Theories of personality

In Ancient Greece, physicians attributed individual differences in personality to imbalances of bodily fluids or humors; other popular theories have included the influence of the stars’ positions at birth, body build and skull shape (Reference Knutson, Heinz and PankseppKnutson 2004). In the 20th century, research into personality moved to the level of the psychological, although still influenced by dominant social assumptions such as gender or racial difference. Freud emphasised the role of innate drives, an early account of what we might now understand as the genetic basis of stress responses. He is also attributed as being the first to describe the concept of ‘defences’ against stress and their effect on the expression of adult personality. Later theorists, such as Klein and Bowlby (in somewhat different ways), emphasised the importance of the interaction between the child’s innate individual features and the environment in the development of normal personality functioning.

In the 1960s, Allport highlighted the role of ‘traits’ in the makeup of personality, which he defined as the ‘dynamic organization […] of those psychophysical systems that determine characteristics of behaviour and thought’ (Reference AllportAllport 1961: p. 28). Factor analysis enabled the description of personality in terms of dimensions such as dominance and affiliation (Reference Freedman, Leary and OssorioFreedman 1951).

Like Allport, some theorists see the self as an organising principle of a number of personality traits, some of which are inherited, and some of which develop in relation to early social experience with others. Others see the self as the subjective experience of personal identity (the ‘I’ of experience) and the personality as the objective aspect (the ‘me’ that others experience). More recent concepts of personality link it with related concepts such as the self and personal identity (Reference McAdamsMcAdams 1992).

Personality and its disorders

The concepts of self, person and identity raise many complex questions for psychiatrists. Space does not permit an analysis of all of these, but an important one is the question of change and stability of the personality. Within general medicine, changes in function are usually associated with pathology. But, in relation to personality function, it is not clear to what extent change can be expected over time or in response to different situations, nor whether such changes indicate pathology or are flexible responses to different demands. Similarly, personality disorder is defined in ICD-10 (World Health Organization 1992) and DSM-IV-TR (American Psychiatric Association 2000) as ‘enduring’ characteristics from early childhood; yet changes in personality after brain trauma have been recognised since the 19th century, and DSM-IV-TR recognises acquired personality disorder after exposure to traumatic stressors. If personality disorder can be acquired as a result of psychological or neurological change, then it can be the result of pathology, and not a permanent feature of a person.

Second, in everyday life, it is obvious that there are discrepancies between how people see themselves and how others see them: but it is not so clear what the discrepancies mean. This has implications for how disorders of personality are detected and assessed. Consider this example: three people experience a man as bullying, two see him as assertive and he sees himself as threatened. Each of these perspectives has some validity, which suggests that assessment of personality, or any presumed disorder of personality, requires not only self-report data but also other-report data. However, unlike other domains of medicine, where informants may have objective and valid information about a patient’s symptoms, it is by no means clear whether informants can provide information that is not confounded by their personal experience of the patient and their interpersonal relationship.

Third, and related, individual emotional experience is probably not purely individual, or at least not purely internal. An individual who experiences a strong emotion is able to transmit that experience to others (and vice versa), probably through the operation of mirror neurons (Reference GalleseGallese 2001), which fire when another’s emotional experience is witnessed. The closer the emotional tie, the more pronounced the experience: we do feel the pain of others, especially those with whom we are in close relationships (Reference Singer, Seymour and O'DohertySinger 2004). Caregivers of infants, both human and non-human, regulate the stress responses of those infants through attachment relationships. This means that a well-functioning personality has to regulate and respond to another’s feelings, not just to their own experience. Some disorders of personality have a greater impact on social relations than others, especially in relationships that involve the attachment system, such as parenting and intimate relationships.

