Suicide in borderline personality disorder: A meta-analysis

See discussions, stats, and author profiles for this publication at:

Suicide in borderline personality disorder: A meta-analysis

Article in Nordic Journal of Psychiatry ? February 2005

DOI: 10.1080/08039480500320025 ? Source: PubMed

CITATIONS

154

4 authors, including: Maurizio Pompili Sapienza University of Rome 695 PUBLICATIONS 10,475 CITATIONS

SEE PROFILE

Roberto Tatarelli Sapienza University of Rome 368 PUBLICATIONS 6,588 CITATIONS

SEE PROFILE

READS

6,091

Paolo Girardi Sapienza University of Rome 576 PUBLICATIONS 11,623 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects: Workplace Suicide Prevention, Mental Health and Resilience View project Adult ADHD: comorbidity, suicide risk and faking View project

All content following this page was uploaded by Maurizio Pompili on 31 May 2014.

The user has requested enhancement of the downloaded file.

REVIEW ARTICLE

Suicide in borderline personality disorder: A meta-analysis

MAURIZIO POMPILI, PAOLO GIRARDI, AMEDEO RUBERTO, ROBERTO TATARELLI

Pompili M, Girardi P, Ruberto A, Tatarelli R. Suicide in borderline personality disorder: A meta-analysis. Nord J Psychiatry 2005;59:319 ?/324. Oslo. ISSN 0803-9488.

Suicide is the major cause of death among patients with borderline personality disorder; however, the literature on completed suicides in such disorder is inconclusive, as suicide rates vary greatly among cohorts of patients. We searched MedLine, Excerpta Medica and PsycLit from 1980 to 2005 to identify papers dealing with suicide in borderline personality disorder. We also searched the World Health Statistics Annual to ascertain the suicide rate in the age groups for specific years and country. We selected eight studies comprising 1179 patients with a diagnosis of borderline personality disorder. Of these patients, 94 committed suicide. Results obtained for each study were processed together to calculate the mean figure for each year of suicides for 100,000 individuals suffering from borderline personality disorder. Our metaanalysis shows that suicide among patients with borderline personality disorder is more frequent when compared with the general population. All study analyses reported that patients with borderline personality disorder committed suicide more often than their counterparts in the general population. Suicide seems more alarming in the first phases of follow-up than during chronic phases of illness. Borderline, Meta-analysis, Suicide.

Maurizio Pompili, Department of Psychiatry, Sant'Andrea Hospital, Via di Grottarossa, 1035-1039, 00189 Roma, Italy, E-mail: maurizio.pompili@uniroma1.it or mpompili@mclean. harvard.edu; Accepted 25 May 2005.

The incidence of completed suicide in borderline personality disorder (BPD) has been unknown until recently. In two long-term follow-up studies of borderline patients treated in residential settings, McGlashan (1) and Stone et al. (2) found that 3% and 9%, respectively, of borderline go on to complete suicide.

Patients with BPD represent 9?/33% of all suicides (3, 4). Mehlum et al. (5) pointed out that a lack of control of high intensity affects such as depression, anxiety or anger may increase the tendency towards suicidal behavior. Thus the notion of BPD as ``the suicidal personality disorder'' may be justified. On the other hand, the self-destructive tendency in subjects with BPD may not primarily be due to the personality disorder (PD) syndrome itself, but rather be caused by some secondary or coexisting Axis I mental disorder, such as a mood disorder, anxiety disorder or substance abuse. Or it may be viewed as a result of a complex interplay between state and trait dimension. Comorbidity with major depression is highly prevalent in BPD (6); however, the effect of this comorbidity on suicidal

behavior is unclear. Comorbidity with major depressive episode has been associated with an increased mortality rate in some (1, 7) but not all (8 ?/10) studies of suicidal behavior in BPD. Comorbidity with major depressive episode has also been associated with an increase in the seriousness and frequency of suicide attempts among inpatients with BPD (6). Other studies have found that comorbidity with major depressive episode is not predictive of a history of suicide attempts with BPD (11) and have found no relationship between comorbid major episode and measures of suicidal intent, lethality or risk (12). Peterson & Bongar (13) studied patients with chronic suicidality that made four or more visits in a year to a psychiatric emergency room; most often, these patients met criteria for BPD. Paris et al. (8) found that the strongest clinical predictor of completed suicide was previous attempts, although very often these gestures are manipulative. These authors also observed that higher education was strongly associated with completed suicide, which was explained by the crushed expectations of an educated person with severe psychopathology.

# 2005 Taylor & Francis

DOI: 10.1080/08039480500320025

M POMPILI ET AL.

