need help with research paper and research experiment
kel11
REVIEW
Implications of Long-term Outcome Research for the Management of Patients with
Borderline Personality Disorder
Joel Paris, MD
Fifteen- and 27-year follow-up studies of patients with borderline personality disorder show that most of them no longer meet full criteria for the disorder by age 40, and that even more show improvement by age 50. The mechanisms behind remission could include matura- tion, social learning, and the avoidance of conflictual intimacy. Affective instability is slower to change than impulsivity. Suicide rates in patients with this disorder are close to 10%, with most completions occurring late in the course of illness; early mortality from all causes exceeds 18%. All of these findings have clinical implications. Although treatment effects must be assessed in the context of naturalistic improvement, therapy can hasten remission. (HARVARD REV PSYCHIATRY 2002;10:315–323.)
From the Department of Psychiatry, McGill University, and the In- stitute of Community and Family Psychiatry, Sir Mortimer B. Davis– Jewish General Hospital, Montreal, Quebec, Canada.
Original manuscript received 14 March 2002; revised manuscript re- ceived 2 May 2002, accepted for publication 9 May 2002.
Reprint requests: Joel Paris, MD, Institute of Community and Family Psychiatry, Sir Mortimer B. Davis–Jewish General Hospital, 4333 Chemin de la Côte Ste. Catherine, Montréal, Québec H3T1E4, Canada.
© 2002 President and Fellows of Harvard College
315
Borderline personality disorder (BPD) has long been known to be a chronic condition.1 The purpose of this paper is to eval- uate treatment for BPD in light of its course, as shown by studies following patients for 15 years or more. If remission (i.e., reduction of symptoms over time) occurs naturally, then the effectiveness of therapy needs to measured in this context.
The literature reviewed in this paper was located by means of Medline and PsycInfo searches for articles published in English between 1990 and 2002 and containing the key- words “borderline personality disorder.” The search identi- fied 1803 publications, of which 938 concerned treatment, and an additional 146 concerned outcome. These publications were supplemented by selected reports published prior to 1990 but frequently referenced in recent literature. This re-
view will specifically consider 23 papers on treatment of BPD and 26 on its outcome that shed particular light on the links between outcome and treatment.
FIFTEEN-YEAR FOLLOW-UP STUDIES OF BPD
In a memorable phrase, Schmideberg1 described the course of BPD as “stably unstable.” Although Schmideberg’s clinical impressions are correct in the medium term, they may not apply to the long-term outcome of patients with the disorder.
The first formal follow-up studies of patients with BPD were conducted in the 1970s.2–4 When cohorts were followed for 5 years, patients changed very little. But this period was too brief to determine the outcome of such a chronic disorder.
Four 15-year follow-up studies of borderline patients were published in the 1980s (see Table 1). The Chestnut Lodge study,5 conducted by McGlashan, and the Austen Riggs study,6 led by Plakun, followed cohorts from private, psycho- analytically oriented hospitals. The Columbia study,7 con- ducted by Stone, followed patients treated on a specialized psychotherapy ward at the New York State Psychiatric Insti- tute. The Montreal study,8 led by Paris, followed a group of patients treated more briefly at an urban general hospital.
Each of these investigations had strengths and weak- nesses. The Chestnut Lodge and Columbia studies located over 80% of patients, while the Austen Riggs and Montreal studies assessed less than a third of their cohorts. Clearly, the studies that located the largest percentage of the original
patients have obvious advantages for generalizability. On the other hand, patients in the Chestnut Lodge, Columbia, and Austen Riggs samples were highly educated and affluent. Such populations do not resemble community profiles of BPD, in which most patients have low educational and socio- economic levels.9 In this respect, the sample in the Montreal study, which included subjects from a wide range of socio- economic backgrounds, was more typical. Another difference among the various samples arose from treatment histories: in the Chestnut Lodge, Columbia, and Austen Riggs studies, all patients received long-term psychotherapy, while in the Montreal study, many received only intermittent treatment.
Subjects in all four studies were followed into early middle age (about age 40). In all cohorts, the majority of subjects were female, the same proportion found in community samples.9 In two studies (Chestnut Lodge and Montreal), out- come was assessed through telephone interviews. In the Columbia investigation, most patients were interviewed by telephone, but some were assessed through information pro- vided by informants. The Austen Riggs study relied on mailed questionnaires.
All studies measured global outcome at follow-up, using either the Health-Sickness Rating Scale10 or the Global As- sessment of Functioning.11 The Montreal study also exam- ined whether patients still met formal criteria for BPD, us- ing the Diagnostic Interview for Borderlines.12
Despite differences in samples and methodology, all four studies of the 15-year outcome of BPD patients produced vir- tually identical results. Mean scores for global functioning were all in the 60s—within the broad range of normality. In all cohorts, rehospitalization was uncommon after the first few years, and by the time of follow-up, most patients were working and had a social network. The Montreal study found that only 25% of the original sample still met diagnostic cri- teria for BPD, with all subscales (dysphoria, impulsivity, dis- turbed relationships, and cognition) showing improvement over time.