The diagnosis of personality disorder

If personality is a property of human organisms, then there is no particular reason to suppose that it cannot become dysfunctional. The difficulty is with the conceptualisation of that dysfunction. If ‘personality’ is conceived as a limited number of traits in each individual, which are largely genetically driven, then a limited number of categories of disorder may seem appropriate. This categorical approach to diagnosis underpins the ICD and the DSM systems, both of which distinguish personality disorders from other psychiatric disorders. The principal features of personality disorders in both systems are summarised in Box 2.

There are a number of criticisms of the international classificatory systems. First, they do not address the primary aspects of personality pathology, namely social relationships with others. Second, a dimensional (rather than categorical) account of personality and its disorders might offer improved descriptions of the social and interpersonal dysfunctional aspects of personality disorder. On this account, personality disorder reflects the abnormal functioning of normal dimensions at different times and in different settings: individuals can develop degrees of severity of personality disorder (e.g. Reference Yang, Coid and TyrerYang 2010) or their personalities might become disordered for a period, then recover. Third, as has already been described above, there is now considerable evidence that personality disorder is not enduring (Reference Seivewright, Tyrer and JohnsonSeivewright 2002, Reference Seivewright, Tyrer and Johnson2004; Reference Shea, Stout and GundersonShea 2002).

Signs and symptoms of personality disorder

Like any other mental disorder, personality disorders have signs and symptoms. These constitute the three major components described in Box 3.

At a general level, patients with personality disorders across all clusters find it hard to make and maintain relationships in any social domain. The more severe the personality disorder, the more socially isolated the individual is likely to be (Reference Yang, Coid and TyrerYang 2010). In broad terms, those with personality disorder demonstrate three distinct patterns of social engagement (or lack of it) with others. Although these patterns do not neatly map onto DSM clusters A, B and C (Box 4), people with cluster A disorders tend to move away from social attachments, either by taking up a frightened position with regard to others, or by flight into safety through isolation. Conversely, people with cluster C disorders move towards others as sources of support and dependence, and even if those sources of support become persecutory, individuals are relatively unable to disengage and assert themselves. Finally, people with cluster B disorders often tend to move against others. Ironically, to move against others, they have to first move towards them. Hence, their interactions are characterised by an ambivalence to social encounters that is most visible in those with borderline and narcissistic personality disorders.

Thus, a man with a mild degree of a cluster A (schizoid) disorder may still be able to form relationships in the workplace, or have some degree of emotional attachment to a friend or family member. A woman with a moderate degree of a cluster B borderline personality disorder is unlikely to be able to maintain emotionally intimate relationships with sexual partners, and may have complex relationships with carers, but may still be able to maintain work relationships. These two individuals will be quite unlike a man with a severe degree of cluster B antisocial personality disorder, who is unlikely to have ever been able to connect to any social group for work or social purposes, and close emotional relationships will be unknown to him (and he may indeed treat them with contempt). His chances of being both a criminal rule breaker and an exploiter of the vulnerable are much higher than for individuals with other types of personality disorder.

The symptoms of personality disorder are listed in Box 5. Symptoms of negative affect vary according to cluster, but include anxiety, irritability, low mood, intense distress, feelings of rage, fear of abandonment, the perception that others are threatening or attacking the individual, and dissociative experiences associated with stress. Individuals with emotionally impulsive borderline personality disorder frequently describe brief periods of pseudo-hallucinations that almost invariably take the form of voices telling them to harm themselves or others. People with antisocial personality disorder may report little distress, but complain of persecutory anxiety or anger with others whom they perceive to have let them down. They may also enjoy feelings of contempt for other people’s distress, excitement in controlling others and hostility towards dependence and neediness.

The signs of personality disorder (Box 6) are manifested in disturbances of social relationships, i.e. at the interpersonal level, at the boundary between the individual and their social world. The signs include repeated behaviours that are socially rejecting, self-destructive or result in social exclusion, and more elaborate and enduring dysfunctional relationship styles. However, behaviours alone cannot determine the presence of a personality disorder; there must also be evidence of disturbance of affect and arousal regulation. It is also important to consider behaviours that do not attract attention, but are still pathological, such as social withdrawal.

