ORIGINAL RESEARCH INVESTIGACIÓN ORIGINAL
Redefining personality disorder: a Jamaican perspective
Redefinición del trastorno de la personalidad desde una perspectiva de Jamaica
Frederick W. Hickling*; Vanessa Paisley
Caribbean Institute of Mental Health and Substance Abuse, University of the West Indies, Kingston, Jamaica
OBJECTIVE: To characterize and assess the factor structure of phenomenological features of DSM-IV personality disorder diagnosis in Jamaican patients and determine any similarities with those of traditional criteria, associations with disorder severity, and/or significant relationships between variables to inform the current debate on the relevance of established personality disorder diagnostics.
METHODS: This was a case-control study. All the patients included were seen by one private psychiatric practice from 1974 to 2007. The study sample group (n = 351) were patients diag nosed as having a personality disorder (DSM-IV Axis II). The control group was composed of patients with DSM-IV Axis I clinical disorders, who had not been diagnosed with a personality disorder, and matched exactly on gender, and closely on age, as well as socioeconomic variables.
RESULTS: Of the 351 individuals in the study sample group, 166 (47.3%) were male and 185 (53.7%) were female; 50 (14.2%) were white and 301 (85.8%) were black; 293 (83.5%) were born and raised in Jamaica; and 202 (57.6%) were from socioeconomic classes I and II. Mean age was 33.92 (standard deviation 10.236). Disaggregating the phenomenology, the conventional DSM-IV personality disorder diagnoses disappeared. Factor analysis of 38 clinical phenomena identified five components: psychosis, major depression, power management problems, psychosexual issues, and physiological dependency. Independent t-tests revealed patients without personality disorder had significantly higher mean scores for psychosis; both groups scored equally for depression; and those with personality disorder had significantly higher mean scores on the remaining factors. Analysis of variance indicated these factors differed significantly for three levels of severity (mild, moderate, and severe).
CONCLUSIONS: The phenomenology clustering into three major groups suggested an Axis I (clinical) diagnostic disorder of impulse control and authority and conflict management.
Key words: Personality disorders; Jamaica.
OBJETIVO: Caracterizar y evaluar la estructura factorial de las características fenomenológicas del diagnóstico de trastorno de la personalidad según el DSM-IV en pacientes jamaiquinos, y determinar sus semejanzas con las de los criterios tradicionales, la asociación con la gravedad del trastorno o las relaciones significativas entre las variables con objeto de aportar información al debate actual sobre la relevancia de los diagnósticos establecidos de trastorno de la personalidad.
MÉTODOS: Estudio de casos y controles, donde todos los sujetos fueron atendidos en una práctica psiquiátrica privada de 1974 a 2007. Los casos (n = 351) correspondieron a pacientes diagnosticados con un trastorno de la personalidad, (Eje II del Manual Diagnóstico y Estadístico de los Trastornos Mentales [DSM-IV, por su sigla en inglés]). El grupo de control estuvo integrado por pacientes con trastornos clínicos del Eje I del DSM-IV, sin diagnóstico de un trastorno de la personalidad, emparejados exactamente en cuanto al sexo y estrechamente en cuanto a la edad y a variables socioeconómicas (n = 351).
RESULTADOS: De los 351 individuos del grupo de la muestra del estudio, 166 (47,3%) eran varones y 185 (53,7%) mujeres; 50 (14,2%) eran de raza blanca y 301 (85,8%) de raza negra; 293 (83,5%) habían nacido y crecido en Jamaica; y 202 (57,6%) pertenecían a las clases socioeconómicas I y II. La media de la edad era de 33,92 (desviación estándar 10,236). Cuando se desagregaron las características fenomenológicas, no concordaban con los diagnósticos convencionales de trastorno de la personalidad según el DSM-IV. El análisis factorial de 38 fenómenos clínicos permitió determinar cinco componentes: psicosis, depresión mayor, problemas de manejo del poder, trastornos psicosexuales y dependencia fisiológica. Las pruebas de la t independientes revelaron que los pacientes sin un trastorno de la personalidad obtuvieron puntuaciones medias significativamente mayores para la psicosis; ambos grupos obtuvieron las mismas puntuaciones para la depresión; y los que padecían un trastorno de la personalidad obtuvieron puntuaciones medias significativamente mayores para los factores restantes. El análisis de la varianza indicó que estos factores diferían significativamente según el nivel de gravedad (leve, moderado o grave).
