Avoidance of Disclosure

Personality disorder may be undetected or inaccurately assessed due to difficulties engaging in the personality assessment process on the part of patients and clinicians, and due to limitations in applying current measures and constructs for this purpose. Specifically, many older adults grew up in an age when it was not acceptable to openly talk about discordant feelings or interpersonal conflicts. Perceived stigma may impede inquiry into deviant inner experiences or behaviours, and would also make it unlikely that an older adult would have received a previous personality disorder diagnosis of which they can inform their current physician²¹. In a complementary manner, some clinicians adhere to the erroneous belief that older adults simply do not have personality disorders¹⁴. In addition, personality disorders may increase risk for developing Axis I disorders and it is often for this subsequent condition that patients seek professional help. Axis I conditions may be easier to recognize and accept, for both the clinician and patient, and are therefore often deemed the focus of diagnoses and treatment.

Assessment Measures

The structured interview is often considered an ideal assessment tool for diagnoses of personality disorder, but is subject to biases and inaccuracies when assessing older adults. As longitudinal evidence is required for a diagnosis of personality disorder, older adults may be asked to recount events and trends from 30–50 years before, and it is often doubtful that such information is valid or reliable²¹. Furthermore, measures of personality disorder cannot fully account for the potential influences of current psychological states, medical or neurological problems, situational or environmental contexts, or sociocultural roles on reporting of current symptoms or behavioural patterns²¹. Self-report inventories face similar vulnerabilities^14,15, and must be supplemented by a practitioner’s clinical judgement.

Diagnostic Criteria

DSM-IV criteria for many personality disorders may not be generalizable to experiences and behaviours displayed by older adults. For example, symptoms of personality disorder such as increased dependency, social withdrawal, or rigid and moralistic thinking may be misinterpreted by clinicians as part of the natural ageing process and therefore may not enable discrimination between normal and pathological behaviour. In a large epidemiological survey which compared younger and older adults with equivalent levels of personality disorder pathology, an item analysis found evidence for age bias in two thirds of personality disorder criteria sets^22,23. Difficulties determining when impairment in social and occupational functioning is significant enough to warrant a diagnosis of personality disorder may also impede diagnosis. In the lives of older adults, where retirement and smaller social circles are common, this may not be an accurate measure of impairment^14,21. Finally, discrete behaviours upon which inferences as to personality are based may be more fallible in older adulthood. As individuals age, changing physical and developmental capabilities may limit their repertoire of behaviour, and thus the likelihood that behaviours are completely representative of their inner character. In this context it may be dangerous to rely on a solely behavioural framework for diagnoses^11,24.

One of the most prominent barriers to accurate detection of personality disorder pathology is uncertainty as to whether and how age-related biopsychosocial changes affect the phenomenology of personality disorder, and a definitive answer remains elusive. Tyrer and Sievewright²⁵ suggested that personality disorders can be divided into mature and immature types, in which the mature types, such as obsessive–compulsive, paranoid, schizoid and schizotypal, remain stable with age. In contrast, the more dramatic disorders, termed immature (borderline, narcissistic, histrionic), are hypothesized to decrease in intensity with age.

Empirical studies of normal personality find evidence for both stability and change over the life span. Major personality traits, such as those represented in the five-factor model (FFM) of personality (neuroticism, extraversion, openness, agreeableness and conscientiousness) have a prominent genetic basis, appear universal across cultures, and demonstrate substantial stability over the life course^10,26,27. Hence, one might expect that maladjusted individuals with personality disorder tend to remain so as they age. However, a recent meta-analysis of 92 longitudinal studies of reports of personality traits over the life span found that ‘people tend to become more socially dominant, conscientious, and emotionally stable through midlife’ (p. 21)²⁸, suggesting a positive maturational effect for most adults likely linked to age-graded roles (e.g. work, marriage, parenthood). Personality processes, such as emotion regulation, also display favourable age-related changes in cross-sectional studies of younger and older adults^29,30. Such findings are in accord with case series reports and cross-sectional studies of late life personality disorder³¹, and suggest that normative, maturational effects may favourably modify the phenomenology of personality disorder in older adults, especially in areas such as affect regulation and impulsivity.