Since this level of improvement had not been seen in 5- year follow-up studies, it seems likely that remission requires more time, probably a decade or more after first presentation. But even with a mean time to follow-up of 15 years, each co- hort showed a wide range of outcomes and of times to remis- sion. The Chestnut Lodge study was the only one to examine separate cohorts defined by postdischarge time. In a group followed for 20 years, a significant decrease in functional level was observed, raising the question as to whether outcome has an “inverted U” pattern over time. (Data shedding light on this issue are examined below.)
Suicide completions were the downside of the outcome story. In the Columbia7 and Montreal13 cohorts, rates were close to 9%. Similar rates of suicide completion for patients with BPD have been confirmed in other settings: 8% in Nor-
316 Paris Harvard Rev Psychiatry
November/December 2002
TABLE 1. Long-Term Studies of the Outcome of Borderline Personality Disorder
Study
Parameter Chestnut Lodge5,18 Columbia7 Austen Riggs6 Montreal8,31,32
Type of hospital Private State Private General Length of follow-up (y) 15 15 15 15/27 % of cohort located 86 91 27 32/26 Number assessed 81 206 54 100/64 Mean age (y) 47 37 40 39/51 Female:male ratio 52:46 70:30 62:38 84:16/83:17 Socioeconomic status High High High Low to high % ever married 70 F, 52; M, 29 ? 67 % with children 48 F, 25; M, 15 ? 59 Mean GAF score 64 67 67 63/63 % still meeting criteria for BPD ? ? ? 25/8 Predictors of continued dysfunction Low IQ, lengthy Low IQ, Self-harm, anger Dysthymia, problems
previous admissions, childhood abuse with mother affective instability
% suicide 3 9 ? 9/10 Mean age at suicide (y) ? 30 ? 30/37 Predictors of suicide completion ? Substance abuse, ? Previous attempts,
major depression higher level of education % death other than suicide 13 13 ? 13/18
F, female; GAF, Global Assessment of Functioning; M, male.
way14 and 10% in Toronto.15 Although the Chestnut Lodge co- hort5 had a much lower rate (only 3%), McGlashan (personal communication, 1991) believes that his sample was unrepre- sentative, given that patients were referred following treat- ment in general hospitals. A suicide rate around 10% would be similar to those found in schizophrenia16 and major mood disorders.17 Findings at 15-year follow-up also show that sui- cide generally occurs comparatively late in the course of ill- ness—that is, after age 30.7,13
Since outcome varies greatly in patients with BPD, it would be highly useful to identify predictive factors. Mc- Glashan18 reported that the two strongest correlates of posi- tive outcome were higher intelligence and shorter length of previous hospitalization, although neither accounted for a large percentage of the variance. McGlashan also found that high levels of affective instability were associated with a lower level of functioning, a finding supported in the Mon- treal study.
Because patients with BPD report having had many prob- lems during childhood,19 researchers have sought to deter- mine whether early developmental experiences are related to outcome. One study20 compared a group of women who still met criteria for BPD with a group of women who did not and found that childhood sexual abuse was more frequent in the women who remained symptomatic. In the Columbia cohort,7
“parental brutality” was linked to outcome, but this measure was not blindly assessed and accounted for only 7% of the variance. The Montreal study21 found a correlation between a chart review–derived measure of problems with mothers during childhood and lower outcome scores. But none of these relationships is strong or consistent enough to be clinically useful.
Research has also failed to identify clinically useful pre- dictors for suicide. This problem is not unique to BPD: even in very large samples of psychiatric patients, it is difficult to identify suicide predictors of practical use.22 For example, al- though the number of previous attempts has some relation- ship to completion in individuals with BPD,13,14,23 most pa- tients with multiple attempts do not complete suicide. The Columbia study7 found that substance abuse predicted com- pletion, but most of the subjects who abused substances did not complete suicide. Authors of the Montreal investigation observed that patients with higher levels of education were more likely to complete suicide.13 This finding was not sup- ported by a Norwegian study,14 which showed that BPD pa- tients at all educational levels are at risk for completed sui- cide and instead suggested a correlation between separation or loss early in life and higher rates of completed suicide (a correlation also seen in a psychological autopsy study of young males with BPD who completed suicide24). Again, none of these reports yielded predictors that account for enough of the variance to be useful clinical markers.
PROSPECTIVE STUDIES OF BPD OUTCOME
The 15-year follow-up studies of BPD all used a “follow-back” method. Prospective designs yield more-reliable baseline data, allowing outcome predictors to be identified more accu- rately. The main limitation of prospective methods is that pa- tients with BPD who agree to be followed over time may have unusual characteristics, such as higher compliance, that make them different from populations clinicians see. More- over, to reduce attrition, prospective studies of BPD have focused on patients in long-term treatment, in whom distin- guishing between treatment effects and naturalistic remis- sion is difficult.