Crumley (14) has shown a high incidence of BPD in the adolescents and young adults aged 15 ?/24 years who engage in suicidal behavior. Paris & Zweig-Frank (15) indicated that this diagnosis significantly increases the risk of eventual suicide. Those at higher risk appeared to be young, ranging from adolescence into the third decade (2, 16). The high rates of suicidal behavior in patients with BPD are reflected by the inclusion of recurrent suicidal behavior, gestures, threats or self-mutilating behavior as diagnostic criteria in the DSM IV. Also, among patients with BPD, impulsivity, assessed as a diagnostic criterion, is associated with the number of suicide attempts independent of comorbid depression or substance use disorder. Soloff et al. (17) observed that hopelessness predicted the lifetime number of suicide attempts and the degree of lethal intent. Hopelessness may contribute to the seriousness of suicidal behavior in BPD, especially in patients with comorbid depression, by increasing the number of attempts, the level of subjective intent and the degree of objective planning. A relation has been found between increased suicidal behavior and comorbidity of substance abuse disorder with BPD (3, 11, 18). Links et al. (18) examined the prognostic significance of comorbid substance abuse in patients with BPD. The patients were followed prospectively over a 7-year period. These researchers found that patients with comorbid substance abuse and BPD perceived themselves to be at significantly more risk for suicide than did the comparison groups of patients having BPD without comorbidity, patients having substance abuse without BPD and patients having borderline traits only.

Harris & Barraclough's (19) comprehensive metaanalysis on suicide as an outcome for mental disorders included five studies dealing with suicide in personality disorders. The five cohorts were mainly heterogeneous spanning from the mid-1960s to the mid-1980s. These authors combining the studies found a suicide risk seven times the expected value.

Data concerning suicide among subjects with BPD may be difficult to obtain because suicide statistics are usually reported among many other variables. Also, cohorts including only borderline patients and followed up for a certain period of time are few; this contrasts with the ever increasing availability of data regarding suicidality among these patients. To our knowledge, a meta-analytic investigation of suicide among patients with a diagnosis of BPDs has still to be performed in the international literature.

Materials and Methods

We conducted careful MedLine, Excerpta Medica and PsycLit searches to identify papers in English during the period 1980 ?/2005. The following search

terms were used: ``suicid*'' (which comprises suicide, suicidal, suicidality and other suicide-related terms), ``borderline'' and ``personality disorder''. In addition, each category was cross-referenced with the others using the MeSH method (medical subject headings). Study selection allowed the inclusion of only articles published in peer-reviewed English-language journals. Suicide of borderline patients is often reported among cohorts of patients that received a diagnosis of personality disorder. We avoided a systematic review of the literature prior to 1980 because of diagnostic criteria heterogeneity. Individuals analyzed in this study received a diagnosis of BPD according to various diagnostic criteria, mostly according to DSM-III or IV criteria and were followed up or were studied retrospectively. We excluded any study that reported suicide in borderline patients as part of the analysis of suicidality among individuals with personality disorders or with any other psychiatric disorders. We also excluded studies that mentioned data about suicide but were not clear about follow-up times, method of statistical analysis, diagnostic criteria and the number of patients analyzed.

The World Health Organization publishes the World Health Statistics Annual , a bulletin with the statistics of all causes of death ismost countries worldwide. Each cause of death is divided into age groups, providing for each group rates and total number of male and female deaths. Statistics in this bulletin are generally per 100,000 individuals.

For each study selected for our analysis, we searched the Bulletin of the World Health Organization (WHO), referring to the year of publication and the country where the study was performed. We identified suicide statistics for a specific year and country, and used only those applicable to the age group indicated in the study. A comparative analysis was performed in order to ascertain whether suicide in BPD may be considered a more frequent phenomenon in comparison to suicide among the general population.

Results

We selected eight studies comprising 1179 patients. In this group of patients, 94 had committed suicide. Follow-up observation ranged from 3 to 27 years. Table 1 provides information about the studies that we used for our analysis (1, 2, 7, 10, 15, 19?/21). Each study has been positioned in the table in chronological order. Table 1 also illustrates the results of our comparative analysis. Fig. 1 shows distribution of suicides among patients with BPD compared with the general population of the same age group in accordance with the country and year of publication of a specific study.

320

NORD J PSYCHIATRY?VOL 59 ?NO 5 ?2005

NORD J PSYCHIATRY ?VOL 59?NO 5?2005

Table 1. Results of the meta-analysis.

