Fluidity of the Subsyndromal Phenomenology of Borderline Personality Disorder Over 16 Years of Prospective Follow-Up

Recent reports from two National Institute of Mental Health (NIMH)-funded large-scale, prospective studies of the long-term course of borderline personality disorder—the McLean Study of Adult Development (1) and the Collaborative Longitudinal Personality Disorders Study (2)—have found high rates of remission of borderline personality disorder after 10 years of follow-up. In the McLean Study of Adult Development, it was found that 93% of the borderline patients achieved a remission that lasted at least 2 years. In the Collaborative Longitudinal Personality Disorders Study, it was found that 85% of borderline patients achieved a remission lasting 12 months or more.

The McLean Study of Adult Development also studied remissions of borderline personality disorder after 16 years of prospective follow-up (3). It was found that 99% of borderline patients had a 2-year remission, 95% had a 4-year remission, 90% had a 6-year remission, and 78% had an 8-year remission. Recurrences of borderline personality disorder were also studied. It was found that 36% had a recurrence after a 2-year remission, 25% had a recurrence after a 4-year remission, 19% had a recurrence after a 6-year remission, and 10% had a recurrence after an 8-year remission.

Taken together, the results of these studies have led patients, their families, and the mental health clinicians treating them to be more hopeful about the long-term prognosis of borderline personality disorder. However, less attention has been paid to the course of the symptoms of borderline personality disorder. At the 6-year follow-up of the McLean Study of Adult Development, it was found that all 24 symptoms studied declined significantly over time but remained significantly more common among borderline patients than axis II comparison subjects (4).

At the 10-year follow-up, we determined time to remission of each of these 24 symptoms and found that 12 were what we termed acute symptoms and the other 12 were what we termed temperamental symptoms (5) (see Table 1). Acute symptoms, such as self-mutilation and help-seeking suicide attempts, resolve relatively rapidly, are the best markers for the disorder (as they are both dramatic and quite specific [6] to borderline personality disorder, particularly in conjunction with one another), are often the main reason for expensive forms of psychiatric care, such as hospitalizations, and are akin to the positive symptoms of schizophrenia. In contrast, temperamental symptoms, such as chronic anger and intolerance of aloneness, resolve relatively slowly, are not specific to borderline personality disorder, are associated with ongoing psychosocial impairment, and are akin to the negative symptoms of schizophrenia.

The present study, which is an extension of the McLean Study of Adult Development mentioned above, builds on our prior work in three important ways. First, we followed our two study groups over 6 additional years of prospective follow-up. Second, we assessed cumulative remission rates lasting 2 and 4 years for the 24 symptoms of borderline personality disorder we are studying. Third, we assessed cumulative recurrence rates for these symptoms—something that to the best of our knowledge has never been studied before. We hypothesize that the acute symptoms are more likely to remit and less likely to recur than the temperamental symptoms over the course of the study.

Method

The methodology of this study has been described previously (4). Briefly, all subjects were initially inpatients at McLean Hospital in Belmont, Massachusetts. Each patient was first screened to determine that he or she 1) was between the ages of 18 and 35; 2) had a known or estimated IQ of 71 or higher; 3) had no history or current symptoms of schizophrenia, schizoaffective disorder, bipolar I disorder, or an organic condition that could cause serious psychiatric symptoms; and 4) was fluent in English.

After the study procedures were explained, written informed consent was obtained. Each patient then met with a masters-level interviewer blind to the patient’s clinical diagnoses for a thorough diagnostic assessment. Three semistructured interviews were administered. These interviews were 1) the Structured Clinical Interview for DSM-III-R Axis I Disorders (7), 2) the Revised Diagnostic Interview for Borderlines (8), and 3) the Diagnostic Interview for DSM-III-R Personality Disorders (9). The interrater and test-retest reliability of these three diagnostic measures (10, 11) have been found to be good-to-excellent.

At each of eight follow-up waves, separated by 24 months, our diagnostic battery was readministered by staff members blind to previously collected information. Good-to-excellent follow-up (within a generation of raters) and longitudinal (between generations of raters) interrater reliability were maintained throughout the course of the study for both axis I and II disorders (10, 11).

We defined remission as no longer meeting criteria for a symptom of borderline personality disorder (no longer being rated as definitely having that symptom) as determined by the Revised Diagnostic Interview for Borderlines (N=22 symptoms) or the Diagnostic Interview for DSM-III-R Personality Disorders (N=2 symptoms not covered by the Revised Diagnostic Interview for Borderlines) for a period of 2 years (or one follow-up period). We also studied remissions lasting 4 years (or two consecutive follow-up periods).

We defined recurrence of a symptom as again meeting criteria for that symptom (again being rated as definitely having that symptom) as determined by the Revised Diagnostic Interview for Borderlines or the Diagnostic Interview for DSM-III-R Personality Disorders after a remission lasting 2 years. We also studied recurrences following remissions lasting 4 consecutive years (or two consecutive follow-up periods).