A major prospective study of patients with BPD was con- ducted at McMaster University in Hamilton, Ontario.25 The researchers followed a cohort of 130 former inpatients, of whom 88 had a diagnosis of BPD and 42 had “borderline traits.” At 7-year follow-up, approximately two thirds of the cohort remained. Two patients had died of natural causes, and six (5%) had committed suicide. Forty-seven percent of the remaining cohort still met criteria for BPD, while 53% showed symptomatic remission. That this level of improve- ment was only about two thirds of the level observed at 15- year follow-up in the Montreal cohort8 supports the concept that BPD patients often require 10 years or more to attain relatively normal levels of functioning. The McMaster study found that severity of initial pathology was the best predic- tor of outcome, accounting for 17% of the variance. Patients with serious comorbid substance abuse (about a quarter of the sample) had a much poorer outcome.26 The main limita- tion of this otherwise informative study was that 7 years is not long enough to observe remission or establish a definitive suicide rate.
Zanarini’s group at McLean Hospital, in Belmont, Massa- chusetts,27 has been conducting a prospective study of a pre- viously admitted cohort, following 290 patients with BPD and 72 patients with other Axis II diagnoses. At 4-year follow-up, 75% of the patients with BPD were still in outpatient ther- apy, but only 36% had been rehospitalized.28 Although these findings are preliminary, the study will eventually provide data on long-term outcome.
The largest-scale current investigation of BPD outcome is the National Institute of Mental Health–funded Collabora- tive Longitudinal Personality Disorders Study, now being con- ducted in Boston, New Haven, New York, and Providence.29
The researchers have been tracking a group of patients with BPD and other Axis II disorders for over 5 years. The most striking finding thus far is that BPD has a waxing and wan- ing course: when assessed every few months, patients may or may not meet criteria, depending on what is occurring their lives.30 This study is still in its early stages but will in time provide data on the long-term outcome of BPD.
Harvard Rev Psychiatry
Volume 10, Number 6 Paris 317
A 27-YEAR OUTCOME STUDY OF BPD
The Montreal group has recently completed a 27-year follow- up of patients with BPD.31,32 Data were obtained on 81 of the 100 individuals studied after 15 years. In the intervening pe- riod, five had died from natural causes, and three had com- mitted suicide. Nine patients known to be alive refused to be evaluated. Thus, 64 subjects (12 men and 52 women) who had then reached a mean age of 51 were interviewed.
Although mean global functioning scores were unchanged, the overall rate of suicide completion increased to 10.3%. The mean age at suicide was 37.3 ± 10.3 years. Thus, completions occurred late in the course of illness, with few among patients in their early 20s, when attempts were very common.
The cohort had an unusually high rate of early death, a finding also observed in the Columbia study.7 At 27-year follow-up, a total of 18.2% of the original sample had died, ei- ther from natural causes or from suicide, a much higher rate than would be expected for a population of this age.33 The high level of mortality associated with BPD is one of the most im- portant findings of research on long-term outcome.
The most striking change between 15 and 27 years in- volved the number of subjects who still met criteria for BPD. Only 8% of the total still had a borderline diagnosis at 27 years, as opposed to 25% at the 15-year point. (In the later study, the Revised Diagnostic Interview for Borderlines,34 a less inclusive measure than the original Diagnostic Interview for Borderlines, was used for diagnosis, but the results would probably have been similar even if the original instrument had been employed.) The subscale showing the most signifi- cant improvement between the two follow-up points involved quality of relationships.
As for Axis I, only 5% of the remaining cohort had active substance abuse or met criteria for major depression at 27 years. Social adjustment was also close to normative values. The main indicator of continued problems was that 22% met criteria for dysthymia. This diagnosis indicates low-level chronic depression but can also be seen as a marker for affec- tive instability. Early-onset dysthymia is found quite often in individuals with BPD35 and is one of the most common Axis I comorbidities in the disorder.36 The continued presence of these symptoms suggests that affective lability changes more slowly over time than does impulsivity. The strongest predic- tors of 27-year outcome were levels of functioning at 15-year follow-up.32 Childhood experiences, as measured by self- report scales, had no relationship to outcome.
The findings of the Montreal study show that remission in BPD continues well into middle age. The results do not sup- port earlier concerns that outcome may follow the pattern of an “inverted U.” By the age of 50, BPD patients were func- tioning much better than they had been 12 years earlier, even though about a quarter of them had residual affective symptoms.
MECHANISMS OF REMISSION
Several mechanisms lie behind the remission process in BPD. It is well established in community populations that impul- sivity tends to decrease with age,37 a process that could re- flect biological maturation. In this context, BPD patients seem to improve in much the same way as those with other impulsive disorders, including alcohol abuse,38 antisocial per- sonality disorder,39 and bulimia nervosa.40
A second mechanism could involve social learning.41 Al- though individuals with BPD are slow to learn from experi- ence, they can still increase their skills over time. Improve- ment could also come from finding supportive relationships and choosing partners who exhibit less pathology.
Long-term outcome data on BPD suggest another mecha- nism: avoidance of intimacy. Many BPD patients drop out early from school, have difficulty in establishing a career, ex- perience periods of unemployment, and have problems find- ing stable friendships. Yet those who improve eventually over- come most of these difficulties. As we have seen, most subjects in the follow-up cohorts obtained employment and estab- lished social networks.
