Unlike most personality disorders that were first described in Europe, the term "borderline personality" was introduced by American psychoanalyst Adolph Stern in 1938 to describe a patient group who did not fully fit the characteristics of psychotic or neurotic patient groups, thus existing on the "borderline" between the two. This concept of BPD persisted into the 1950s and 1960s. The identification and labeling of patients as "borderline" first arose during the era when psychiatry was dominated by the psychoanalytic paradigm. The classification system for mental disorders was primitive and dichotomous, with classification tied to patient capacity for analysis. Patients considered analyzable, and thus treatable, were diagnosed with neuroses, while those considered not analyzable, and therefore untreatable, were deemed to have psychoses [6].

The introduction of operationalized diagnoses for BPD and other disorders based on observable criteria in the 1980 DSM-III was considered a significant advancement in the field. However, the concept of and diagnostic criteria for BPD during and after the 2000 DSM-IV-TR became increasingly criticized on several grounds. For example, the description of BPD was non-specific. In the DSM III and the DSM-IV-TR, clinicians were instructed to diagnose BPD when five out of nine criteria were met. But, with this paradigm, individuals diagnosed with BPD could have as few as one criterion in common. This led to the same diagnosis given to patients with various criterion permutations, producing a heterogeneous patient group [11]. This issue is thought to have been largely resolved with the alternative DSM-5 criteria.

Research of BPD during the 13 years between the DSM-IV-TR and the DSM-5 clarified the understanding of BPD and prompted revisions to the diagnosis [4]. The greatest overall change between the DSM-IV and the DSM-5 has been the elimination of the multi-axial classification system, whereby BPD and other personality disorders were assigned a separate axis (Axis II). Several factors contributed to this change. The distinction between Axis I and Axis II disorders in earlier DSM editions received little empirical validation and increasingly became disputed in light of evolving research and clinical evidence. Personality disorders were traditionally conceptualized as the product of environmental factors, while Axis I disorders were viewed as having a biologic or organic cause. This dominant paradigm influenced the introduction of the multi-axial classification system in the DSM-III. While environmental stressors can contribute to personality disorder development, the same is also true with many Axis I disorders such as major depressive disorder (MDD) and PTSD. Also, BPD does not conform to traditional conceptions of personality disorders as ego-syntonic conditions; the symptoms of BPD are clearly ego-dystonic and lead patients to seek treatment for these symptoms [11,12]. Another criticism of DSM-IV-TR criteria was the combination of unstable, stress-induced symptoms and stable personality characteristics, also termed dimensional traits [13].

In the 2010s, a new model for diagnosing personality disorders was presented to the DSM-5 Task Force, and it was strongly and unanimously approved. However, the American Psychiatric Association Board of Trustees voted to sustain the DSM-IV-TR diagnostic system for personality disorders, including unchanged criteria for BPD, in the main section of DSM-5 due to insufficient evidence to validate the new proposed model [14]. The proposed new model is maintained in the DSM-5 as an "alternative DSM-5 model for personality disorders," and professionals have reported good clinical utility. These proposed BPD criteria are organized into two sections: impairments in personality (self and interpersonal) functioning and pathologic personality traits [4]:

The results of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) study, the first large-scale, community study of personality disorders in the United States, were published in 2008. This study found the overall lifetime prevalence rate for BPD was 5.9% (18 million people), with similar rates in men (5.6%) and women (6.2%). The prevalence of BPD was higher in Native American men; younger adults who were separated, divorced, or widowed; and persons with lower levels of education and socioeconomic status. Those with lower BPD prevalence were Hispanic men and women and Asian women [1].

In general, prevalence studies of BPD have found that prevalence in women is three times higher than in men in clinical settings, with similar gender prevalence in community settings [17,26,27]. Although epidemiologic studies have not formally assessed age of onset of BPD, data extrapolation from onset of self-harm (the most predictive symptom of BPD) suggests the onset of BPD occurs before 12 years of age in 32.8% of patients, begins between 13 and 17 years of age in 30.2%, and at 18 years of age or older in 37% [28,29]. The association of higher BPD prevalence with lower education, income, and socioeconomic class suggests these adversity factors predispose to developing BPD, although this association is likely bidirectional; BPD symptoms may contribute to poor educational achievement, lower income, and social class [18,26].

In general, prevalence studies of BPD have found that prevalence in women is three times higher than in men in clinical settings, with similar gender prevalence in community settings [17,26,27]. Although epidemiologic studies have not formally assessed age of onset of BPD, data extrapolation from onset of self-harm (the most predictive symptom of BPD) suggests the onset of BPD occurs before 12 years of age in 32.8% of patients, begins between 13 and 17 years of age in 30.2%, and at 18 years of age or older in 37% [28,29]. The association of higher BPD prevalence with lower education, income, and socioeconomic class suggests these adversity factors predispose to developing BPD, although this association is likely bidirectional; BPD symptoms may contribute to poor educational achievement, lower income, and social class [18,26].