In a particular subgroup of people who behave antisocially, lack of response to other’s distress is noted to be a distinguishing feature (Reference Cleckley and MosbyCleckley 1964). Later researchers, principally Robert Hare, have confirmed the existence of a subgroup of individuals (mainly in the cluster B antisocial subgroup) who have both a pronounced lack of empathy for others and display predatory or cruel behaviours towards others. It has been suggested that these individuals do not recognise facial signals of distress in others and do not detect emotional tone, probably as a result of amygdala dysfunction (Reference BlairBlair 2003). In the most severe cases, this lack of response can be associated with extremes of cruelty and violence to others, and is captured by the clinical concept of ‘psychopathy’ (Reference HareHare 1991).

Although the international classificatory systems have yet to accept psychopathy as a ‘disorder’, its assessment is standard in most forensic psychiatric settings. The standard measure of psychopathy is useful for distinguishing ‘milder’ from more ‘severe’ or ‘extreme’ forms of antisocial personality disorder. However, clinicians not working with forensic patients are unlikely to see such individuals.

On the basis of the above, it may be argued that dysfunctional social relationships are a diagnostic feature of personality disorders that distinguish them from mental illnesses. The counterargument is that many serious mental disorders (especially chronic psychotic disorders and mood disorders) also have profound effects on social relationships. Social isolation is a common problem for many service users with severe mental illness and it is usually caused by relationship breakdown occasioned by aspects of the illness. The difference may be in degree, and the types of relationship breakdown, rather than an absolute difference.

Social difficulties in relationships with mental health professionals

These interpersonal and social difficulties are also inevitably manifest in relationships with healthcare professionals. Patients with personality disorder generally do not take on the conventional ‘sick role’, in which the patient is compliant, obedient and grateful. For these reasons, attempts to care for such patients on the basis of conventional therapeutic relationships are unlikely to succeed, and staff need particular competencies to provide a service for such individuals (Home Office 2005).

Individuals in the different clusters have different approaches to help-seeking from mental health professionals (Reference Tyrer, Milchard and MethuenTyrer 2003). Those in clusters A and C and the more antisocial individuals may rebuff help, or (in the case of antisocial people) denigrate caregivers and therapies. Cluster C patients with dependent personality disorders, however, commonly form submissive, clinging relationships with clinicians rather than rebuffing help. In contrast, people with more borderline personality pathology have highly ambivalent attachment patterns, which mean that they may seek and then reject help. They are likely to be fearful of asking for help, and this fear can cause increasing arousal and ultimately hostility towards those they are approaching for help. There is evidence that help-seeking and care-seeking behaviours (including engagement in and adherence to therapy) are influenced by attachment experiences in childhood (Reference HendersonHenderson 1974; Reference Dozier, Lomax and TyrrellDozier 2001). This suggests that there is a link between personality disorder and attachment history.

Prevalence and incidence

Prevalence data are similar worldwide and recent figures from the World Health Organization show no important or consistent differences across countries (Reference Huang, Kotov and De GirolamoHuang 2009). In the UK, the prevalence of any personality disorder is about 4% overall (Reference Coid, Yang and TyrerCoid 2006) – considerably higher than the prevalence of psychotic disorders.

In primary care, the prevalence of personality disorder is around 10–12%, and it consists mainly of patients with depressive and somatising symptoms. However, the prevalence of personality disorder in general psychiatric out-patients is 33%, rising to about 40% in eating disorder services and 60% in substance misuse services (Reference Herzog, Keller and LavoriHerzog 1992; Reference Sanderson, Wetzler and BeckSanderson 1994; Reference Rounsaville, Kranzler and BallRounsaville 1998; Reference Moran, Jenkins and TyleeMoran 2000; Reference Torgersen, Kringlen and CramerTorgersen 2001).