Intimacy, however, is more difficult to achieve. Barden- stein and McGlashan42 observed gender differences in out- come among BPD patients in the Chestnut Lodge cohort. Women with BPD had a somewhat poorer long-term outcome than did men. Whereas men benefited from strong commit- ments to work, intimacy for borderline women was problem- atic; many women became increasingly symptomatic when their marriages broke down.
The success of marriage in BPD patients may depend in part on the personality characteristics of their partners.43 Al- though no systematic studies have been conducted on choice of partners in individuals with BPD, some women with the disorder seem to be attracted to narcissistic men, who ini- tially find them attractive and later become abusive or aban- doning. Marriage with a spouse who assumes a care-taking role might be more stable, although clinical experience sug- gests that this dyad can also run into trouble due to insuffi- cient limit-setting on impulsive behavior.
As shown in Table 1, in the Columbia cohort7 only 52% of the women and 29% of the men had ever been married; only 25% of the women and 15% of the men had ever had children. Among the patients who had married, the divorce rate was 33%. Although this figure is not excessive compared to the national average, when marriages broke down, only 10% remarried—a rate much lower than the national average.44
In the Montreal study, the marriage rate was 67% (similar to the 70% observed in the Chestnut Lodge cohort), and the divorce rate was 36%. But at 27-year follow-up, only 42% were then living in a stable relationship; 41% remained childless.
We lack systematic data to confirm the hypothesis that
318 Paris Harvard Rev Psychiatry
November/December 2002
BPD patients can achieve symptomatic improvement by avoiding intimacy. Nonetheless, clinical interviews with individuals who no longer meet criteria for the disorder of- ten elicit descriptions of learning not to fall in love. Highly charged relationships create numerous problems for young patients with BPD. Over time, some learn that intimacy is dangerous for them. For these patients, being comfortably alone and finding other, less conflictual ways of establishing social support networks and contacts reduces the chance of serious difficulty.
Therapists may benefit from taking these observations into account. The success of treatment is not determined solely by the attainment of stable intimacy. On the contrary, some patients with BPD might be encouraged to avoid such involvement. Many who improve find that having less- intimate friends, belonging to a social community, or having a pet provides more stability than could have been achieved through intimacy.
Little research has been done on the effects of parenthood on women with BPD. A surprisingly large number remain childless. Parenthood also requires the management of inti- macy, and it is possible that some mothers can develop “borderline relationships” (i.e., clinging dependency) with their children. In one study45 the children of mothers with BPD were found to be highly symptomatic, and family life was observed to be very dysfunctional.46 Although systematic observations on the subject are lacking, the author’s clinical experience suggests the possibility that some women give up impulsive behaviors for the sake of their children. In such cases, they may no longer meet criteria for a diagnosis of BPD; they “graduate” to a DSM diagnosis of “personality disorder, not otherwise specified,” or their symptoms are diminished.
CHRONICITY,TREATMENT, AND REMISSION
Since BPD is chronic but improves with time, treatment re- sponse must be assessed in the context of naturalistic re- mission. Thus, when clinicians claim that patients with BPD improve after many years of treatment, one cannot know whether the positive outcomes are the result of therapy or of “waiting out” the pathology until it remits.
Nonetheless, therapy for BPD can hasten the natural pro- cess of remission. A meta-analysis of treatment studies of pa- tients with personality disorders47 found that whereas a model of the natural history of BPD shows 3.7% of patients with BPD to remit each year, active psychotherapy improves this rate sevenfold. Although this figure is encouraging, it is probably overly optimistic. The meta-analysis was drawn from a small data set that included uncontrolled or partially controlled studies of varying provenance. As noted earlier, even in patients who receive consistent therapy, remission rarely occurs within the first 5 years. Moreover, although
making comparisons among samples is impossible, the re- sults in high-treatment and low-treatment follow-up cohorts were quite similar.
To assess the long-term impact of therapy on the course of a disorder with many remissions and relapses, patients need to be followed over time. Most treatment studies last for a year or less—sufficient for major depression but not for BPD. This caution applies particularly to psychopharmacologi- cal studies, which have demonstrated short-term sympto- matic improvement in BPD (mainly in relation to impulsivity) with a variety of agents, including neuroleptics, selective serotonin-reuptake inhibitors, and mood stabilizers.48 How- ever, research has not shown whether such improvements are stable over time.
Time is also important in measuring the effects of psycho- therapy. For example, Linehan’s cohort of patients in a ran- domized controlled trial of dialectical behavior therapy con- ducted in the 1980s49 was treated for a year and followed up a year later.50 At the end of therapy, the patients who received dialectical behavior therapy were overdosing less, slashing less, and requiring less hospital care than were controls, who received “treatment as usual,” but they did not achieve full remission. Most continued to suffer from high levels of dys- phoria—symptoms that were still present at follow-up. It is unfortunate that this cohort, treated about 15 years ago, has never received long-term follow-up.