Longitudinal studies of personality disorder are limited, but extant research provides further insight into stability and change in the phenomenology of personality disorder over the life span. McGlashan³² performed a longitudinal study on inpatients diagnosed with BPD, unipolar depression, or schizophrenia. He obtained follow-up data for over 288 participants (mean age of 47) at an average of 15 years after discharge (range 2–32 years). Compared to the other groups, the BPD group displayed the highest level of global functioning at follow-up, particularly in those conducted 10–19 years after discharge compared to subsequent follow-ups. Instrumental and global functioning increased for participants with BPD, whereas their interpersonal relationships often remained problematic or non-existent. McGlashan speculated that as patients with personality disorder age and lose employment capabilities, earlier improvements in instrumental functioning become less relevant and the lack of close relationships becomes more significant, potentially resulting in diminished functioning later in the life span. Similarly, a 16- to 45-year longitudinal follow-up of male inpatients diagnosed with antisocial personality disorder (ASPD) found that over half had improved or remitted, but that many also had enduring problems with occupational performance, interpersonal discord and social isolation³³.

Perry³⁴ reviewed 26 longitudinal studies on personality disorders, with the majority focusing on BPD. At an average follow-up of 8.7 years, only 57% of BPD patients continued to meet diagnostic criteria. This relationship was reportedly linear, with approximately 3.7% of borderline cases remitting each year. A more recent 10-year follow-up study of inpatients with BPD found that 88% achieved remission (many within 2 years), and that symptoms such as impulsivity were likely to resolve, while chronic dysphoria and interpersonal symptoms persisted^20,35.

In sum, investigations into the temporal stability of personality traits, disorders and processes find evidence for substantial continuity, particularly for major traits and for interpersonal functioning, but also for favourable change in select areas, such as emotion regulation and impulsivity. Although we are unaware of studies that follow personality disorder into older adulthood, such findings suggest shifts in the phenotypic expression of personality disorder, and provide further empirical support for recommendations that clinicians consider possible attenuation and exacerbation of personality disorder traits and symptoms over the life span, apply time course criteria flexibly, and appreciate the salience of subsyndromal presentations of personality disorder in late life^11,24.

In addition to the barriers described above, clinicians who work with older adults face added diagnostic challenges, as personality disorder is often embedded in a complex matrix of age-related changes in cognitive and physical health, functional ability, interpersonal roles, and coping resources and behaviours. Although co-morbidity is the rule rather than the exception in geriatric practice, the DSM-IV-TR ‘General Criteria for Personality Disorder’¹⁶ require that manifestations of personality disorder be disentangled from major psychiatric syndromes, medical and neurological disorders, situational behaviours or coping styles, and social/cultural roles. Among the most vexing issues are difficulties differentiating personality disorder from (i) an unremitted Axis I disorder; (ii) context-dependent roles and behaviours; (iii) subtle neurocognitive impairment; or (iv) a difficult doctor–patient relationship.

Personality Disorder Versus Unremitted Axis I Disorder

In practice settings an older person’s long-standing and unremitted Axis I pathology, especially if associated with features of irritability, negativism and caregiver exhaustion, is often attributed to personality pathology. Although personality pathology is correlated with poorer treatment outcomes and persisting decreases in functioning in older depressed samples^3–5, it is not a ‘diagnosis of exclusion’ and such a conclusion may confuse personality disorder with illness duration and/or treatment responsiveness. This challenge may be addressed through careful application of the general diagnostic criteria for personality disorder, supplemented by dimensional personality measures and informant reports, with special attention to temporal relationships between the onset of Axis I and Axis II symptoms. If core features of personality disorder persist following maximal treatment of Axis I syndromes with combination therapies, a definitive diagnosis of personality disorder may be warranted.

Personality Disorder Versus Context-Dependent Roles and Behaviours

Older adults confronted with significant, ongoing life stressors such as serious medical illness (e.g. becoming disabled), interpersonal loss or role transition (e.g. becoming a caregiver), or the need to navigate a novel social environment (e.g. long-term care setting) may exhibit behaviours that mimic aspects of personality pathology. For example, confronted with medical problems or physical limitations, some elderly may express their distress somatically as if to manipulate the responses of caregivers; others may become antagonistic and demanding; still others may regress to a child-like dependency. Although significant psychosocial stressors may lead underlying per sonality disorder features to emerge²⁴, it is often the case that such behaviours are context-dependent, or only loosely tied to personality traits rather than disorders. A longitudinal history, supplemented by informant data and dimensional measures of personality, may help clarify whether personality disorder symptoms and behaviours preceded the change in life context, thus meriting a diagnosis of personality disorder.

Personality Disorder Versus Subtle Neurocognitive Impairment

Neurological disorders such as Alzheimer’s disease, Parkinson’s disease, and vascular and frontotemporal dementias are common in late life, and both their characteristic features and associated behavioural manifestations are well recognized. Recent evidence suggests that personality changes (such as increased egocentricity, emotional lability, rigidity or apathy) may represent early or prodromal manifestations of dementia³⁶