MW Hamilton). Cassell, London. (A classic text of great importance in the development of ideas about personality disorder.)