4.2. Personality Disorders
Master Psychology
________________________________________________________________
, Theme 2. Assessment and treatment of personality disorders
Sources
Emmelkamp & Kamphuis – Chapter 2, 5-9, in particular Chapter 6 (p. 127-137)
Martin et al. (2010) in Dobson – Chapter 10
Gunderson (2016)
Adshead et al. (2012)
Emmelkamp & Kamphuis – Chapter 2. Diagnosis and assessment
Validity of the concept of personality disorder
- Do PDs exist? What is normal and abnormal personality variation, and are they qualitatively
or quantitively different? How many PDs are there?
- There are at least 3 levels to consider: The conceptualization of the constructs themselves,
the formulation of the constituent (DSM) diagnostic criteria sets, and the instruments used
to assess the constructs.
The constructs
- Are normal and abnormal personality qualitatively different, or is the difference a matter of
degree?
- Categorical, all-or-none formulations of PDs imply that there should be an identifiable,
nonarbitrary cut-off point to demarcate where normal personality ends and abnormal
personality begins; but others have doubted this approach.
- Meehl (1992) has developed taxometrics, a family of statistical procedures that test
between categorical (taxonic) and dimensional (nontaxonic) models.
- Shown that PDs represent a mix of latent categories and dimensions, so neither
categorical or dimensional approaches of latent structure have generalized
applicability throughout Axis-II.
- Taxometric testing showed that schizotypal PD, schizotypy, and antisocial PD there is
a categorical approach, and for borderline PD there is a dimensional approach.
- Another approach is to define personality pathology by the nature and associated domains
of impaired functioning, e.g., inflexibility, self-defeating, lack of humor etc.
- Is personality pathology (Axis-II) different from clinical syndromes (Axis-I) enough to warrant
a separate axis?
- Distinguishing factors should be stability, age of onset, treatment response, insight,
comorbidity and symptom specificity, and etiology.
- However, others argued that these criteria are not specific to Axis-II disorders, and
that PDs are variants of major clinical syndromes of Axis-I.
- In conclusion, it is an elusive goal to come up with a clinically useful demarcation criterion
for personality pathology, so dimensional models have been developed instead.
- There is little evidence to justify a separate axis for personality pathology, as the
differences with the Axis-I disorders appear rather unsystematic.
, The DSM personality disorders
- Does the specific set of DSM PDs exist?
- The DSM aimed for a categorical, all-or-none representation of the PDs, although the PD-
NOS and subthreshold diagnoses allow for some dimensionality.
- Categories seem efficient for e.g., treatment and heuristics in decision-making.
- PDs are diagnosed in two steps, the first being meeting the general criterion (the first short
story, criterion A), and then counting the presence of specific criteria.
- Evolvement of PDs in the DSM:
- DSM 2: PDs did not have a separate axis, and involved narratives rather than
operational criteria.
- DSM 3: Started with the operational criteria, and fine-tuning them in the revised
DSM 3, but there was still little empirical evidence.
- DSM 4: Dropped some diagnoses (e.g., masochistic and sadistic PD), some disorders
got more or less criteria.
- Unlikely that the DSM 4 and 5 have the last word about personality pathology.
- Deciding on optimal levels of lumping and splitting is important, e.g., are avoidant and
schizoid PD the same?
- Authors have argued that they differ significantly on the motivation for social
isolation, so they need to be ‘split’.
- One way to empirically find out the structure of personality pathology as shown in the DSM,
is through factor analyses.
- The results have been mixed, some find a good fit, others not.
DSM criteria sets
- Current DSM uses polythetic criteria that imply an all-or-none diagnosis, so a PD is classified
as a dichotomous, categorical phenomenon, someone has it or not, so a specified minimum
number of criteria can be defined to demarcate when normal personality ends and a PD
begins; also, each criterion weights equally towards the diagnosis and no criterion is
essential.
- Polythetic criteria for equal weight: Issues of heterogeneity and diagnostic efficiency:
- No single criterion is absolutely required or essential to the disorder, but they are
alternative definers of the disorder, with a certain minimum number needed for the
diagnosis to be present.
- Polythetic criteria invite heterogeneity within diagnostic classes, so quite a lot of
diversity of individuals meet the same diagnosis.
- Also, all the criteria are of equal importance, and receive equal weight towards the
diagnosis.
- Alternative models are additive models (more criteria met leads to higher
probability of the presence of the diagnosis) or weighting models (some criteria are
more equal than others in contributing to accurate diagnosis).
- E.g., for schizoid PD:
High sensitivity and low specificity: No close friends other than first-degree
relatives.
Low sensitivity and high specificity: Indifference to praise and critics of
others.