Statistical Analyses

The Kaplan-Meier product-limit estimator (of the survival function) was used to assess cumulative rates of remission lasting 2 and 4 years of the 24 symptoms of borderline personality disorder being studied. We defined time to attainment of these outcomes as the follow-up period at which these outcomes were first achieved.

The Kaplan-Meier product-limit estimator was also used to assess cumulative rates of recurrence after remissions lasting 2 and 4 years. We defined time to recurrence as the number of years after first attaining these remissions until criteria for the symptom were met again.

Finally, Cox proportional survival analyses were used to compare the borderline patients and axis II comparison subjects in terms of these time-to-event outcomes; these analyses yield a hazard ratio and 95% confidence interval for the comparison of the two diagnostic groups. Statistical comparisons were only made when the risk set in each group was no less than 20 subjects. To partially adjust for multiple testing, we set the significance cut-off at the 0.005 level.

Results

The study sample and the diagnostic characteristics of the sample have been described previously (4). Two hundred and ninety patients met both Revised Diagnostic Interview for Borderlines and DSM-III-R criteria for borderline personality disorder and 72 met DSM-III-R criteria for at least one nonborderline axis II disorder (and neither criteria set for borderline personality disorder). The following diagnoses were found among the 72 comparison subjects (N=34): antisocial personality disorder (N=10, 13.9%), narcissistic personality disorder (N=3, 4.2%), paranoid personality disorder (N=3, 4.2%), avoidant personality disorder (N=8, 11.1%), dependent personality disorder (N=7, 9.7%), self-defeating personality disorder (N=2, 2.8%), and passive-aggressive personality disorder (N=1, 1.4%). Another 38/72 of these comparison subjects (52.8%) met criteria for personality disorder not otherwise specified (which was operationally defined in the Diagnostic Interview for DSM-III-R Personality Disorders as meeting all but one of the required number of criteria for at least two of the 13 axis II disorders described in DSM-III-R).

Baseline demographic data have also been presented previously (4). Briefly, 77.1% (N=279) of the subjects were female and 87% (N=315) were white. The average age of the subjects was 27 years (SD=6.3), the mean socioeconomic status was 3.3 (SD=1.5) (where 1=highest and 5=lowest) (12), and the mean Global Assessment of Functioning score was 39.8 (SD=7.8) (indicating major impairment in several areas, such as work or school, family relations, judgment, thinking, or mood).

In terms of continuing participation, which has also been described previously (3), 87.5% (N=231/264) of surviving borderline patients (13 died by suicide, and 13 died of other causes) were reinterviewed at all eight follow-up waves. A similar rate of participation was found for axis II comparison subjects, with 82.9% (N=58/70) of surviving patients in this study group (one died by suicide, and one died of other causes) being reassessed at all eight follow-up waves.

The cumulative rates of remission from the symptoms of borderline personality disorder lasting 2 or 4 years are presented in Table 2. As can be seen, estimated rates of 2-year remissions were very high for both groups. The cumulative rates of 2-year remission of the 24 symptoms studied reported by borderline patients ranged from a low of 62% to a high of 100%, with a median of 93%. The cumulative rates of 2-year remissions of these symptoms reported by axis II comparison subjects ranged from a low of 77% to a high of 100%, with a median of 97%. Additionally, borderline patients were significantly slower to remit for 2 years on 12 of the symptoms than axis II comparison subjects; five symptoms were not compared due to the small number of axis II comparison subjects in the risk set.

The cumulative rates of 4-year remissions of many of the 24 symptoms studied were substantially lower than the cumulative rates of 2-year remissions reported by both borderline patients and axis II comparison subjects. For borderline patients, these rates of 4-year remissions ranged from a low of 34% to a high of 100%, with a median of 74%. For axis II comparison subjects, these rates of 4-year remissions ranged from a low of 56% to a high of 100%, with a median of 89%. Borderline patients were also significantly slower to remit for 4 years on 13 of the symptoms than axis II comparison subjects; five symptoms were not compared due to the small number of axis II comparison subjects in the risk set.

The cumulative rates of recurrence after remission of the 24 symptoms of borderline personality disorder being studied are presented in Table 3. By the time of the 16-year follow-up, cumulative rates of recurrence for these symptoms reported by borderline patients after a 2-year remission ranged from a low of 30% to a high of 98%, with a median of 75%. The comparable figures for axis II comparison subjects ranged from a low of 14% to a high of 94%, with a median of 58%. After 4-year remissions, rates of recurrence for borderline patients ranged from a low of 18% to a high of 96%, with a median of 56%. In terms of axis II comparison subjects, symptom recurrence after a 4-year remission ranged from a low of 0.0% to a high of 85%, with a median of 47%. Borderline patients had a significantly faster time to recurrence after a 2-year remission on three of the symptoms; there were no significant differences between the two diagnostic groups after a 4-year remission for any of the symptoms compared.