Prospective longitudinal studies have identified environmental and parental factors that significantly contribute to BPD development. The number of BPD symptoms at 28 years of age has been found to be significantly and directly correlated with early attachment disorganization or maltreatment; maternal hostility, inconsistency, and/or over-involvement; aversive or hostile parental behavior; low parental affection; family disruption related to the father's presence; and family life stress. Maternal hostility and early life stress contributed independently to the prediction of BPD symptoms at 28 years of age [31,32]. Other studies identified additional factors associated with BPD development, including childhood physical abuse or neglect, sexual abuse, maladaptive parenting, maladaptive school experiences, and the demographic characteristics of low family socioeconomic status, family welfare support recipient status, and single-parent family status [33].

During the 1980s, the findings that childhood sexual abuse was prevalent in histories of patients with BPD led to the theory of childhood trauma as a primary etiologic factor in BPD development. Further research confirmed that while childhood trauma was highly prevalent in BPD, childhood sexual abuse was not necessary or sufficient for BPD development and did not account for much of the variance in causation [18]. In inpatient and outpatient settings, 40% to 70% of patients with BPD report childhood sexual abuse. Although traumatic childhood experiences, including childhood sexual abuse, are strong risk factors for later developing BPD, fewer than 10% of those with a history of childhood sexual abuse develop BPD, effectively eliminating childhood sexual abuse as a primary cause [34,35,36].

Specific adolescent risk factors for adult BPD have been identified. Substance use disorders, especially alcohol, during adolescence are a significant factor. In addition, depression and/or disruptive behavior disorders in childhood or adolescence, including conduct disorder, oppositional defiant disorder, and attention deficit hyperactivity disorder, are predictive of adult-onset BPD. A history of repetitive, intentional self-harm in childhood or adolescence is more common in adults with BPD than the general population [37].

Persons with BPD have high lifetime rates of other psychiatric disorders, including bipolar disorder (10% to 20%), MDD (71% to 83%), substance use disorder (50% to 65%), panic disorder (34% to 48%), social phobia (23% to 47%), PTSD (47% to 56%), and eating disorders (7% to 26%). Co-occurring personality disorders are also common in BPD, including avoidant (43% to 47%), dependent (16% to 51%), obsessive-compulsive (18% to 26%), and paranoid (14% to 30%) personality disorders [12,38].

While several medical conditions are more common in patients with BPD than the overall population, exact figures on prevalence are difficult to find. One study suggests that individuals with BPD had a higher risk of almost all somatic comorbidities (e.g., diabetes, obesity, cardiovascular disease) and a worse prognosis than individuals with other personality disorders [39].

Patients with BPD are heavy utilizers of intensive healthcare services, resulting in higher related healthcare costs than patients with other personality disorders or MDD [44,45]. Following suicide attempts or intentional self-injury, patients with BPD are typically hospitalized, and such episodes result in an average hospital stay of 6.3 days per year and roughly one emergency room visit every two years; these rates are 6 to 12 times higher than those of MDD [44,46,47]. Relative to patients with MDD, those with BPD are more likely to use almost every type of psychosocial treatment (except self-help groups) and most classes of psychotropic medications [44]. However, a prospective six-year study of patients with BPD found that while rates of hospitalization and day or residential treatment were high at study initiation, they significantly declined over time. Similar patterns were observed for intensive psychotherapy, while use of less intensive psychosocial therapy and polypharmacy remained stable during follow-up. At any time during the six-year period, 40% of patients took three or more concurrent medications, 20% took four or more, and 10% took five or more. Thus, outpatient utilization remained constant and inpatient utilization slowly declined over time [48].

It is essential to note that BPD is no longer viewed as solely the result of parental or primary caregiver behaviors that shape passive, inert children. It is now recognized that innate temperament and behaviors in a child influence parental behaviors by passively evoking parental behaviors and by actively soliciting certain types of parental interactions [50,51]. This has been demonstrated in twin studies that found elicitation of maternal warmth was substantially controlled by child temperament [52].

A temperament that predisposes sensitivity to interpersonal stress contributes to the development of BPD. In these infants and children, heightened distress states may trigger fearful response in a vulnerable, depressed, anxious, ill, or traumatized caregiver, further diminishing his or her already compromised availability to the child. In particular, child traits of interpersonal hypersensitivity and stress reactivity evoke parental reactions of fearfulness or helplessness and withdrawal, which significantly affect the vulnerable child. Parents of a pre-BPD child are likely to exhibit adverse responses when confronted by increasing neediness or anger in the child, with child and parent factors both contributing to an escalating series of negative and difficult interactions that contribute to adult BPD [53].

Proneness to distress, particularly at separation, is a core feature of ambivalent and disorganized attachment patterns, the childhood counterparts of adult BPD attachment dysfunction. Infants with insecure attachments show greater distress-prone temperaments and irritability and are more likely to express the ambivalent form of attachment. Ambivalent infants engage in hyperactivation behaviors intended to elevate their visibility and increase engagement from an inconsistently attentive parent; these behaviors include clinging, anger, resistance to contact, and failure to soothe in the presence of their parents. Most ambivalently attached children also show the features of disorganized attachment, while a subgroup of infants with disorganized attachment exhibit the amplified distress and difficulty in soothing that is observed in ambivalent attachment. This latter infant subgroup has a heightened vulnerability for developing BPD. Thus, infants born with highly distress-prone temperaments and raised under non-optimal conditions of parental attention and interaction are at greater risk of evolving into ambivalent and/or disorganized attachment and BPD [54].