Nonetheless, the likelihood of eventual improvement should be reassuring for therapists. It has been suggested that patients with BPD should be told to expect remission over time. Clinicians clearly do take outcome into account when they provide supportive therapy that offers a “hold” for patients until they improve.
Despite the chronicity of BPD, it may not always be nec- essary to treat patients continuously for years. Many move in and out of therapy, and only about a third remain in long- term treatment, even when it is offered.51,52 We do not know whether patients who remain in therapy fare differently from those who do not. But clinicians can capitalize on trajectories of instability. Several outcome researchers,53–55 aware of the chronicity of BPD, have recommended intermittent treat- ment, as long as reentry is readily available when crises occur.
OUTCOME AND AFFECTIVE INSTABILITY
Affective instability (AI) is a central feature of BPD. Individ- uals with this disorder can be anxious in the morning, angry in the afternoon, and suicidally depressed in the evening. Linehan56 has hypothesized that emotional dysregulation (a concept essentially equivalent to AI) is associated with a rapid response to environmental stimuli and a slow recovery time. She suggests that this trait is the core feature of BPD and that it is constitutional.
Harvard Rev Psychiatry
Volume 10, Number 6 Paris 319
Often confused with depression or mania, AI has impor- tant phenomenological differences from these conditions.57,58
It is highly responsive to environmental cues and involves in- consistency of mood over time. Moreover, although mood sta- bilizers and antidepressants in BPD often lead to declines in impulsivity, they have fewer effects on affective symptoms.59
These observations suggest that AI could have a unique biology.
Although much research has been done on the biology of impulsivity, much less exists on that of affective instability. Jang and colleagues60 used behavioral genetic methods to measure genetic and environmental contributions to AI. As with most traits, about 40% of the variance was genetic, whereas another 50% was attributable to unshared environ- ment. Yet even when traits are genetically influenced, they may be modified by psychotherapy.61,62
Long-term outcome findings suggest that most patients with BPD improve more slowly on the affective than on the impulsive dimension. Although both traits are influenced by genetic factors, impulsivity may be more likely to change with age and life experience. These observations have clinical implications.
When treating patients with BPD, therapists can estab- lish a hierarchy of goals for different traits. (A similar strate- gic framework has been recommended in clinical guides to treatment by Linehan56 and Gunderson.63) Impulsivity is usually the first target, since acting out prevents treatment from addressing other goals. Once the patient is calm enough to work in therapy, treatment can address AI in a number of ways. These methods involve developing tolerance of emo- tions, decentering emotions, and modifying cognitive ap- praisals.56,64 Even if AI never entirely disappears from the picture, it can be tamed. This happens when the circum- stances that provoke such responses occur less often and when patients learn to modulate emotions before they become overwhelming.
SUICIDALITY IN BPD
Suicidality in patients with BPD can be frightening. Concern might seem to be justified by a 10% completion rate, yet 90% of BPD patients do not die by suicide. Even though most make multiple threats and attempts, exactly which individuals will eventually succeed in killing themselves cannot be predicted.
Outcome research on BPD offers a surprising degree of re- assurance about the danger of completion. Suicide usually occurs surprisingly late in the course of the illness—gener- ally after age 30. This contrasts with the alarmingly high lev- els of suicide threats and attempts seen in younger patients. Thus, suicide in BPD does not necessarily occur in the midst of a crisis. Although younger patients do commit suicide, the fact that outcome studies7,13,31 found a mean age of 30–37 years among completers supports the clinical observation
that suicide becomes most likely in a state of withdrawal and hopelessness, often after a series of failed treatments.
In light of these conclusions, the management of suicide threats and gestures in BPD can be conservative and less in- terventive. Alarmed clinicians tend to go out of their way to “prevent” suicide, but there is no evidence that therapy actu- ally prevents completion.55 Hospitalization, the most common intervention, may provide temporary relief but can also lead to regressive complications.63 Although managed care has made the option of admission less available, there is no def- inite evidence that patients with BPD are suffering as a re- sult of this trend.
Experts on suicide65 and on the treatment of BPD63,66,67
have taken the position that constant concern about suicide completion can derail therapy. Linehan,56 whose training is behavioral, emphasizes the reinforcing quality of hospital- ization or increasing the frequency of contact. Thus, inter- ventions intended to respond to suicidal behaviors can im- plicitly reward them. Hospital treatment was originally designed to manage acute episodes of suicidality in patients with a mood disorder. In such cases, admission allows clini- cians to provide effective interventions such as antidepress- ant and electroconvulsive therapy. But hospital stays are much less appropriate for chronically suicidal patients with BPD, for whom medical treatment is rarely definitive, and among whom suicidality tends to continue after discharge.
Therapists might do better to manage suicidality by fo- cusing on the implicit communication of distress behind thoughts and actions.55 Since suicidality “goes with the ter- ritory” of BPD, therapists must tolerate these patients’ anxi- ety. By and large, threats of suicide can be understood as an individual’s attempt to “raise the volume,” in the belief that this is the only way to be heard. When patients discover that therapists respond empathically to inner suffering and can offer ways of reducing their pain, suicidality may decline. Moreover, since BPD patients have difficulty breaking the links between thought and action, therapists can model this capacity by tolerating and resisting the contagion of despair and anxiety, and not acting on these feelings (e.g., by send- ing patients to the emergency room).