These general findings about cumulative rates of remission and recurrence for borderline patients and axis II comparison subjects are important, particularly as they indicate that the course of the symptoms of the disorder is more fluid than the rates of remission and recurrence of the disorder itself (1, 3). However, the cumulative rates of these outcomes of acute and temperamental symptoms are more revealing clinically. All 12 acute symptoms (and no temperamental symptoms) remitted for 2 years at higher rates than the median value for borderline patients of 93%. One of these acute symptoms was in the affective realm (affective instability), and two were in the cognitive realm (quasi-psychotic thought and serious identity disturbance, which is categorized as cognitive in nature because it is based on overvalued ideas about the self). The others were either in the impulsive realm (substance abuse/dependence, sexual deviance [mostly promiscuity], self-mutilation, and help-seeking suicide attempts) or in the interpersonal realm (stormy relationships, devaluation/manipulation/sadism, demandingness/entitlement, treatment regressions, and counter-transference problems/“special” treatment relationships). Eleven of these acute symptoms (all but affective instability) (and one temperamental symptom: dependency/masochism) remitted for 4 years at higher rates than the median value for borderline patients of 74%.

In terms of recurrences after a 2-year remission, 11 of these acute symptoms (all but affective instability) (and one temperamental symptom: dependency/masochism) recurred at lower rates than the median value for borderline patients of 75%. In terms of recurrences after a 4-year remission, 11 of these acute symptoms (all but affective instability) (and one temperamental symptom: dependency/masochism) recurred at lower rates than the median value for borderline patients of 56%.

We also conducted exploratory analyses to determine cumulative rates of remissions of these 24 symptoms lasting 6 and 8 years, as well as cumulative rates of recurrences of these symptoms after these remissions. It was found that rates of remission and recurrence declined substantially. For example, 74% of borderline patients with a baseline history of self-mutilation experienced an 8-year remission of this symptom, and of those who remitted after an 8-year remission, only 17% experienced a recurrence. However, due to sparseness of subjects, particularly comparison subjects, who experienced these events, detailed results of these analyses are not presented.

Discussion

Four main findings have emerged from this study. The first is that 2-year remissions of the 24 symptoms studied were very common for those in both study groups. The median cumulative 2-year remission rate was found to be 93% for borderline patients and 97% for axis II comparison subjects. Not surprisingly, the rates of 2-year remission found after 16 years of prospective follow-up were somewhat higher for those in both study groups than the rates found after 10 years of prospective follow-up reported in one of the earlier studies of this sample mentioned above (5). More specifically, the median rates for borderline patients were 93% versus 80%, and the median rates for axis II comparison subjects were 97% versus 95%.

The second main finding is that the rates of recurrence of the 24 symptoms being studied were quite high after a 2-year remission but somewhat lower after a 4-year remission. More specifically, the median cumulative recurrence rate for borderline patients was 75% after a 2-year remission and 56% after a 4-year remission. A similar pattern was found for axis II comparison subjects: 58% and 47%. Some of the decline in the rates of recurrence may be due to the shorter length of follow-up after a 4-year remission.

The third main finding is that axis II comparison subjects exhibited higher remission rates and lower recurrence rates of these symptoms than borderline patients. This is consistent with results from studies of remissions and recurrences of the disorder itself (1, 3). However, it should be noted that the base rates of these symptoms were typically substantially higher and thus more problematic for those with borderline personality disorder.

The fourth main finding is that acute symptoms had higher remission rates and lower recurrence rates than temperamental symptoms. This might be seen as a treatment effect, as all five of the comprehensive empirically based psychotherapies for borderline personality disorder (13–17) are focused on acute symptoms. However, very few of the subjects in this study were ever in one of these treatments because all of them are hard to access due to an insufficient number of trained practitioners. It may be that the supportive therapies most patients reported being in also focused on these symptoms due to their association with turbulence in the therapeutic relationship, emergency department visits, and inpatient stays.

In contrast, 11 temperamental symptoms (all but dependency/masochism) were consistently found to have relatively low rates of remission and high rates of recurrence. Five of these symptoms are affective in nature: chronic/major depression, chronic feelings of helplessness/hopelessness/worthlessness, chronic anger/frequent angry acts, chronic anxiety, and chronic loneliness/emptiness. It may be that these dysphoric affects are in this stubborn group of symptoms because they represent inadequately treated mood and/or anxiety disorders, which are common co-occurring disorders among borderline patients (18). However, a high proportion of borderline patients used psychosocial treatments (19) and psychotropic medications (20) over time. It may also be that these symptoms represent a reasonable response to the limited lives led by a substantial percentage of borderline patients—lives marked by social isolation and serious vocational impairment (21).