Disorganized attachment is the precursor to unresolved attachment, one of two attachment forms in BPD. Its expression mimics the relational style of adults with BPD and involves contradictory approach and avoidance with dissociative responses to caregivers. Roughly 15% of infants show disorganized attachment patterns by 1 year of age, which predicts controlling patterns of attachment relations by 3 to 6 years of age and behavior problems when entering school. Infant use of disorganized attachment strategies is heightened when raised in environments of low socioeconomic status (24%), parental psychopathology (30% to 60%), and infant maltreatment (60% to 70%). Higher cortisol stress responses are found in infants with disorganized versus organized attachment strategies, reflecting a genetic basis of the serotonergic abnormalities and high cortisol responses to separations found in adults with BPD [55,56].

Early psychoanalytic theorists identified problems with separation from caregivers as a developmental failure central to vulnerability for BPD, with insecure attachment and traumatic separation experiences the pathogenic factors that accounted for abandonment fears in BPD [57,58]. Subsequent empirical investigation confirmed the association between dysfunctional early caretaking experience (e.g., frequent separations, parental over- or underinvolvement) and BPD diagnosis [59,60]. These previous findings have been substantively enhanced and refined by more recent investigations, including the identification from multiple lines of evidence of specific caregiver contributions to childhood development of early attachment disturbance and interpersonal hypersensitivity.

Psychopathology is highly prevalent in the parents of patients with BPD, although it is difficult to determine the exact prevalence due to limited studies. The prevalence of maternal BPD is 10% to 15%. These mothers are more insensitive to their infants at 2 months of age, and their children are likely to show disorganized attachments at 1 year of age. Also prevalent in parents of patients with BPD are substance abuse, depressive disorders, eating disorders, and antisocial or other personality disorders [68,69,70]. An estimated 30% are negative for psychiatric disorder history [53]. The prevalence of affective, impulsive, and interpersonal phenotypes is increased in families of patients with BPD, including findings that 50% of relatives have affective instability, 33% have impulsivity, and 28% have disturbed relationship styles comparable to their borderline offspring with BPD [68,71,72].

Roughly 65% of children with disorganized attachment undergo a change between 18 months and 6 years of age, whereby attachment becomes organized around the goal of controlling interaction with the primary attachment figure. This is likely an adaptive response designed to increase dysfunctional parent involvement (i.e., alleviate parental inability to meet the child's comfort and security needs) [73]. After 18 months of age, dysfunctional parenting is more likely to be personalized in children with negative emotion as anger at the parent, which a hypersensitive parent may experience as personally rejecting. Controlling behaviors of these children toward their caregiver have been observed as young as 3 years of age [73].

A genetic basis of BPD was identified in several family studies that found family loading for the disorder and significantly higher prevalence of BPD in first-degree relatives of patients with BPD than in the general population [70]. Research from a 2019 total population study estimated aggregation and hereditability among family members, which showed a pattern of decreased familial association with genetic relatedness [76]. The concordance hazard ratio of BPD was 11.5 for monozygotic twins and 7.4 for dizygotic twins [76]. Among full siblings, the hazard ratio indicated a 4.7 times greater risk of BPD, compared with maternal half-siblings (2.1 times) and paternal half-siblings (1.3 times). Cousin relations were also part of the study, and it was found that the hazard ratio was 1.7 for cousins whose parents were full siblings, 1.1 for cousins whose parents were maternal half-siblings, and 1.9 for cousins whose parents were paternal half-siblings. Heritability has been found to range from 44% to 60%, with individual specific environmental factors accounting for the remaining variance. This extent of genetic influence exceeds that of anxiety disorders and depression but is less than that of bipolar affective disorder or schizophrenia [18,76,77].

Neuropeptide research helps identify maladaptive brain processes involved in stress response and interpersonal sensitivity. As discussed, persons with BPD are typically exposed to high levels of stress during childhood from unstable and insecure attachments. Stress from fragmented insecure attachments interacts with and is amplified by genetic factors, and high stress levels continue through adulthood. Elevated cortisol response to psychosocial stress in subjects with BPD reflects physiologic alteration in stress management [92]. Endogenous opioid and oxytocin neuropeptide systems also mediate stress response and facilitate prosocial tendencies. Persons with BPD show functional alteration in both systems. In healthy individuals, oxytocin administration enhances interpretation of mental states from social cues or "mind-reading" and collaboration in social exchange tasks. But in people with BPD, oxytocin administration paradoxically increases mistrust and decreased cooperation in a social exchange, suggesting the oxytocin system in BPD increases mistrust and interpersonal instability in social activity [93,94,95]. However, this effect was not uniform; individual differences (e.g., attachment styles, rejection sensitivity) were observed [95]. One study included 31 patients with BPD and 31 healthy controls. Serum oxytocin levels at baseline were found to be significantly lower in patients with BPD than in healthy control subjects, whereas rejection sensitivity and childhood traumas were found to be significantly higher. No difference was found between the patient and control groups in terms of attachment styles, yet it was determined that there may be differences between the oxytocin levels of patients with BPD according to their attachment styles [96].