In some situations hospitalization of BPD patients is un- avoidable—for example, after a life-threatening suicide at- tempt, or to treat a psychotic episode.68 However, admitting patients every time they become suicidal works against the goals of therapy. Linehan56 advises her patients to avoid the hospital; if they end up there nonetheless, they can be held overnight, but no longer.
Clinicians often justify their fear that patients with BPD will commit suicide as a concern about “medical-legal issues.” The implication is that a therapist whose patient commits suicide will be held responsible for his or her death through litigation. Yet, as reviewed elsewhere,55 the majority of law- suits after suicide do not concern the treatment of chronically
320 Paris Harvard Rev Psychiatry
November/December 2002
suicidal patients with BPD, but of acutely suicidal patients with Axis I disorders.
Borderline patients may need more containment than is possible in outpatient therapy. In such cases, there is strong evidence for the value of day treatment.69–71 Unlike full ad- mission, partial hospitalization provides structure that limits regressive complications; it is also more cost effective.
DIRECTIONS FOR FUTURE RESEARCH
The remission of BPD, a disorder whose effects on functioning can be as serious as those of a psychotic illness,72 remains something of a mystery. Yet BPD is only one of many mental disorders that improve over time. Many impulsive disorders show similar recovery with age; even schizophrenia has a pat- tern of relative remission of positive symptoms by age 50.73
(Severe mood disorders, which often continue into old age, are an exception.74)
The main limitation of existing research is that it provides only a “snapshot” of BPD at different time points rather than a continuous assessment that could pinpoint mechanisms of remission and relapse. For this reason, the ongoing work at McLean and the Collaborative Longitudinal Personality Dis- orders Study have the potential to shed more light on the course and outcome of BPD and to provide more-precise guidelines for treatment.
REFERENCES
1. Schmideberg M. The borderline patient. In: Arieti S, ed. Ameri- can handbook of psychiatry, vol 1. New York: Basic, 1959:398– 416.
2. Werble B. Second follow-up study of borderline patients. Arch Gen Psychiatry 1970;23:3–7.
3. Carpenter WT Jr, Gunderson JG, Strauss JS. Considerations of the borderline syndrome: a longitudinal comparative study of borderline and schizophrenic patients. In: Hartocollis P, ed. Bor- derline personality disorders: the concept, the syndrome, the pa- tient. New York: International Universities Press, 1977:231–53.
4. Pope HG Jr, Jonas JM, Hudson JI, Cohen BM, Gunderson JG. The validity of DSM-III borderline personality disorder: a phe- nomenologic, family history, treatment response, and long-term follow-up study. Arch Gen Psychiatry 1983;40:23–30.
5. McGlashan TH. The Chestnut Lodge follow-up study, III: Long- term outcome of borderline personalities. Arch Gen Psychiatry 1986;43:20–30.
6. Plakun EM, Burkhardt PE, Muller JP. 14-year follow-up of borderline and schizotypal personality disorders. Compr Psy- chiatry 1985;26:448–55.
7. Stone MH. The fate of borderline patients: successful outcome and psychiatric practice. New York: Guilford, 1990.
8. Paris J, Brown R, Nowlis D. Long-term follow-up of borderline patients in a general hospital. Compr Psychiatry 1987;28:530–5.
9. Swartz M, Blazer D, George L, Winfield I. Estimating the preva-
lence of borderline personality disorder in the community. J Per- sonal Disord 1990;4:257–72.
10. Luborsky L. Clinicians’ judgment of mental health. Arch Gen Psychiatry 1962;7:407–17.
11. Endicott J, Spitzer RL, Fleiss JL, Cohen J. The Global Assess- ment Scale: a procedure for measuring overall severity of psy- chiatric disturbance. Arch Gen Psychiatry 1976;33:766–71.
12. Gunderson JG, Kolb JE. Discriminating features of borderline patients. Am J Psychiatry 1978;135:792–6.
13. Paris J, Nowlis D, Brown R. Predictors of suicide in borderline personality disorder. Can J Psychiatry 1989;34:8–9.
14. Kjelsberg E, Eikeseth PH, Dahl AA. Suicide in borderline pa- tients—predictive factors. Acta Psychiatr Scand 1991;84:283–7.
15. Silver D, Cardish R. BPD outcome studies: psychotherapy im- plications. Presented at the 144th Annual Meeting of the Amer- ican Psychiatric Association, New Orleans, May 1991.
16. Wilkinson DG. The suicide rate in schizophrenia. Br J Psychiatry 1982;140:138–41.
17. Guze SB, Robins E. Suicide and primary affective disorders. Br J Psychiatry 1970;117:437–8.
18. McGlashan TH. The prediction of outcome in borderline person- ality disorder: part V of the Chestnut Lodge Follow-up Study. In: McGlashan TH, ed. The borderline: current empirical research. Washington, DC: American Psychiatric Press, 1985:61–98.