CONCLUSIONES: El agrupamiento de las características fenomenológicas en tres grupos principales sugirió un diagnóstico (clínico), correspondiente al Eje I, de trastorno del control de los impulsos y del manejo de la autoridad y los conflictos.
Palabras clave: Trastornos de la personalidad; Jamaica.
Despite modifications that have been made to the DSM and ICD classification systems,1 there is still much debate over their representation of personality disorders (1, 2), clinical applicability, and proposed replacement (DSM-V) (3-5). Clinicians have found that the criteria of the current DSM (4th edition text revision, DSM-IV-TR) do not accurately depict complex and distinct personality syndromes and have therefore proposed a new classification system that utilizes comprehensive descriptions of patients' inner experiences as well as their overt behaviors (6). This suggested shift could include mental health diagnostics that take into account the phenomenological features of dysfunction and attempt to categorize them in terms of the type of self-reported personal-social impairment experienced by the patient.
Cross-cultural applicability of the DSM system is also problematic, as the operational criteria used by the DSM to classify personality disorders are derived from Eurocentric definitions and perceptions of personality and mental illness (7). Significant dissimilarities have been found between personality traits of persons from different countries, with cultures becoming less similar as geographic or historical separation increases (8). This suggests that certain personality traits of Western cultures are not appropriate in or transferable to Eastern cultures (8), underscoring the difficulty in using "one-size-fits-all" criteria to classify personality. The recognition that personality dimensions and traits may differ from one culture to another has made the reformulation of the classification of personality disorders a priority for DSM-V (9).
The work of Walton et al. (1970) (10) challenged the conceptualization of personality disorder as being separate from mental illness by acknowledging the repeated co-occurrence and association between psychiatric illnesses and personality deviations, and the influence of personality deviations on prognosis. A pilot study using Walton's perspective was carried out from 1974 to 1980 in Jamaicans diagnosed with DSM-III personality disorder. The phenomenological features of personality disorder diagnoses were disaggregated and then re-aggregated through statistical analysis. The results indicated three distinct clusters, forming a clinical triad of abnormal thoughts, feelings, and actions manifested as 1) power management and authority problems; 2) dependency issues (psychological and physiological); and 3) psychosexual problems (11). The current study expands on this pilot study using a larger cohort of patients seen over a period of 33 years.
The objective of the current study was to characterize and assess the factor structure of phenomenological features of the DSM-IV diagnosis of personality disorder in Jamaican patients and determine 1) any similarities with those of the personality disorder features and categories outlined in the DSM-IV; 2) associations with disorder severity, as conceptualized in Walton's research; and/or 3) significant relationships between these variables to help inform the current debate on the relevance of established personality disorder diagnostics.
MATERIALS AND METHODS
This research was a case-control study. Cases were obtained from the naturalistic clinical setting of a private psychiatric practice in Kingston, Jamaica.
The study sample group included all patients seen by one private psychiatric practice from 1974 to 2007 and diagnosed as having an Axis II personality disorder based on 1) the criteria of the DSM-III, the revised DSM-III (DSM-IIIR), and the DSM-IV classification systems; 2) subjective phenomena identified by the patient; and 3) objective phenomena observed by external correspondents and the therapist. A control group composed of patients with Axis I clinical disorders who had not been diagnosed with a personality disorder was drawn from a database from the same private psychiatric practice. The study sample and control groups were matched exactly on gender, and closely on age, as well as socioeconomic variables, based on the United Kingdom's Registrar General's Social Class (RGSC) five-class scheme according to occupation (12).
The lead author conducted clinical examinations on all patients, obtaining and recording the demographic, clinical, phenomenological, and socioeconomic information, and determining the DSM diagnoses, which were supported by handwritten, verbatim case notes and based on the psychobiological model of Adolf Meyer and the Edinburgh tradition of David Henderson. Patient diagnoses were originally derived from DSMIII and DSM-III-R specifications as well as DSM-IV specifications (13) but were all converted to DSM-IV diagnostic stipulations for the purposes of this study.