Emotion regulation is mediated by frontolimbic brain regions that include, among other structures, the amygdala, hippocampus, hypothalamus, dorsolateral and right dorsomedial prefrontal cortex, orbital frontal cortex, anterior cingulate cortex, and insula [100]. These and other structures are interconnected by function and structure and are recruited to modulate subcortical responses to emotional stimuli and inhibit behavioral impulses [101]. Dysfunction of this circuitry contributes to emotion dysregulation. Emotion regulation is best understood as an individual and interpersonal process that begins with early attachment and continues to later peer and romantic relationships. Early attachment and interpersonal relationships distribute the effort of emotion regulation through co-regulation [102].

Early attachment figures and relationships are the initial source of co-regulation. Bonding usually occurs quickly and unconditionally during a period of rapid development when neural links are being formed between the prefrontal cortex and structures that underlie emotion and memory including the amygdala, nucleus accumbens, and hippocampus [103]. These early relationships appear to have lasting effects on attachment style and emotion regulation. The caregiver-child relationship is the first experience in which the child learns to influence the caregiver's emotions and behaviors, and the child experiences self-regulation of behavior and emotions through caregiver actions [102].

Emotion dysregulation is a complex process of multiple interactive genetic and environmental components that begins in infancy and develops over the lifetime. The widely used definition of emotion dysregulation as the inability to flexibly respond to and manage emotions is overly broad and non-specific to BPD, lacking in nuance, and a hindrance to BPD research and clinical practice. Emotion dysregulation is characterized by four components: emotion sensitivity, negative affect, deficient appropriate emotion regulation strategies, and maladaptive emotion regulation strategies [109].

In BPD, emotion sensitivity is thought to have a biologic origin with presence in early life. It involves heightened emotional reactivity to environmental stimuli and is primarily associated with negative mood states such as anger, fear, and sadness. This heightened emotional reactivity also involves the emotions of others. Emotion recognition studies in BPD show negativity bias in emotion recognition (i.e., negative emotions in others are over-identified) and poor accuracy in correctly identifying facial emotions [110,111].

Ability to identify one's emotions is important for emotion regulation. Related to emotion awareness is capacity to distinguish among emotional states, termed emotional granularity. Persons with high emotional granularity can reliably and accurately differentiate emotional states (e.g., sadness from anger); those low in emotional granularity often describe emotional states in global terms, such as feeling good or feeling bad. Persons with BPD have shown deficient emotion awareness, low emotional granularity (as evidenced by poor emotional clarity and mood and emotion labeling), and greater affect polarity ("all-or-nothing" thinking) [114,115].

This led researchers to investigate whether a high lethality subtype of BPD could be identified. One study found that patients at highest risk for suicide had greater illness severity, vocational failures, and estrangement from family and friends. Low lethality subjects had better overall psychosocial functioning but more negativism, lifetime substance abuse, and histrionic and/or narcissistic personality disorder comorbidity. Suicide attempts of this latter group may reflect dramatic "communicative gestures," which show little change in medical lethality with repeat attempts [196]. Long-term longitudinal data have identified factors with greatest prediction of suicide attempt: diagnosis of MDD, substance use disorder, or PTSD; non-suicidal self-harm behavior; sexual assault as an adult; caretaker death from completed suicide; affective instability; and more severe dissociation. Prediction of suicide attempts in patients with BPD is complex and involves assessment of co-occurring psychiatric disorders, prominent BPD symptoms (i.e., self-harm, affective reactivity, and dissociation), adult adversity, and family history of completed suicide [197].

Maladaptive behaviors to regulate emotion can lead to emotion dysregulation problems very obvious to others. If negative affect becomes sufficiently intense, the person will likely choose maladaptive over adaptive behaviors. While maladaptive behaviors can produce immediate reduction in negative affect and are simpler to employ than adaptive behaviors, they have negative consequences and can become ineffective with long-term use [117]. The maladaptive cognitive strategies of rumination and thought suppression, often used in BPD, actually increase negative affect in the long term [118,119]. Experiential avoidance is also common in BPD and is characterized by behaviors to escape unwanted experiences [120]. The impulsive, suicidal, and self-injurious behaviors common in BPD are behaviors specifically used to regulate affect [109].

As they enter adulthood, persons with BPD may undergo multiple hospitalizations resulting from poor impulse control, suicidality, or quasipsychotic and dissociative symptomatology. As discussed, BPD decompensation accounts for a sizeable proportion of psychiatric hospitalizations. Employment history is often characterized by multiple job losses or career changes, and interpersonal relationships are continually volatile and chaotic. Fluctuations in gender identity, sexual orientation, and personal values are common, reflecting cognitive distortions and fragmented sense of self. By their 30s, affective instability and impulsivity generally begin to lessen, and forming a relationship with a supportive and patient sexual partner or simply retreating to a more isolated lifestyle may promote earlier stabilization of disruptive emotional lability [129,130].