19. Zanarini MC. Childhood experiences associated with the devel- opment of borderline personality disorder. Psychiatr Clin North Am 2000;23:89–101.
20. Paris J, Zweig-Frank H, Guzder J. Psychological risk factors in recovery from borderline personality disorder. Compr Psychia- try 1993;34:410–3.
21. Paris J, Nowlis D, Brown R. Developmental factors in the out- come of borderline personality disorder. Compr Psychiatry 1988; 29:147–50.
22. Pokorny AD. Prediction of suicide in psychiatric patients: report of a prospective study. Arch Gen Psychiatry l983;40:249–57.
23. Kullgren G. Factors associated with completed suicide in bor- derline personality disorder. J Nerv Ment Dis 1988;176:40–4.
24. Lesage AD, Boyer R, Grunberg F, Vanier C, Morrisette R, Ménard-Buteau C, et al. Suicide and mental disorders: a case- control study of young men. Am J Psychiatry 1994;151:1063–8.
25. Links PS, Heslegrave R, Van Reekum R. Prospective follow-up study of borderline personality disorder: prognosis, prediction of outcome, and Axis II comorbidity. Can J Psychiatry 1998;43:265– 70.
26. Links PS, Heslegrave RJ, Mitton JE, Van Reekum R, Patrick J. Borderline personality disorder and substance abuse: conse- quences of comorbidity. Can J Psychiatry 1995;40:9–14.
27. Zanarini MC, Frankenburg FR. Attainment and maintenance of reliability of Axis I and II disorders over the course of a longitu- dinal study. Compr Psychiatry 2001;42:369–74.
28. Zanarini MC, Frankenburg FR, Khera GS, Bleichmar J. Treat- ment histories of borderline inpatients. Compr Psychiatry 2001; 42:144–50.
29. Gunderson JG, Shea MT, Skodol AE, McGlashan TH, Morey LC, et al. The Collaborative Longitudinal Personality Disorders Study: development, aims, design, and sample characteristics. J Personal Disord 2000;14:300–15.
30. Grilo CM, McGlashan TH, Skodol AE. Stability and course of per-
Harvard Rev Psychiatry
Volume 10, Number 6 Paris 321
sonality disorders: the need to consider comorbidities and conti- nuities between Axis I psychiatric disorders and Axis II person- ality disorders. Psychiatr Q 2000;71:291–307.
31. Paris J, Zweig-Frank H. A 27-year follow-up of patients with bor- derline personality disorder. Compr Psychiatry 2001;42:482–7.
32. Zweig-Frank H, Paris J. Predictors of outcome in a 27-year fol- low-up of patients with borderline personality disorder. Compr Psychiatry 2002;43:103–7.
33. National Center for Health Statistics. Mortality data from the National Vital Statistics System, 1999. [Available on the World Wide Web at http://www.cdc.gov/nchs/about/major/dvs/mcd/ msb.htm].
34. Zanarini MC, Gunderson JG, Frankenburg FR, Chauncey DL. The Revised Diagnostic Interview for Borderlines: discriminat- ing BPD from other Axis II disorders. J Personal Disord 1989; 3:10–8.
35. Pepper CM, Klein DN, Anderson RL, Riso LP, Ouimette PC, Lizardi H. DSM-III-R Axis II comorbidity in dysthymia and ma- jor depression. Am J Psychiatry 1995;152:239–47.
36. Zanarini MC, Frankenburg FR, Dubo ED, Sickel AE, Trikha A, Levin A, et al. Axis I comorbidity of borderline personality disor- der. Am J Psychiatry 1998;155:1733–9.
37. Vaillant GE. Adaptation to life. Cambridge, Massachusetts: Har- vard University Press, 1977.
38. Vaillant GE. The natural history of alcoholism revisited. Cam- bridge, Massachusetts: Harvard University Press, 1995.
39. Black DW, Baumgard CH, Bell SE. A 16- to 45-year follow-up of 71 men with antisocial personality disorder. Compr Psychiatry 1995;36:130–40.
40. Keel PK, Mitchell JE, Miller KB, Davis TL, Crow SJ. Long-term outcome of bulimia nervosa. Arch Gen Psychiatry 1999;56:63–9.
41. Bandura A. Social learning theory. Englewood Cliffs, New Jer- sey: Prentice-Hall, 1977.
42. Bardenstein KK, McGlashan TH. The natural history of a resi- dentially treated borderline sample: gender differences. J Per- sonal Disord 1988;2:69–83.
43. Paris J, Braverman S. Successful and unsuccessful marriages in borderline patients. J Am Acad Psychoanal 1995;23:153–66.
44. Riley G. Divorce: an American tradition. New York: Oxford Uni- versity Press, 1991.
45. Weiss M, Zelkowitz P, Feldman RB, Vogel J, Heyman M, Paris J. Psychopathology in offspring of mothers with borderline personality disorder: a pilot study. Can J Psychiatry 1996;41: 285–90.
46. Feldman RB, Zelkowitz P, Weiss M, Vogel J, Heyman M, Paris J. A comparison of the families of mothers with borderline and non- borderline personality disorders. Compr Psychiatry 1995;36: 157–63.