In forensic services and prisons, the prevalence of personality disorder is 70%, and the principal subtypes are antisocial, borderline and narcissistic (Reference Singleton, Gatwood and MeltzerSingleton 1998). Prisoners’ problems include lack of empathy, social hostility and contempt for weakness, as well as affect dysregulation. In specialist forensic personality disorder treatment settings, virtually all patients have comorbid psychiatric disorders such as substance misuse or depression, and most fulfil criteria for several personality disorders (Reference Duggan, Mason and BanerjeeDuggan 2007).

Prevalence data can be misleading because of selection bias. Services for behavioural conditions such as eating disorder, substance misuse or antisocial behaviour are likely to be ‘selecting’ for comorbid personality disorders that manifest in the particular behaviour. It is important for clinicians not to generalise about personality disorders as a whole on the basis only of the group they see in their service settings.

Pathogenesis

Social dysfunction in adulthood (including both self-harm and antisocial behaviour) is more likely when genetic vulnerability for arousal and affect dysregulation interacts with environmental adversity and negative life experiences during early development (National Scientific Council on the Developing Child 2010). Genetic studies suggest that personality disorder is strongly heritable (Reference Jang, Livesley and VernonJang 1996); one model hypothesises a vulnerability to the replication of genes for proteins that are relevant to neurophysiological processes such as arousal, response times and homeostatic mechanisms. However, it is also clear, from both retrospective and prospective studies, that early childhood adversity is highly relevant to the development of psychological signs and symptoms of personality disorder (Reference Kessler, McLaughlin and Greif GreenKessler 2010). Physical abuse and neglect appear to increase the risk of developing all types of psychiatric morbidity, especially substance misuse. Sexual abuse appears specifically to increase the risk of developing depression and borderline personality disorder.

There are three major theoretical approaches to explaining how early childhood relationships and maltreatment result in the adult interpersonal dysfunction found in personality disorder.

Course of the disorder

Personality disorders are long-term, chronic disorders, with varying degrees of severity (Reference Tyrer and JohnsonTyrer 1996; Reference Yang, Coid and TyrerYang 2010). Some forms have a relapsing and remitting nature, depending on environmental factors and comorbidity. There is evidence that a proportion of individuals gradually experience remission of the disorder over time (Reference ParisParis 2003). Given the link with affect and arousal regulation, one would expect patients with personality disorder to exhibit more signs and symptoms when they are distressed, aroused or depressed. This may explain why psychotropic mood-stabilising or sedating medication is helpful for some individuals with personality disorder.

A rule of thumb is that symptoms of personality disorder will be exacerbated during periods of stress, particularly if the stress is linked to relationships with partners, parents or dependants. As a result, people with personality disorders may behave in socially alienating ways at times of stress – ironically, at the time of their greatest need. Patients with the disorder may consequently be excluded from help or they may reject help, without realising that they are doing so. Appropriate clinical skills are therefore of paramount importance in terms of acknowledging such features as signs of a disorder rather than an expressed view that the person is rejecting services or trying to manipulate professionals.

Severe personality disorder

Personality disorders have not typically been graded in terms of severity, partly because of the tradition of using categorical descriptions. However, it has long been recognised that not all personality disorders cause the same degree of dysfunction. Reference MillonMillon (1981) suggests that borderline, paranoid and schizotypal personality disorders are the most severe types within a social system, because these personality styles characteristically produce significant social incompetence and isolation. Reference BlackburnBlackburn (2000) suggests that individuals with dependent, histrionic, narcissistic and antisocial personality disorders are deemed to be the least severely disordered in terms of social adaptation. However, if severity is defined by the degree of disorganisation caused by a personality disorder within a society, then antisocial and narcissistic personality disorders may both be defined as ‘severe’. Here the meaning of the word ‘severe’ depends on whether one understands it in terms of individual social adaptation or the impact of the individual on a social group.