Longitudinal data also found high levels of healthcare service utilization, with gradual reduction in the use of emergency rooms, inpatient admissions, and other expensive services [44,48]. Patients with BPD are heavy users of medications and polypharmacy, but a higher number of medications is correlated with worse clinical course. This may reflect over-reliance on medication as treatment despite modest (at best) benefit. Underuse of appropriate psychosocial therapies may contribute to a worsening clinical course, compelling the prescription of additional medications [18].

The full expression of BPD psychopathology occurs in the interpersonal context, and patients with BPD impose a legitimate and, on some dimensions, unique challenge to providers involved in their care. The propensity of patients with BPD to attempt suicide is probably the greatest source of provider stress. Patients with BPD are especially prone to feeling rejected and then reacting with rage, and the manner by which patients with BPD may endanger their lives can be unusually distress-provoking [138]. Difficult patient characteristics and prognostic pessimism have contributed to discrimination and bias in broader society and in the mental health system.

Recurrent suicidal threats or actions in response to fears of abandonment are by themselves strongly indicative of a BPD diagnosis. However, BPD remains underdiagnosed, and often misdiagnosed, in large part because the characteristic recurrent crises, emotional volatility, and self-injurious behavior are perceived as willful manipulative choices rather than expressions of illness [43,147,148]. A substantial gap exists between the education and practice of mental health care, and the current educational system for mental health professionals does not pay adequate attention to BPD or other personality disorders. Training health practitioners in BPD-related educational interventions can enhance positive attitudes and change practice toward people with BPD [149].

The DSM-IV-TR explicitly stated to exercise great caution when diagnosing BPD in patients younger than 18 years of age, largely from the belief that personality and behavioral patterns during adolescence are predominantly transient. In other words, adolescents may "outgrow" borderline symptoms, so diagnosing them before 18 years of age is premature [160,161]. However, more recent research on BPD and personality development indicates that an adolescent diagnosis of BPD is valid, and ignoring BPD as a possible disorder in adolescents may hamper effective clinical intervention [162]. The typical onset of self-harm before 12 years of age suggests an important window to screen and provide early intervention for these children and their families [29].

BPD is typified by significant impairments in identity, self-direction and interpersonal functioning, and pathologic overexpression of negative affectivity, disinhibition, and antagonism [4]. These pathologic deficits in personality functioning and pathologic personality traits are expressed as intense and disproportionate levels of anger, euphoria, depression, and anxiety, sometimes with rapid switching between mood states. This emotional intensity and instability leads to impulsive behavior, confusion, and shifting long-term goals, career objectives, friendships, gender identity, and values. Not infrequently, persons with BPD feel unfairly treated or misunderstood, bored, empty, and without a sense of who they are. Symptoms can be exacerbated by events in the environment that trigger emotional memory of past trauma or unresolved events [165].

In the absence of significant symptom expression in adolescence, the full onset of BPD can be triggered by events in adulthood considered normal developmental milestones, such as leaving home or starting an intimate relationship. Sometimes trauma is the triggering event, such as injury in a motor vehicle accident or sexual assault. Such events seem to precipitate the onset of BPD in predisposed persons such that BPD characteristics become fully expressed with attention to their condition for the first time [36,165].

The new proposed diagnostic criteria for BPD in the DSM-5 are reliable and replicable and reflect the observable manifestation of dysregulated interpersonal, behavioral, identity, and cognitive domains [109]. However, the symptom criteria do not fully capture the foundational basis of the psychopathology. Abnormal personality traits in BPD have been attributed to four factors, with each factor representing an underlying temperament or phenotype:

Abandonment fears may be confused with separation fears, and patients with BPD also frequently have intense separation anxiety concerning attachment figures. The self-image, affect, cognition, and behavior of the person with BPD can abruptly and profoundly change when he or she perceives an imminent separation, rejection, or loss of external structure. Interpersonal hypersensitivity can result in intense abandonment fears and inappropriate anger, triggered even when confronted by criticism or with a time-limited separation. Abandonment fears stem from an intolerance of being alone and a need to have other people with them, and frantic, impulsive actions to avoid abandonment can include self-injurious or suicidal behaviors [3].

Anger and aggression are likely to appear in any attachment relationship when abandonment is perceived, but most of all in the intimate relationships of persons with BPD. Perception of emotional distance or physical separation, coupled with intense fear of abandonment and loneliness, can provoke intense anger or rage. Studies of men and lesbians with BPD in treatment for domestic violence have found identical emotional and behavioral processes surrounding anger and violence toward intimate partners. For these patients, violence is used as a strategy to prevent abandonment by maintaining the connection to a partner through coercion and/or fear [166].

Splitting is the psychological construct describing the subconscious process that compartmentalizes bad, toxic, and/or terrifying representations of self and other from the good, rewarding, positive, and comforting representations [3]. This aspect of BPD is difficult for others, especially family members, to understand. A reality basis is rare; instead, patient perceptions are filtered by internal images of self and others that are exaggerated, distorted, and superimposed on relationships with others and on themselves [139].