Case notes included information on the referral source, history of presenting complaint, systematic inquiry, mental status examination, phenomenology (in detail), diagnosis, and treatment plan. Approximately 30 of the 351 cases were discussed with a consultant clinical psychologist, trained in the United States but practicing in Jamaica (14), who assisted in the diagnostic formulation of the cases using the criteria of the DSM - III. General diagnostic criteria were used to confirm the presence of personality disorder, and the 11 categories specified in the DSM-III were used to specify type of personality disorder.
Version 17.0 of the Statistical Package for the Social Sciences (SPSS) (Chicago, IL, USA) was used for all statistical analyses. Five Axis I and 10 Axis II patient diagnoses were identified for the study population. The symptoms of these disorders were disaggregated and recorded as individual phenomenological markers of dysfunction, with a total of 38 phenomenological variables recorded as present or absent.
All phenomenological variables were entered into an exploratory principal component factor analysis to extract the factor solution using a variance maximizing (varimax) rotation. Item loadings of 0.40 and higher were considered significant. Items were summed and re-coded to create the respective factors, and severity was scored from 0 to 9 (with zero signifying lack of presence). Correlations were determined to assess the direction and strength of the relationship between the factors.
Chi-squared analysis was used to assess the association between phenomenological markers and the Axis I and Axis II diagnoses, and the differences between patients diagnosed with Axis I and Axis II disorders in relation to the factors identified. Multiple regression analyses were carried out to determine if the identified factors were able to predict the type of Axis II cluster based on patients' diagnoses. Independent t-tests and one-way analysis of variance (ANOVA) were used to measure mean differences between variables.
The sample's demographic characteristics are reported in Table 1. Of the 351 individuals in the study sample group, 166 (47.3%) were male and 185 (53.7%) were female; 50 (14.2%) were white and 301 (85.8%) were black; 293 (83.5%) were born and raised in Jamaica; and 202 (57.6%) were from socioeconomic classes I and II. Mean age was 33.92 (standard deviation 10.236). In the study sample group ("Axis II"), significantly more individuals were diagnosed with major depression (38.5%) versus other Axis I disorders (psychosis, 6%; substance-abuse disorder, 17.9%; and anxiety disorder, 19.1% [χ2 = 166.212, P < 0.001]). In the control group ("Axis I"), just under half of those previously diagnosed with major depression also qualified for an Axis II (personality disorder) diagnosis (44.7%; n = 135 [χ2 = 5.951, P < 0.001]). Significantly more control group ("Axis I") patients were found to have symptoms related to Axis I (clinical) disorders versus study sample group ("Axis II") patients, most of whom presented prototypical symptoms of those diagnoses (see Table 2). However, after being disaggregated, most individual symptoms of depression were almost equally distributed among patients from both groups.
Exploratory factor analysis of phenomenological variables
The Cronbach's alpha value was 0.71 for all phenomenological variables in the study, indicating a sample size adequate for obtaining stable and reliable results. All phenomenological variables identified from psychiatric assessment were entered into a principal component factor analysis. A total of nine factors were identified, with a clear break in the scree plot at the sixth principal component, indicating a six-factor model of disorder. The varimax-rotated factors, which cumulatively accounted for 47.2% of the variance in the 38-item set, were retained. Eigenvalues for the six factors ranged from 1.26 to 5.87, with each factor accounting for 3.5% to 16.3% of the variance. For validation purposes, items that loaded on factors 7 to 9 were removed, the analysis repeated, and another sixfactor model identified. The sixth component of this new model was altered and the resulting six-factor solution improved, accounting for 52.7% of the total variance. Eigenvalues for these six factors ranged from 1.04 to 5.58, with each factor accounting for 3.4% to 18.6% of the variance. The pattern of factors in the second factor analysis (after rotation) is shown in Table 3.