Perhaps the prototypic affective symptom of borderline personality disorder is intense anger. This may well be due to the negative reactions that it engenders in clinicians, family members, and friends—negative reactions of fear and/or revulsion that can lead to struggling with the patient or distancing oneself from him or her.

The other temperamental symptoms showing this pattern of relatively low rates of remission and high rates of recurrence were odd thinking/unusual perceptual experiences, nondelusional paranoia, general impulsivity, intolerance of aloneness, abandonment/engulfment/annihilation concerns, and counterdependency/serious conflict over help/care. These symptoms, which span the cognitive, impulsive, and interpersonal sectors of borderline psychopathology, shared this pattern in a less severe and less consistent manner than found for affective symptoms. However, they too are among the more temperamental and less turbulent of the symptoms of borderline personality disorder.

Taken together, these findings have important clinical implications. In general, they give clinicians a cognitive map of what to expect over time. It seems that the affective symptoms of the disorder are the most chronic—briefly remitting and then recurring. This suggests that clinicians should not be discouraged that these symptoms persist or try to address them with very aggressive polypharmacy that can have negative health consequences, including substantial weight gain and obesity (22), and have no empirical basis (23). Rather, helping patients to accept these feelings may be the most useful strategy and in fact is a core feature of dialectical behavioral therapy (13). These results also suggest that many of the impulsive symptoms are the most likely to remit and the least likely to recur. As most clinicians practice treatment as usual, these results may help to alleviate the fear that many clinicians have that their patients will die either through indirect means that are self-inflicted, such as accidentally overdosing, or actually kill themselves. Rather, these results suggest that many of these symptoms will resolve in time without specialty intervention.

The cognitive and interpersonal symptoms of borderline personality disorder are comprised somewhat evenly of both acute and temperamental symptoms. Here too the results of the present study can provide a cognitive map of what to expect in these symptomatic realms over time. Quasi-psychotic thought and serious identity disturbance are categorized as acute symptoms, while odd thinking/unusual experiences and nondelusional paranoia are categorized as temperamental symptoms. Gunderson (24) has long suggested that added support and structure are the most effective interventions for quasi-psychotic symptoms. This approach may also be useful for overvalued ideas of inner badness (odd thinking), depersonalization (unusual perceptual experiences), and distrust (non-delusional paranoia) as these symptoms are on a continuum of severity, with quasi-psychotic thought being a more serious symptom than odd thinking, unusual perceptual experiences, and non-delusional paranoia.

In terms of the nine interpersonal symptoms studied, five are categorized as acute in nature (e.g., stormy relationships and serious treatment regressions), while four are categorized as temperamental in nature (e.g., intolerance of aloneness and undue dependency/masochism). Clinical experience suggests that the acute symptoms are more difficult for clinicians to handle because they are so action-oriented. While it is heartening that they have high rates of remission and relatively low rates of recurrence, clinicians wonder what they can do to speed this result. Many patients with borderline personality disorder have some awareness that being manipulative or demanding, for example, interferes with their having the supportive relationships they crave. Therapists can build on this awareness in a way that is useful by tactfully confronting these behaviors, labeling them self-destructive, and pointing out more adaptive ways of getting one’s needs met. Also, in our clinical experience, the interpersonal symptoms of borderline personality disorder that are temperamental in nature are quite treatment-resistant and tend to persist in patients after their more acute interpersonal symptoms have resolved somewhat. Clinicians have two options in dealing with these symptoms. They can try to lessen the severity of these symptoms, or they can try to help their patient come to terms with the more limited life with which they are associated. Both are reasonable strategies, and the best choice depends on the patient’s competence and resilience over his or her lifetime.

This study has two main limitations. One limitation of this study is that all of the patients were seriously ill inpatients at the start of the study. Another limitation is that about 90% of those in both patient groups were in individual therapy and taking psychotropic medications at baseline, and about 70% were participating in each of these outpatient modalities during each follow-up period (19). More specifically, 65% of borderline patients were in individual therapy, and 71% were taking standing medications at the 16-year follow-up, with 58% taking antidepressants, 26% taking anxiolytics, 28% taking antipsychotics, and 29% taking mood stabilizers (20). Thus, it is difficult to know whether these results would generalize to a less disturbed group of patients. It is also difficult to know whether these results would generalize to people meeting criteria for borderline personality disorder who were not in treatment, which in the present study was typically nonintensive outpatient treatment as usual in the community.

Taken together, the results of this study suggest that the symptoms of borderline personality disorder are quite fluid, with remissions and recurrences being common. However, the more clinically urgent acute symptoms of borderline personality disorder seem to have a better prognosis than the less turbulent temperamental symptoms of the disorder.