Patients with BPD can experience dissociation symptoms, whereby the feeling and perception of self and/or environment has an unreal quality. These symptoms often occur during situations of extreme stress. These patients can also be unrealistically self-conscious, believing that others are critically looking at or talking about them. These lapses of reality in the patient with BPD are distinct from other pathologies because, with proper feedback, they are usually able to correct their distortions of reality [3].

Recurrent suicidal attempts, gestures, threats, or self-injurious behaviors are a hallmark of BPD, and this behavior is so prototypical with BPD that in the absence of other patient background information, recurrent self-destructive behaviors indicate a high probability of BPD. Self-harming acts often start in early adolescence or younger. Self-injurious behaviors or suicidal gestures are usually precipitated by interpersonal stressors, such as threats of separation or rejection, or by misinterpretation of emotion or communication in others as meaning that abandonment is imminent. When present, this clinical feature can greatly assist in the differential diagnosis in patients with dominant features of depression or anxiety [3].

The reliability of the diagnostic assessment for personality disorder has been considerably improved by the introduction of standardized interview schedules. There are two structured interview techniques widely used for diagnosing personality disorder with high specificity and reliability: the Structured Clinical Interview for DSM-5 Personality Disorders (SCID-5-PD) and the International Classification of Diseases, 10th Revision International Personality Disorder Examination (IPDE-ICD-10) [148,168,169]. The SCID-5-PD aligns with DSM-5 criteria for personality disorders diagnosis, while the IPDE-ICD-10 aligns with diagnostic criteria using the ICD-10 and DSM-IV-TR [148,168,169], One issue, common to many of the instruments, is the excessive length of time required for administration, with the interview for either method taking approximately one to three hours, depending on experience and skill level of the interviewing clinician [148,169]. A patient questionnaire is also available that can be completed in approximately 20 to 30 minutes, and will lessen the time of the interview. It should be noted that despite strengths such as reliability and direct correlation with DSM criteria, structured interviews and questionnaires may not fully capture the complexity and dynamics of patient mental health status. The diagnosis of BPD is most easily established by asking patients whether they believe the criteria for the disorder fits them and by listening to patients describe interpersonal interactions. Patients with BPD may be more likely to accept the assessment process by participating in the diagnosis. As discussed, patients and their families often find it helpful to be informed of the diagnosis and are relieved to learn that others share similar symptoms for which there are effective treatments [43].

The differential diagnostic issues surrounding BPD have changed over time. The original question of whether BPD was an atypical form of schizophrenia was dismissed with findings of the modest overlap in phenomenology and absence of familial or genetic connection. However, a novel case-control genome-wide study of BPD and comorbid conditions demonstrated a genetic overlap of BPD and other mental disorders, especially with bipolar disorder, and to a lesser degree with MDD and schizophrenia [171]. More research is needed to further explore the potential genetic overlap and association between the conditions.

Distinguishing BPD from bipolar disorder, and especially bipolar disorder II, can present a diagnostic dilemma due to the shared, overlapping symptoms. Both disorders have in common a substantial risk of suicide or suicide attempt, impulsivity, and inappropriate anger. However, symptoms that differentiate BPD include self-mutilation, self-injurious behavior without suicidal intent, and a frequent history of childhood abuse. Insecure attachments, reflected by intense abandonment fears, are hallmarks of BPD and uncommon in bipolar disorder. Patients with BPD have higher levels of impulsivity, hostility, and acute suicidal threats relative to those with bipolar disorder. Careful history taking usually elicits a differing time course of mood lability. Patients with BPD are extremely sensitive to rejection and do not have episodes of mania. Mood lability is often triggered by interpersonal sensitivity; mood lability in bipolar disorder tends to be autonomous and persistent [43,129,173].

The most frequent diagnostic error is confusing the chronic emotional instability and affect storms of patients with BPD with true hypomanic or manic behavior. This differentiation is easier with bipolar I, while the assumption of hypomanic behavior can form the basis for a bipolar II diagnosis. The diagnosis of bipolar disorder requires at least one episode of a manic (bipolar I) or hypomanic (bipolar II) episode. Accurate assessment of such an episode is essential and is done by patiently ascertaining whether the patient has one or several periods of three to four days (or longer) of dominant and unusually euphoric, angry, or irritated mood, with a sense of heightened energy, affective dyscontrol, little need to sleep, hyperactivity, and unusual behavior that contrasts with the patient norm. The behavior can involve inappropriate sexual exposure or behavior, gross recklessness with money or other properties, socially inappropriate approaches to others, and possibly elevated sexual drive along with overall expansiveness of mood and behavior. A full manic episode often involves loss of reality testing, grossly inappropriate social behavior with patient unawareness of the behavioral deviation, and possibly hallucinations or delusions that can lead to intervention by others [153].