The clinical interpretation of the sixfactor solution was assessed by the current study's research team and the following factor titles were assigned: psychosis (Factor I); conflict management problems (Factor II); control issues (Factor III); psychosexual problems (Factor IV); major depression (Factor V); and physiological dependency (Factor VI). The items comprised by each factor were summed to create continuous variables that were named as outlined above.
Because there was some overlap in the variables comprised by Factors II and III, and all of the comprised variables were conceptualized as being representative of problems with interpersonal power management, the items on these two factors were clustered and summed to form one continuous variable ("power management problems"). The internal consistency of the variables comprised by Factors I, II, and III, as measured by Cronbach's alpha, was good (α = 0.77), ranging from 0.724 to 0.837 for individual items.
A moderate relationship was found between power management problems and psychosexual problems (r = 0.460, P < 0.001), and a weak relationship was found between power management problems and physiological dependency (r = 0.202, P < 0.001), as well as psychosexual problems and physiological dependency (r = 0.204, P < 0.001).
Associations between factors and personality disorder
Significant relationships were iden - tified between a personality disorder diagnosis and power management problems (χ2 = 172.847, P < 0.001); physiological dependency (χ2 = 82.347, P < 0.001); and psychosis (χ2 = 22.225, P < 0.001). Among those diagnosed with an Axis II personality disorder, 79.7% reported power management problems, 88.0% had physiological dependency, and 88% had some feature of psychosis (versus 30.2%, 57.5%, and 97.2% respectively for those diagnosed with an Axis I clinical disorder). The factors of major depression and psychosexual problems were not significant. The proportion of patients with symptoms within the major depression factor was about equal for persons diagnosed with an Axis II personality disorder versus those diagnosed with an Axis I clinical disorder (82.9% and 83.5% respectively).
Predictive ability of factors for DSM-IV-TR cluster diagnosis
A step-wise multiple regression equation was constructed to assess the predictive factors associated with the three types of DSM-IV-TR Axis II clusters. The anal ysis revealed a statistically significant regression model for type-of-cluster diagnosis (F(6 329) = 15.656, P < 0.001). Power management problems, psychosexual problems, psychosis, and major depression together explained only 19% of the variance in type-of-cluster diagnosis. Analysis of the β coefficients of the sig nificant regression functions showed psychosis was the best predictor of type of cluster (β = -0.295, t(1 331) = -5.385, P < 0.001), followed by power management problems (β = -0.294, t(1 331) = -5.732, P < 0.001); psychosexual problems (β = 0.234, t(1 331) = 4.600, P < 0.001); and major depression (β = 0.135, t(1 331) = 2.359, P < 0.001). Due to the distribution of features of psychosexual problems and major depression within the sample, these factors are likely to be positively related to cluster diagnosis. The results indicate a weak correlation between the factors formed from the phenomenological features of personality disorder and the DSM-IV-TR Axis II clusters.
Comparison of factor scores by diagnosis
The control group had a higher mean score for the psychosis measure (M= 2.8, standard deviation [SD] = 1.93) than patients who were originally diagnosed with a personality disorder (M = 2.0, SD = 1.30, t(700) = 6.72, P < 0.001). On the other hand, patients in the study sample had significantly higher mean scores for four of the other measures: conflict management problems (M = 2.2, SD = 1.74, t(698) = -16.69, P < 0.001); control issues (M = 1.7, SD = 1.59, t(699) = -17.77, P < 0.001); psychosexual problems (M = 3.9, SD = 0.94, t(699) = -24.18, P < 0.001); and physiological dependency (M= 0.9, SD = 0.33, t(700) = -9.65, P < 0.001). Patients in the study sample also had higher mean scores for the combined "power management problems" measure (M = 2.9, SD = 2.35, t(698) = -17.26, P < 0.001). There were no significant differences between the two groups for the major depression factor.