47. Perry JC, Banon E, Ianni F. Effectiveness of psychotherapy for personality disorders. Am J Psychiatry 1999;156:1312–21.
48. Soloff PH. Psychopharmacology of borderline personality disor- der. Psychiatr Clin North Am 2000;23:169–92.
49. Linehan MM, Armstrong HE, Suarez A, Allmon D, Heard HL. Cognitive-behavioral treatment of chronically parasuicidal bor- derline patients. Arch Gen Psychiatry 1991;48:1060–4.
50. Linehan MM, Heard HL, Armstrong HE. Naturalistic follow-up of a behavioral treatment for chronically parasuicidal borderline patients. Arch Gen Psychiatry 1993;50:971–4.
51. Skodol AE, Buckley P, Charles E. Is there a characteristic pat- tern to the treatment history of clinic outpatients with border- line personality? J Nerv Ment Dis 1983;171:405–10.
52. Waldinger RJ, Gunderson JG. Completed psychotherapies with borderline patients. Am J Psychother 1984;38:190–202.
53. McGlashan TH. Implications of outcome research for the treat- ment of borderline personality disorder. In: Paris J, ed. Border- line personality disorder: etiology and treatment. Washington, DC: American Psychiatric Press, 1993:235–59.
54. Silver D. Psychotherapy of the characterologically difficult pa- tient. Can J Psychiatry 1983;28:513–21.
55. Paris J. Chronic suicidality among patients with borderline per- sonality disorder. Psychiatr Serv 2002;53:738–42.
56. Linehan MM. Cognitive-behavioral treatment of borderline per- sonality disorder. New York: Guilford, 1993.
57. Gunderson JG, Phillips KA. A current view of the interface be- tween borderline personality disorder and depression. Am J Psy- chiatry 1991;148:967–75.
58. Bolton S, Gunderson JG. Distinguishing borderline personality disorder from bipolar disorder: differential diagnosis and impli- cations. Am J Psychiatry 1996;153:1202–7.
59. Coccaro EF, Kavoussi RJ. Fluoxetine and impulsive aggressive behavior in personality-disordered subjects. Arch Gen Psychiatry 1997;54:1081–8.
60. Jang KL, Livesley WJ, Vernon PA, Jackson DN. Heritability of personality disorder traits: a twin study. Acta Psychiatr Scand 1996;94:438–44.
61. Paris J. Nature and nurture in psychiatry: a predisposition- stress model of mental disorders. Washington, DC: American Psychiatric Press, 1999.
62. Kandel ER. A new intellectual framework for psychiatry. Am J Psychiatry 1998;155:457–69.
63. Gunderson JG. Borderline personality disorder: a clinical guide. Washington, DC: American Psychiatric Press, 2001.
64. Barlow DH, ed. Clinical handbook of psychological disorders: a step-by-step treatment manual. 3rd ed. New York: Guilford, 2001.
65. Maltsberger JT. Calculated risks in the treatment of intractably suicidal patients. Psychiatry 1994;57:199–212.
66. Fine MA, Sansone RA. Dilemmas in the management of suicidal behavior in individuals with borderline personality disorder. Am J Psychother 1990;44:160–71.
67. Dawson D, MacMillan HL. Relationship management of the bor- derline patient: from understanding to treatment. New York: Brunner/Mazel, 1993.
68. Hull JW, Yeomans F, Clarkin J, Li C, Goodman G. Factors asso- ciated with multiple hospitalizations of patients with borderline personality disorder. Psychiatr Serv 1996;47:638–41.
69. Piper WE, Rosie JS, Joyce AS, Azim HFA. Time-limited day treatment for personality disorders: integration of research and practice in a group program. Washington, DC: American Psy- chological Association, 1996.
70. Bateman A, Fonagy P. Effectiveness of partial hospitalization in the treatment of borderline personality disorder: a randomized controlled trial. Am J Psychiatry 1999;156:1563–9.
71. Bateman A, Fonagy P. Treatment of borderline personality dis- order with psychoanalytically oriented partial hospitalization: an 18-month follow-up. Am J Psychiatry 2001;158:36–42.
322 Paris Harvard Rev Psychiatry
November/December 2002
72. Skodol AE, Gunderson JG, McGlashan TH, Dyck IR, Stout RL, Bender DS, et al. Functional impairment in patients with schizo- typal, borderline, avoidant, or obsessive-compulsive personality disorder. Am J Psychiatry 2002;159:276–83.
73. Harding CM, Brooks GW, Ashikaga T, Strauss JS, Breier A. Ver- mont longitudinal study of persons with severe mental illness,
II: Long-term outcome of subjects who retrospectively met DSM- III criteria for schizophrenia. Am J Psychiatry 1987;143:727–35.
74. Winokur G, Tsuang MT. The natural history of mania, depres- sion, and schizophrenia. Washington, DC: American Psychiatric Press, 1996.
Harvard Rev Psychiatry
Volume 10, Number 6 Paris 323
Copyright of Harvard Review of Psychiatry is the property of Taylor & Francis Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.