Potential confusion between BPD and PTSD also arises from repeated findings that ongoing, chronic sexual, physical, or psychological traumatization, particularly in early childhood, constitutes an important etiologic factor in the development of a severe personality disorder, particularly BPD [153]. Additionally, there is a syndrome of severe early trauma leading to sequelae including a BPD-like syndrome, called complex PTSD, whereby trauma is the central concern and requires therapeutic priority. These patients experience great difficulty in trust and cognitive processing, rendering BPD treatments ineffective. However, as discussed, trauma in most patients with BPD is superimposed on a genetically determined pre-existing sensitivity, and although these patients experience psychophysiologic difficulty in processing trauma and communicating about these adverse events, as adults they benefit from therapies for BPD. Thus, patients with BPD, unlike those with PTSD, respond to treatments that focus on feelings and not trauma and challenge them to take control of their lives [18].

In contrast to patients with BPD who present different aspects of their internal world from one moment to the next, patients with narcissistic personality disorder mask the fragmentation and weakness of their identity under a brittle and fragile grandiose self that they present to the world and to themselves [153]. Patients with a severe narcissistic personality disorder may present symptoms strikingly similar to those of patients with BPD, including general impulsivity, chaos in relations with significant others, severe breakdown in their capacity for work and emotional intimacy, and parasuicidal and self-mutilating behavior. These patients are also prone to antisocial behavior, which requires the differential diagnosis among different types of narcissistic pathology with different levels of antisocial features [153].

Important differential features include the patient with narcissistic personality disorder's difficulty accepting any dependent relationship, their severe lack of investment in relations with significant others except in exploitative or parasitic relationships, and an aloofness that contrasts with the highly ambivalent yet clinging and dependent relationships of patients with BPD. Patients with narcissistic personality disorder can show extreme fluctuations between feelings of inferiority and failure and corresponding depressive reactions and an inordinate sense of superiority and grandiosity reflected by contemptuous and dismissive behavior toward others, including therapists. Patients with BPD may alternate their relationship between clinging dependency/idealization and angry rejection and dismissal, but do not show the chronically contemptuous and dismissive attitude of narcissistic patients. Resulting from these characteristics, patients with narcissistic personality disorder are usually isolated socially, even if they are externally part of a social network. They lose their friends and do not maintain relationships over an extended period of time, and their objective loneliness contrasts with the complicated, contradictory, yet enmeshed relationships of patients with BPD [153].

Kernberg stated that the adverse reactions to therapy that were highly commonplace in patients with BPD resulted from unconscious guilt (as an element of masochistic character structures), unconscious envy that underlies patient need to destroy what is received from their therapist, and unconscious identification with a primitive and sadistic object that underlies patient need to destroy the therapist as a good object [182]. Patient failure to improve with psychoanalytic therapy was solely attributed to pathologic motivations in the patient with BPD [6]. Improvements were rare exceptions rather than the rule. Although unknown at the time, in many patients with BPD traditional psychoanalytic therapy promoted symptom exacerbation from unintended toxic interaction between therapist approach and core BPD psychopathology [184]. Specifically, therapist neutrality encouraged patient projection and fueled abandonment fears, and therapist passivity promoted patient fears of disinterest and neglect. Therapist interpretations of negative motivations were experienced by patients as blaming and invalidating.

Psychotherapy is the current foundation of BPD treatment. Development of a secure attachment to the therapist is generally essential for patient improvement, but this does not come easily given the inherent intense needs and fears of attachment relationships. Patient symptoms can be difficult for professionals to manage, as they may assume the role of protective caretaker and can become angry and fearful when the patient suddenly reverts to dangerous or maladaptive behaviors. Patients with BPD may also abruptly terminate even highly skilled therapists. While this may be experienced as a failure by the provider, even brief therapy exposure is often later shown to have served a valuable purpose in helping the patient through a difficult period and in helping remove patient resistance to seeking and engaging subsequent therapists [3].

Group therapy for BPD can help in developing the following skills [184]:

For patients with BPD, hospitalization is usually restricted to the management of crises (including, but not limited to, situations in which patient safety is precarious) and is short in duration. Hospitals provide a safe place where the patient has an opportunity to gain distance and perspective on a particular crisis and where professionals can assess the patient's psychological and social problems and resources. It is not uncommon for medication changes to take place in the context of a hospital stay, so professionals can monitor the impact of new medications in a controlled environment [3].

However, hospital admission carries liabilities unique to patients with BPD. The American Psychiatric Association's recommendation of hospitalization whenever patients with BPD are suicidal has been criticized by experts for several reasons [193]. Patients can internalize the invalidating message of their inability to get through a crisis without hospitalization. Progress in therapy can be rendered impossible by repeated hospitalization, as this becomes the learned coping strategy for distress. Hospitalization also prevents addressing the interpersonal problem that triggered the crisis in the first place and can reinforce pathologic behaviors and make the patient worse [194].

The most dangerous features of BPD are self-harming behaviors and suicide risk. Suicidal ideation (i.e., ruminating and fantasizing about suicide) is pervasive in the BPD population [3]. Up to 10% of persons with BPD complete suicide, a rate 50 times greater than in the general population. More than 70% of those with BPD attempt suicide at least once, and patients with BPD attempt suicide an average of 3.3 times in their life [195]. Suicide attempts in BPD tend to peak when patients are in their 20s and 30s, although suicidality can occur in any age group [8].