Severity of disorder
One-way ANOVA was used to assess differences in power management problems, psychosexual problems, and physiological dependency for the three levels of disorder severity (mild, moderate, and severe). Because the assumption of homogeneity of variance was violated, the Brown-Forsythe F-ratio is also reported. Results indicated the three severity groups differed significantly for power management problems (F(3 314.92) = 92.514, P < 0.001); psychosexual problems (F(3 398.06) = 198.044, P < 0.001); and physiological dependency (F(3 464.08) = 53.193, P < 0.001). Post hoc Tukey tests indicated the severe group (n = 56; mean = 3.82, SD = 1.99) had a significantly higher mean score for power management problems versus both the moderate group (n = 184; mean = 3.09, SD = 2.50, P = 0.030) and the mild group (n = 109; mean = 2.00, SD = 1.96, P < 0.000). For physiological dependence, both the severe group (n = 57; mean = 0.96, SD = 0.19) and the moderate group (n = 185; mean = 0.91, SD = 0.28) had a significantly higher mean score than the mild group (n = 109; mean = 0.78, SD = 0.42, P = 0.034, and P = 0.040 respectively). There were no significant differences between the moderate and severe groups on this measure. For psychosexual problems, there were no significant mean differences between the three severity groups (see Table 4).
The proposed re-conceptualization of personality disorder in the upcoming DSM-V has faced major scrutiny due to its confusing, inconsistent, and incoherent criteria (5). The purpose of the current case-control study of a Jamaican sample was to determine whether the individual phenomenological features of personality disorder would cluster into patterns similar to those for traditional personality disorder features and categories.
When the phenomenological features of personality disorder are disaggregated and analyzed, the resulting clusters do not resemble the conventional categories of personality disorder as specified by the DSM-IV-TR. Instead, they cluster into three distinct categories of factors that seem to represent problems of psychoemotional nature, indicating a singular, completely separate concept, stemming from problems with impulse control and authority and conflict management, to replace the current iteration of personality disorder classification.
The components of the factors representing power management problems, psychosexual problems, and physio - logical dependency are features of dysfunction that are variable and not representative of personality traits typically assessed by other [traditional] measures of personality disorder. The authors of the current study theorize that the underlying basis of these factors may be a neurobiological dysfunction, the presentation of which is colored by the personality traits that an individual possesses. In this way the dysfunction may be similar to an Axis I disorder. It has been previously suggested that features of personality disorder are no different from the symptomology of mental disorders as both are "caused by biological and psychosocial factors and, like those of other disorders, they wax and wane over time" (3). Schneider's (1923) (15) distinction between abnormal personality (as an extreme of normalcy) and personality disorder (that causes harm/suffering to self and society) predates and supports this perspective.
The authors of the current study suggest a conceptual shift be made, reconfiguring the classification systems by remerging "personality disorder" with Axis I disorders as suggested by Livesley et al. (3). Axis II would serve as the location for listing important personality indicators that could help clinicians understand how the dysfunction is manifested and how various personality traits may be at work in ways both beneficial and harmful for an individual's functioning and treatment outcome. Axis II classification of the personality traits of patients (with or without a personality disorder diagnosis) would help clinicians select the most appropriate intervention for each case (9).
The fact that power management problems, psychosexual problems, psychosis, and major depression together explained only 19% of the variance in type-of-cluster diagnosis indicated a weak correlation between the factor structure formed from the phenomenological features of personality disorder and traditional (DSM-IV-TR) cluster diagnoses in this cohort of patients in Jamaica. Together this suggests that the disaggregated phenomenology of patients with personality disorder do not match the conventional DSM diagnostic categories.
Depression and personality disorder
The association between depression and personality disorder has long been established: a lengthy history of depression can predispose individuals to the condition, and stressful life events (common among those with personality disorders) can lead to depressive episodes (16). Several patients in the study sample had either undergone treatment for depression (38.5%) in which features of personality disorder were revealed, or reported symptoms of depression (30.4%) over the course of their treatment for personality disorder. Furthermore, of all the co-morbid Axis I diagnoses, persons diagnosed with personality disorder were significantly more likely to also have a didiagnosis of major depression than any of the other Axis I disorders that were diagnosed. Although individuals with features of personality disorder were not significantly more likely than those without these features to be depressed, the high rate of depression (82.9%) found in this population suggests that major depression, regardless of the chronology of diagnosis, is likely to be associated with having problems with impulse control and authority and conflict management. If this is the case, treatment of this underlying pathology will resolve the symptoms of the depression. Further research is required to explore the possible relationship between these two conditions.