No. of studies 1 2

3

4

5

6

7

8

Author, year of publication and diagnostic criteria (reference)

Pope et al., 1983 (USA) ?/ DSM-III (20) Akiskal et al., 1985 (USA) ?/ DSM-III; Gunderson-Singer (21) McGlashan, 1986 (USA) ?/ DSM-III; Gunderson criteria (1) Stone et al., 1987 (USA) ?/ DSM-III; Kernberg criteria (2) Stone, 1989 (USA) ?/ DSM-III; Kernberg criteria (7) Modestin and Villinger, 1989 (Switzerland); DSM-III (22) Kjelsberg et al., 1991 (Norway), Gunderson criteria (10) Paris and Zweig-Frank, 2001 (Canada), Gunderson criteria (15) Total

Size of cohort

33 100 81 251 299 26 289 100 1179

Follow-up (years)

Expected suicides Expected suicide in a

in a year in 100,000 year in a population

individuals if they

of 100,000

all suffered from individuals (general

Suicides

BPDb

population)

Expected suicide in a year in a population of 100,000 males

Expected suicide in a year in a population of

100,000 females

Age groups researched in the WHO Bulletin in accordance with the

mean age of the cohorts

3.5

2

3

4

1732 1333

10.6

25.20

7.40

16.1a

25.20

6.96

25 ?/34 25 ?/34

15

2

165

16.0a

25.41

6.61

25 ?/34

16

19

473

11.2

20.50

4.40

15 ?/24

16.5

27

4.6

2

547 1672

12.0 25.2a

21.30 36.60

4.30 13.70

15 ?/24 35 ?/44

20

21

363

24.9

29.80

11.30

25 ?/34

27

17

630

14.1

27.30

7.00

35 ?/44

105.6

94

6915

130.0

SUICIDE IN BORDERLINE PERSONALITY DISORDER

aNot adjusted for gender. bThe number of suicides was calculated as suicides)/100,000/number of patients in the study)/follow up. Mean of suicides in a population suffering from BPD (projection of a population of 100,000 individuals)0/898; standard deviation for a population suffering from BPD (projection of a population of 100,000 individuals) 0/660; mean of suicides expected in the general population (referring to 100,000 individuals) 0/16.6; standard deviation for the general population 0/6.

321

M POMPILI ET AL.

10000

Number of suicides (in a year in a population of 100,000 individuals)

1000

100

10

1

1

2

3

4

5

6

7

8

Number of study

Patients with BPD

General population

Trend

Trend

Fig. 1. Distribution of suicides among patients with BPD compared with the general population of the same age group in accordance with the country and year of publication of a specific study. Dashed line, general population; straight line, patients with BPD.

The results obtained for each study were processed together to calculate the mean figure for each year of suicides per 100,000 individuals suffering from BPD. The mean of suicides expected in a population suffering from BPD was calculated as follows: a[suicides (study 1)'/suicides (study 2)'/ . . . suicides (study 8)]/8. A comparison between mean figure of suicide in patients with BPD and in the general population has statistical relevance and allowed us to draw statistically significant results.

The standard deviation of a population suffering from BPD has been calculated as follows:

s 0 sP ffiffiffiffiffiffiNi0ffiffiffiffi1ffi(ffiffixffiffiffiiffiffi(ffiffiffiffiffiffixffi?ffi)ffiffi2ffi N

Discussion

Suicide among patients with BPD is a major issue and in some studies, the number of suicides is disturbingly high. This is not a new finding, because BPD, by definition (DSM-IV), is characterized by ``recurrent suicidal behavior, gestures or threats, or self-mutilating behavior''. Nevertheless, our metaanalysis showed that suicide among patients with BPD is a heterogeneous phenomenon when different cohorts are compared. Suicide was more frequent in all

cohorts analyzed compared with the general population. Even more striking is, however, the great difference among cohorts. It is difficult to recognize background elements that precipitated suicide in these cohorts in a given length of time. One possible finding emerging from our study is the identification that the number of suicides would appear to be higher in short-term followups. If this assumption were true, suicide risk would be higher during the first phases of the taking care rather than during the chronic phases of the illness.

Comorbidity of psychiatric and personality disorders is increasingly recognized as a major factor in suicide. Comorbid disorders, especially those belonging to DSM-IV Axis I, may play a central role in precipitating suicide in BPD and should always be indicated in studies of mortality of these patients.

Our study has a number of limitations. Firstly, papers included in this meta-analysis may be only a part of the literature; we only used cohorts available from studies published in medical journals, which may exclude data available in other reports. Secondly, the studies analyzed used different diagnostic criteria; these, although similar to each other, may represent a source of bias. One of the main problems of this study is the heterogeneity of cohorts included in the metaanalysis, which might be to some degree incompatible with one another. DSM IV diagnostic criteria for personality disorders are usually the most common for

322

NORD J PSYCHIATRY?VOL 59 ?NO 5 ?2005

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download