Master Psychology
________________________________________________________________
, Theme 2. Assessment and treatment of personality disorders
Sources
Emmelkamp & Kamphuis – Chapter 2, 5-9, in particular Chapter 6 (p. 127-137)
Martin et al. (2010) in Dobson – Chapter 10
Gunderson (2016)
Adshead et al. (2012)
Emmelkamp & Kamphuis – Chapter 2. Diagnosis and assessment
Validity of the concept of personality disorder
- Do PDs exist? What is normal and abnormal personality variation, and are they qualitatively
or quantitively different? How many PDs are there?
- There are at least 3 levels to consider: The conceptualization of the constructs themselves,
the formulation of the constituent (DSM) diagnostic criteria sets, and the instruments used
to assess the constructs.
The constructs
- Are normal and abnormal personality qualitatively different, or is the difference a matter of
degree?
- Categorical, all-or-none formulations of PDs imply that there should be an identifiable,
nonarbitrary cut-off point to demarcate where normal personality ends and abnormal
personality begins; but others have doubted this approach.
- Meehl (1992) has developed taxometrics, a family of statistical procedures that test
between categorical (taxonic) and dimensional (nontaxonic) models.
- Shown that PDs represent a mix of latent categories and dimensions, so neither
categorical or dimensional approaches of latent structure have generalized
applicability throughout Axis-II.
- Taxometric testing showed that schizotypal PD, schizotypy, and antisocial PD there is
a categorical approach, and for borderline PD there is a dimensional approach.
- Another approach is to define personality pathology by the nature and associated domains
of impaired functioning, e.g., inflexibility, self-defeating, lack of humor etc.
- Is personality pathology (Axis-II) different from clinical syndromes (Axis-I) enough to warrant
a separate axis?
- Distinguishing factors should be stability, age of onset, treatment response, insight,
comorbidity and symptom specificity, and etiology.
- However, others argued that these criteria are not specific to Axis-II disorders, and
that PDs are variants of major clinical syndromes of Axis-I.
- In conclusion, it is an elusive goal to come up with a clinically useful demarcation criterion
for personality pathology, so dimensional models have been developed instead.
- There is little evidence to justify a separate axis for personality pathology, as the
differences with the Axis-I disorders appear rather unsystematic.
, The DSM personality disorders
- Does the specific set of DSM PDs exist?
- The DSM aimed for a categorical, all-or-none representation of the PDs, although the PD-
NOS and subthreshold diagnoses allow for some dimensionality.
- Categories seem efficient for e.g., treatment and heuristics in decision-making.
- PDs are diagnosed in two steps, the first being meeting the general criterion (the first short
story, criterion A), and then counting the presence of specific criteria.
- Evolvement of PDs in the DSM:
- DSM 2: PDs did not have a separate axis, and involved narratives rather than
operational criteria.
- DSM 3: Started with the operational criteria, and fine-tuning them in the revised
DSM 3, but there was still little empirical evidence.
- DSM 4: Dropped some diagnoses (e.g., masochistic and sadistic PD), some disorders
got more or less criteria.
- Unlikely that the DSM 4 and 5 have the last word about personality pathology.
- Deciding on optimal levels of lumping and splitting is important, e.g., are avoidant and
schizoid PD the same?
- Authors have argued that they differ significantly on the motivation for social
isolation, so they need to be ‘split’.
- One way to empirically find out the structure of personality pathology as shown in the DSM,
is through factor analyses.
- The results have been mixed, some find a good fit, others not.
DSM criteria sets
- Current DSM uses polythetic criteria that imply an all-or-none diagnosis, so a PD is classified
as a dichotomous, categorical phenomenon, someone has it or not, so a specified minimum
number of criteria can be defined to demarcate when normal personality ends and a PD
begins; also, each criterion weights equally towards the diagnosis and no criterion is
essential.
- Polythetic criteria for equal weight: Issues of heterogeneity and diagnostic efficiency:
- No single criterion is absolutely required or essential to the disorder, but they are
alternative definers of the disorder, with a certain minimum number needed for the
diagnosis to be present.
- Polythetic criteria invite heterogeneity within diagnostic classes, so quite a lot of
diversity of individuals meet the same diagnosis.
- Also, all the criteria are of equal importance, and receive equal weight towards the
diagnosis.
- Alternative models are additive models (more criteria met leads to higher
probability of the presence of the diagnosis) or weighting models (some criteria are
more equal than others in contributing to accurate diagnosis).
- E.g., for schizoid PD:
High sensitivity and low specificity: No close friends other than first-degree
relatives.
Low sensitivity and high specificity: Indifference to praise and critics of
others.