There are several challenges inherent in measuring personality disorder severity based on current diagnostic guidelines. Both the DSM and ICD systems rate personality disorder severity based on the patient's description of his/her condition. The use of nonstandard nomenclature in the diagnostic process has significantly impaired psychiatrists' ability to adequately ascribe severity, and has been criticized for limitations in its logical basis and clinical application (10). Like Walton et al. (10), the authors of the current study suggest that "social disruption" and "social insight" (as self-reported by patients) be used as indicators of severity, and a graded rating of "mild," "moderate," or "severe" be based on the level of dysfunction. Concurrence of the current findings with Walton et al.'s premise supports this proposition.
In a recent article in New York Times Magazine, author Ethan Watters (17) states that the United States has "for many years been busily engaged in a grand project of Americanizing the world's understanding of mental health and illness," and bemoans the world's steady adoption of European and American values of mental illness. Elsewhere, the editor of the British Journal of Psychiatry acknowledged that his journal was "bound to see the panorama of psychiatry through British spectacles" (18). The present study attempts to address this problem by investigating personality disorder in another culture, using the case-control method. In their research, the authors grapple with the difficulty of applying European and American concepts to classify personality disorder in a Jamaican sample. Based on their results, the authors propose a complete reframing of personality disorder conceptualization.
While many of the phenomenological features of personality disorder identified in the Jamaican study sample are similar to the dimensions postulated by Widiger et al. (19), the clustering pattern is entirely different, forming a unique, singular construct. The authors propose this new phenomenon, which appears to be associated with problems with impulse control and authority and conflict management, as a replacement for traditional DSM criteria in the reclassification of personality dysfunction, and suggest the term Shakatani as a possible name for the condition. The concept of Shakatani- derived from the Swahili words shaka ("problem") and tani ("power")-stems from an early Jamaican anthropological study by Kerr (20) and sociological work by Stone (21, 22) that chronicle the tensions caused by economic oppression, racism, biased methods of education, and the economic and psychological insecurity caused by centuries of colonial domination by the British.
This study had several limitations. First, because this was a naturalistic study carried out by a single psychiatrist over a period spanning nearly four decades, it is impossible to verify missing or ambiguous data, or the accuracy of diagnoses. Second, the reliance on one main assessor (who made approximately 96% of the patient diagnoses and assessments of phenomenological features) may have resulted in biased data recording, which would have affected the way in which the data clustered in the factor analysis. The use of a second rater or a standardized rating scale for comparative assessment of the main interviewer's assessments would have increased the diagnostic accuracy and robustness of statistical findings. Third, the use of a combination of phenomenological variables and DSM-IV personality disorder variables may have contributed to the characteristic differences of the personality disorder factors identified by the study (versus those for traditional diagnostics) and reduced their comparability versus standard DSM-IVTR clusters and types, thereby weakening the strength of the conclusions about their relationship. Finally, the fact that the patients were culled from a private psychiatric practice skews the sample in the direction of patients who could afford assessment and treatment.
The findings of this study challenge the validity of the nosological entity of Axis II personality disorder. The authors propose a novel Axis I unitary concept of problems with impulse control and authority and conflict management as its replacement, and suggest the term Shakatani as a possible name for the condition. The authors of the current study suggest that 1) individual and cultural differences in the manifestation of personality disorder may be attributed to cultural differences in how various personality traits are perceived, and 2) their new conceptualization of this dysfunction as a completely different, singular phenomenon may help address current challenges in re-defining personality diagnostic disorder diagnosis, applying the diagnosis cross-culturally, and measuring disorder severity using traditional criteria.
Acknowledgments. This project was funded by the Culture, Health, Arts, Sports and Education Fund (CHASE) (Kingston, Jamaica).
1. Livesley WJ, Schroeder ML, Jackson DN, Jang KL. Categorical distinctions in the study of personality disorders: implications for classification. J Abnorm Psychol. 1994;103(1):6-17.
2. Livesley WJ. Suggestions for a framework for an empirically based classification of personality disorder. Can J Psychiatry. 1998;43(2):137-47.