Severity of personality disorder implies something quite different if the categorical approach of the international classificatory systems is employed. There is some disagreement between the ICD and DSM systems on the types and number of personality disorders that they respectively recognise as existing. It has been suggested that the threshold for making a diagnosis of personality disorder is lower in ICD-10 (Reference Tyrer and JohnsonTyrer 1996), although DSM-IV recognises eleven different types of the disorder, compared with the eight endorsed by ICD-10. The two systems also differ in their respective recommended guidelines for diagnosing personality disorder. The DSM-IV is a more rigid system that advocates a checklist approach to diagnosis, in that a specific number of observable behaviours have to be present for a diagnosis (even though the preamble to DSM-IV warns against a ‘cook book’ approach). The ICD-10, a trait-based system, allows the clinician a greater degree of flexibility in establishing a diagnosis.

Reference Tyrer and JohnsonTyrer & Johnson (1996) suggest a five-point rating of severity: 0, ‘no personality disorder’; 1, ‘personality difficulty’ (any subthreshold criterion of a personality disorder); 2, ‘simple personality disorder’ (one or more personality disorders within the same cluster); 3, ‘complex personality disorder’ (personality abnormalities spanning more than one cluster as diagnosed by the international classificatory systems); and 4, ‘severe personality disorder’ (two or more personality disorders in more than one DSM cluster with one being antisocial personality disorder). A recent study using this rating found different prevalences of personality disorders of different severity: the most severe cases were seen in specialist services and milder conditions were seen in primary care (Reference Yang, Coid and TyrerYang 2010).

Severity could be defined in terms of the harm done to others, even if this would apply only to small subgroup of people with personality disorders. In 1999, a Royal College of Psychiatrists’ working party reviewed the diagnosis of psychopathic and antisocial personality disorders (Royal College of Psychiatrists 1999), and proposed an additional category, defined as those who manifest ‘gross societal disturbance’. A ‘gross’ disturbance was defined as having one cluster B diagnosis and a personality disorder in at least one other cluster also. No detailed rationale was provided for such a grouping, nor did the document explain further what might constitute ‘gross societal disturbance’. However, that same year the UK Home Office and the Department of Health identified a subgroup of offenders with personality disorders who were violent and defined them as having ‘dangerous and severe personality disorder’ or DSPD (Home Office 1999). The DSPD condition, perhaps extrapolating from the recommendations of the College, was defined as the presence of (a) two or more personality disorders, (b) a Psychopathy Checklist score (Reference HareHare 1991) of over 30, and (c) a functional link between the disorder and the violence. The link between severity of personality disorder and severity of risk to others is made quite explicit. This makes sense for a public protection agenda, but has clinical limitations. One objection is that there are many individuals who are dangerous to others as a result of having a single personality disorder, but who do not fulfil criteria (a) or (b). The most obvious group in this category are perpetrators of child and partner abuse.

If personality disorder were to be rated in terms of severity, then this would make it more like depressive disorders and intellectual disability, and less like schizophrenia and bipolar affective disorders, which are described in terms of categorical types rather than dimensional severity. One advantage of a severity rating is that it might help determine workforce competencies and other resource allocation requirements. It might also lay the foundation for research into prognosis and outcomes. For example, several new community services for personality disorder have taken a ‘complex needs’ approach to determining the service configuration and skills mix appropriate for each individual. They categorise a personality disorder as severe if the individual’s problems involve several areas of care. Thus, a patient might have attachment difficulties and also require containment in a secure unit. Such patients typically fulfil criteria for several personality disorders and also have comorbid Axis I conditions.

Personality disorder and mental illness

It is commonly argued that personality disorder is not a mental illness, and/or that it is qualitatively different from mental disorders such as schizophrenia. One argument seems to be principally that people with personality disorders do not lose their capacity for reality testing (e.g. do not have symptoms such as hallucinations and delusions). However, there are a number of counterarguments to this. First, loss of reality testing is not the sole test of whether someone has a mental illness or not. Addiction disorders and neurotic disorders such as depression are still classified as mental illnesses. Further, the presence of psychotic symptoms does not determine mental illness status, since psychotic symptoms occur in many physical disorders, such as delirium or the encephalitides. Finally, even if one does use psychotic symptoms as a defining diagnostic test for mental illness, there is plenty of evidence that some types of personality disorder do involve psychotic symptoms, albeit usually of brief duration.

