Here is the second half (without the references):

The Atypical-Bipolar II Connection

Major depressive states with reverse vegetative signs (so-called "atypical features") are commonly encountered in this unstable temperamental terrain. Three recent studies have provided greater clarification about this complex interface of volatile affective temperament and atypical affective states.

In collaboration with clinical researchers in Pisa,[16] we demonstrated that 72% of 80 depressive patients with DSM-IV atypical features simultaneously met the criteria for bipolar II; 60% had antecedent cyclothymic temperament. In addition, 94% were rated as interpersonally sensitive. As expected, using the DSM-IV axis II schema, both cluster B (borderline-histrionic) and cluster C (avoidant) personality disorders were prevalent.

Deltito and colleagues[17] studied 20 consecutive patients diagnosed borderline by experienced clinicians at Westchester-Cornell, "validated" independently by Gunderson's DIB. They then rated them by descending order of certainty of bipolarity: in light of what the current literature indicates as established bipolarity ( bipolar I + bipolar II), the conservative rate for bipolarity in this well-characterized, though small. sample of borderlines was 44%; taking the most liberal definition of bipolarity (including pharmacologic-hypomania, cyclothymic temperament, and family history for bipolar disorder), 81% of borderline patients could be considered lying on the border of within the bipolar spectrum. This provocative work, though preliminary, represents the first head-to-head comparison of borderline personality and bipolarity[11] (and obviously is in need of replication).

The author's research as part of the National Institute of Mental Health Collaborative Study of Depression[18] has shown that clinical features reminiscent of borderline features (SADS Item 12) were strongly predictive of which major depressives would, over a prospective observation period of 11 years, switch to bipolar II (Table 2). The temperament of these patients was a mélange of interpersonal sensitivity and mood lability. This study underscores the importance of temperamental factors in borderline psychopathology, as well as their value in predicting bipolar outcome. Stated more tersely, borderline personality, interpersonal sensitivity, mood labile temperamental traits, and bipolar II seem to represent overlapping expressions of the same diathesis. (That mood lability is not pathognomonic for borderline personality, and occurs in bipolar II as well, has been replicated by Henry and colleagues.[19])

Table 2. Prospective Prediction of Bipolar II Outcome in 8.6% of 559 Patients With Major Depressive Disorder*

- Younger age at onset
- High depressive recurrence
- Greater marital disruption
- Higher score on phobic anxiety
- More "borderline" clinical features
- Higher score on interpersonal sensitivity
- High on trait energy-activity and daydreaming
- High on trait mood lability
- 42% sensitivity
- 86% specificity

Borderline as a Casualty of the Axis I-Axis II Distinction

One of the inadequacies of our current nosologic schema of personality disorders is that the long-term functioning of patients suffering from major mood disorders is described primarily in "characterologic" language (Axis II in DSM-IV), which is conceptually removed from the "temperamental" language that had been used in classical European psychiatry during the first part of this century. For instance, today bipolar patients are often described as "dramatic," "erratic," "unstable," impulsive," "passive-aggressive," "histrionic," "narcissistic," or "borderline," as if affective temperaments had little to contribute to our understanding of these personality disorders. O'Connell and colleagues[20] appropriately pointed out that structured interviews tend to misclassify subthreshold affective disturbances as dramatic personality disorders. The author's work[21] and subsequent research by Levitt and colleagues[22] have actually shown significant overlap between the cyclothymic temperament and borderline personality disorder.

There are advantages to returning to the more natural affective temperamental language of describing the premorbid, intermorbid, and postmorbid phases of major affective disorders. In this framework,[7,21,23] affective temperaments represent the substrate from which the more florid episodes develop. Using the analogy of earthquakes, I have elsewhere[9] compared the predisposing terrain and affective instability in 2 types of depression. In many affective ill patients, the temperamental terrain is not visibly pathologic but refers to a vulnerable fault that can be destabilized periodically, erupting into extreme pathology that could lead to self-destruction. In this more classical affective type, the patient has relatively normal -- or even supernormal or hyperthymic -- functioning between episodes. In other patients, the temperamental terrain is characterized by greater instability and intermittent or nearly continuous emotional "mini-earthquakes"; these patients seem "protected," though not entirely, from major melancholic episodes. The patient with this second type of temperamental dysregulation suffers from protracted intermittent emotional disequilibrium and restlessness without necessarily having full-blown syndromal affective episodes. These are then considered to be "atypical" or "borderline" cases where the terrain is so pathologically unstable that it may be difficult to discern the superimposed episodes that are an accentuation of the basic pathology.