3. Livesley WJ, Jang KL, Vernon PA. Phenotypic and genetic structure of traits delineating personality disorder. Arch Gen Psychiatry. 1998;55(10):941-8.
4. Skodol AE, Oldham JM, Bender DS, Dyck IR, Stout RL, Morey LC, et al. Dimensional representations of DSM-IV personality disorders: relationships to functional impairment. Am J Psychiatry. 2005;162(10):1919-25.
5. Livesley WJ. Confusion and incoherence in the classification of personality disorder: commentary on the preliminary proposals for DSM-5. Psychol Inj Law. 2010;3(4):304-13.
6. Schedler J, Weston D. Refining personality disorder diagnosis: integrating science and practice. Am J Psychiatry. 2004;161(8):1350-65.
7. Rivas LA. Controversial issues in the diagnosis of narcissistic personality disorder: a review of the literature. J Ment Health Couns. 2001;23(1):22-35.
8. Allik J. Personality dimensions across cultures. J Pers Disord. 2005;19(3):212-32.
9. Krueger RF, Skodol AE, Livesley WJ, Shrout PE, Huang Y. Synthesizing dimensional and categorical approaches to personality disorders: refining the research agenda for DSM-V Axis II. Int J Methods Psychiatr Res. 2007;16 Suppl 1:S65-73.
10. Walton HJ, Foulds GA, Littman SK, Presly AS. Abnormal personality. Br J Psychiatry. 1970;116(534):497-510.
11. Hickling FW, Martin J, Harrisingh-Dewar A. Redefining personality disorder in Jamaica. In: Hickling FW, Matthies BK, Morgan K, Gibson RC, editors. Perspectives in Caribbean psychology. Kingston (Jamaica): Caribbean Institute of Mental Health and Substance Abuse, University of the West Indies; 2008. Pp. 263-88.
12. Rose D. Official social classifications in the UK [Internet]. Soc Res Update. 1995 July; 9:1-7. Available from: http://sru.soc.surrey.ac.uk/SRU9.html Accessed 9 December 2009.
13. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington: American Psychiatric Association; 2000.
14. Doorbar R. Reflections of a psychologist in Jamaica. In: Hickling FW, Matthies BK, Morgan K, Gibson RC, editors. Perspectives in Caribbean psychology. Kingston (Jamaica): Caribbean Institute of Mental Health and Substance Abuse, University of the West Indies; 2008. Pp. 516-36.
15. Schneider K. Die psychopathischen persönlichkeiten. Berlin: Springer; 1923.
16. Huang Y, Kotov R, de Girolamo G, Preti A, Angermeyer M, Benjet C, et al. DSM-IV personality disorders in the WHO World Mental Health Surveys. Br J Psychiatry. 2009;195(1):45-53.
17. Watters E. The Americanization of mental illness [Internet]. New York Times Magazine. 2010 Jan 8; [about 6 screens]. Available from: http://www.nytimes.com/2010/01/10/magazine/10psyche-t.html?fta=y Accessed 25 January 2010.
18. Tyrer P. From the editor's desk. British made foreign journal. Br J Psychiatry. 2010;196(1):86.
19. Widiger TA, Livesley WJ, Clark LA. An integrative dimensional classification of personality disorder. Psychol Assess. 2009;21(3):243-55.
20. Kerr M. Personality and conflict in Jamaica. London: Collins/Sangster's; 1963.
21. Stone C. The anatomy of aggression. The Daily Gleaner. 1992 Oct 21. Kingston (Jamaica): Gleaner Company Ltd.; 1992.
22. Stone C. Values, norms and personality development in Jamaica. Paper presented at the National Consultation on Values and Attitudes, University of the West Indies, Kingston, Jamaica, March 23, 1992. Available from: http://gtuwi.tripod.com/stonearticle.htm
Manuscript received on 26 October 2010.
Revised version accepted for publication on 13 May 2011.
▲ Send correspondence to: Frederick W. Hick ling, email@example.com
1 DSM refers to the Diagnostic and Statistical Manual of Mental Disorders published by the American Psychiatric Association and ICD refers to the International Statistical Classification of Diseases and Related Health Problems produced by the World Health Organization