There are other good reasons to think that personality disorder has more in common with the Axis I conditions than traditionally supposed. First, it has an aetiology that brings about pathology and, like many medical conditions, it afflicts only a minority of the population, who are consequently quantitatively different from the norm. Second, personality disorder produces a pattern of symptoms and signs that are common to a group, and make group members resemble each other. There is evidence of abnormal brain structure and function in personality disorders, and overlap of psychopathology between Axis I and II disorders in terms of cognitive/perceptual organisation, impulse control, affect regulation and anxiety modulation (Reference Sarkar, Duggan, Sarkar and AdsheadSarkar 2012).

Third, there is the issue of comorbidity or co-occurrence of Axis I and Axis II disorders, which suggests at least a close relationship between them. For example, at least one cluster A disorder (schizotypal personality disorder) lies on a continuum with schizophrenia; cluster B disorders, especially borderline and antisocial personality disorders, co-occur with substance misuse (Reference Knutson, Heinz and PankseppKnutson 2004); and cluster C disorders preferentially co-occur with increased rates of somatoform disorders (Reference Tyrer, Gunderson and LyonsTyrer 1997). People with either cluster B or C disorders have a higher risk of comorbidity with all types of Axis I disorders (Reference Dolan-Sewell, Krueger, Shea and LivesleyDolan-Sewell 2001). Conversely, the coexistence of personality disorder with an Axis I disorder can lead to poorer outcomes (Reference Newton-Howes, Tyrer and JohnsonNewton-Howes 2006). In one study of patients with mental disorders managed by community teams, 40% fulfilled criteria for at least one personality disorder (Reference Newton-Howes, Tyrer and AnagnostakisNewton Howes 2010). There seems to be an interaction between two or more mental health conditions such that there is an additive (if not multiplicative) effect in terms of ‘load’ of clinical symptomatology and socioclinical outcomes.

In contrast to the expanding evidence base that there is at least a complex relationship between Axis I and Axis II disorders, there is no new evidence in support of the assertion that (a) personality disorder is fundamentally different from mental illness, or (b) that Axis I and Axis II disorders are alternatives that cannot coexist. Although the multiaxial nature of the DSM should make it obvious that disorders can coexist, it is still common to hear clinicians refer to these conditions as if they were alternatives.

MCQs

Select the single best option for each question stem

1. 1 Personality function includes:

   1. a regulation of individual levels of arousal, impulsivity and affect

   2. b ability to carry out activities of daily living

   3. c being able to relate adversely to others

   4. d ability to manage financial responsibilities

   5. e absence of mental illness.

2. 2 ICD-10 and DSM-IV state that personality disorders:

   1. a begin in adulthood

   2. b are not developmental in nature

   3. c are not mental disorders

   4. d are not due to other mental disorders

   5. e do not deviate markedly from sociocultural norms.

3. 3 The clinical components of personality disorders include:

   1. a disturbed relationships with caregivers

   2. b individual regulation of drug and alcohol consumption

   3. c lack of socially appropriate behaviours

   4. d lack of cognitive and reality distortions

   5. e onset in adulthood.

4. 4 Clinical features of personality disorders include:

   1. a abnormalities in somatic self-identity

   2. b normal psychological self-identity

   3. c lack of attacks on self and others

   4. d absence of psychotic episodes

   5. e absence of major affective episodes.

5. 5 Personality disorders are:

   1. a valid and legitimate clinical syndromes

   2. b not mental disorders

   3. c untreatable

   4. d significantly different from mental illnesses

   5. e abnormal illness behaviours, not genuine illnesses.

MCQ answers

1

a

2

d

3

a

4

a

5

a