Borderline as the "Darker Side" of Cyclothymia

Our work has actually demonstrated that the temperamental terrain between depression and manic-depression is bridged by a spectrum of subtle bipolar disorders with an extremely variable course.[7,21,23] Mood switches are recurrent, biphasic and abrupt, and may be seasonal and sometimes exacerbated by antidepressants. The term "explosive" captures the abruptness of the affective switches, each phase lasting for hours, days, and, sometimes, weeks. These patients are rarely euthymic. Their mood shifts often follow a circadian pattern (ie, waking up convinced of the futility of existence), but can also be reactive to interpersonal altercations, often rather trivial in nature, but emotionally charged for the patient. Even when provoked by such situations, the resultant emotional outbursts are more like avalanches than understandable reactions proportional to the proximate provoking situation.[23] One must infer an endogenous propensity to extreme emotional reactivity to these patients. Given such emotional tempests, it is no wonder that most clyclothymes accumulate an extreme array of social disturbances by their mid-20s[21]: repeated romantic failure, episodic promiscuity, financial extravagance, uneven work or school record, dilettantism, geographic instability, polysubstance abuse, and joining various eschatologic cults. Such instability appeared to be secondary to lifelong biphasic mood swings below the threshold for full-blown bipolar disorder. Subsequent studies in a community sample[24] have reported similar interpersonal havoc and social disruptions.

The instability in the biography of cyclothymics is especially accentuated in those with predominantly irritable traits.[23] These individuals are habitually dysphoric, prone to anger, hypercritical and complaining, with a penchant for ill-humored joking. They would thus easily offend their loved ones, often leading to verbally abusive behavior when only minutes or hours earlier they had vowed "eternal" love. At other times, interpersonal crises escalate because of their pouting and obtrusive behavior. In brief, the morose temperamentality of the irritable cyclothymic provides the unstable base from which interpersonal tempests arise.

Recent data[25] from a French national collaborative study has shown that the notion of cyclothymia and hypomania as positive "sunny" traits and behaviors represents just one facet of soft bipolarity. This driven-euphoric facet should be contrasted with the irritable-tempestuous or "darker" side of bipolarity. The correlation of cyclothymia reaches significance (.37) only with the latter facet. In brief, depressions arising from a cyclothymic baseline are often characterized by dysphoric hypomanic periods, and are likely to be misdiagnosed as erratic personality disorders. Their high familial load for affective (including bipolar) disorder support their inclusion as a more unstable variant of bipolar II that can be best be characterized as "cyclothymic depressions."

Another study[26] relevant to the "darker" side of bipolarity, which is still unpublished, derives from the author's collaboration with the University of Pisa. In 107 atypical major depressive patients, logistic regression revealed that cyclothymic temperament accounted for much of the relationship between atypicality and borderline personality. The cyclothymic-sensitive disposition seems to represent the common denominator in the complex syndromic pattern of anxiety, mood, and impulse disorders. We explained these findings and considerations as support for the contention that atypical depression, borderline personality, cyclothymia, and bipolar II represent overlapping manifestations of a common underlying psychobiologic diathesis.

Unless their temperamental vulnerabilities are understood properly, the comorbidity and erratic clinical presentations of these borderline patients can baffle clinicians. As a result, some researchers prefer to characterize these vulnerabilities along the lines of sociopathy and related personality disorders.[27] Such characterization, in my opinion, misses the core emotionality of borderline patients that can be observed in both patients and in their biologic kin.[28] I submit that interpersonal sensitivity, mood reactivity, and lability are more germane to the origin of borderline psychopathology and its comorbidities. It is their innate -- and, to some extent, developmentally acquired -- intense reactivity to others that creates their turbulent relationships, and indeed their entire biography.

As discussed elsewhere,[10] the tragedy of borderline patients is that their impulsive drive, which thrusts them into the theater of human interactions -- coupled with their negative affectivity -- accentuates and thereby validates their sense of being rejected emotionally, maltreated, and abused; the more unfortunate among them do get actually abused by psychopathic family members (usually a step-parent).

Concluding Remarks

Discussion of psychopharmacologic and formal psychotherapeutic interventions in borderline personality conceived as a soft bipolar variant are beyond the scope of this paper. Suffice it to say that the affective framework for borderline personality described in this paper has major implications for clinical management. Foremost among them is that affective reconceptualization of borderline pathology may substantially reduce the therapists' countertransference[29] because now the patient is viewed as affectively ill, rather than "character flawed." Treatment should be undertaken with the requisite competence and confidence for a serious mental disorder. The suicide risk -- a potentially fatal consequence of the intense affective dysregulation -- should be conceptualized and clinically managed as rigorously as in any patient with serious mood disorder. The affective dysregulation and the impulsivity that underlies such risk may, in principle, be preventable with mood stabilizers, including carbamazepine and divalproex. This is a vital public health priority.

These patients often come from disturbed families and appear at risk for emotional instability due to both genetic factors (eg, bipolarity, alcoholism) and developmental factors (eg, disruption in early attachment bonds and other traumatic experiences). The neuroendocrine and sleep neurophysiologic correlates of their exquisite affective vulnerabilities have been documented elsewhere.[10] Since borderline patients -- in view of their negative affectivity -- often develop malevolent object representations of significant others in their lives,[30] clinicians must not assume that the parents of borderline patients are or were "monsters." Parents' guidance is often crucial to these patients' mastery of maturational tasks. On the other hand, Kurt Schneider's[31] wise admonition should not be forgotten: "On their bad days, keep out of their way as far as possible."