Self-harm behavior assumes many forms, and patients with BPD often self-injure without suicidal intent. Most self-injury involves cutting, but it can also involve burning, hitting, head banging, or hair pulling. Some self-destructive behaviors are not perceived by patients as self-harming; among these are unprotected promiscuous sex, driving under the influence, and binging and purging [3]. Non-suicidal self-harm can also occur during hospitalization and may be expressed by treatment-sabotaging behavior. In one study, 63% of psychiatric inpatients who exhibited medically self-sabotaging behaviors had BPD. Behaviors and motivation included purposely avoiding needed medical treatment and/or prescribed medication(s) to hurt oneself; gravitating toward a dangerous situation hoping to be physically hurt; and damaging oneself on purpose to seek medical treatment [201].

Many clinicians have found concepts from dynamic deconstructive psychotherapy useful in understanding affective instability and suicide risk in patients with BPD [209]. This model frames the emotionally labile reactions to the environment as a switch between different states of being, or pseudo-personalities, such that patients can alternately present as helpless and childlike (helpless victim state), angry and self-righteous (angry victim state), or depressed and suicidal (guilty perpetrator state). These states reflect different sets of polarized and poorly integrated attributions of self and others, and not strategies of manipulation [205].

Emotional hyper-reactivity is most likely to manifest when limits are set or when a patient (mis)perceives an attachment relationship is about to dissolve. In both contexts, patients are likely to over-react in an emotionally volatile, angry, and possibly regressive manner. In the primary care setting, the patient with BPD may encounter the experience of refusal, or limit setting, in numerous situations, for example, clinician refusal to order a particular laboratory study, prescribe a requested medication or drug class, or make a particular referral. Refusal dynamics can also emerge with patient requests for unnecessary time-off-work excuses, automobile handicap flags, or disability status [212].

Therapy selection should best match the characteristics of the patient with BPD [129]. Psychodynamic therapy is suggested for patients with:

Supportive therapy approaches should be selected if the following patient characteristics are present:

Dialectical behavioral therapy is based on the theoretical principle that maladaptive behaviors, including self-injury, are attempts to manage intense overwhelming affect of biosocial origin. Dialectical behavioral therapy incorporates the two key elements of a behavioral, problem-solving approach blended with acceptance-based strategies, with an emphasis on dialectical processes. This therapy approach emphasizes balancing behavioral change, problem-solving, and emotional regulation with validation, mindfulness, and acceptance of patients. Therapists follow a detailed procedural manual [68,213]. The term "dialectical" refers to the philosophical principle of opposite truths, such that constructs can be opposing yet true at the same time. A core dialectic in this therapy is accepting patients where they are in the moment and working to help them change [194]. The five components of dialectical behavioral therapy are [194,213]:

Systems Training for Emotional Predictability and Problem Solving (STEPPS) is a manualized, CBT-based skills development package intended to supplement primary BPD therapy. The components of STEPPS include BPD psychoeducation, emotion management skills training, and behavior management skills training. It includes a two-hour session for family members and significant others, including members of the treatment team, to introduce them to the concepts and skills enabling them to provide support and reinforcement to the patient. STEPPS is delivered in 20 two-hour weekly group meetings led by two co-therapists, with each session addressing a specific emotion or behavior management skill. Throughout the 20 weeks, patients are instructed to monitor their thoughts, feelings, and behaviors to facilitate the recognition and monitoring of changes in the intensity and frequency of emotional episodes [40,202,215].

Mentalization-based therapy is a psychodynamic approach based on attachment and cognitive theory. Mentalization refers to the ability to accurately imagine the mental states of others, a normal developmental milestone attained by stable early attachment relationships. Patients with BPD are thought to have deficits in mentalization resulting from problematic early attachments. The core features of BPD are thought to reflect this failure to develop mentalizing ability and the resultant profound disorganization of self-structure [194,218].

Transference-focused psychotherapy is based on Kernberg's conceptualization of the core problem of BPD—excessive early aggression led the young child to split positive and negative images of him/herself and his or her mother [221]. The pre-borderline child is unable to merge positive and negative images and corresponding affects to attain a more realistic and ambivalent view of self and others. The primary goal of transference-focused psychotherapy is to reduce symptomatology and self-destructive behavior by modifying representations of self and others as enacted in the transference relationship. Clarifications, confrontations, and transference interpretations are the primary techniques of this twice-weekly psychotherapy [218,219].

While dialectical behavioral therapy, mentalization-based therapy, transference-focused psychotherapy, and interpersonal psychotherapy modified for BPD have all shown substantial benefit, no single therapy approach has emerged with greatest efficacy, which suggests that clinicians should offer the therapy modality that best matches their training, theoretical orientation, and preferences [223]. This point is underscored by repeated findings that treatment outcomes in patients with BPD is particularly influenced by the individual therapist [40].

Structured clinical management is based on a counseling model that resembles supportive therapy, with the addition of case management, advocacy support, problem solving, a crisis plan, medication review, and assertive follow-up if the patient begins missing appointments. Medication is used as an adjunct, when clinically indicated. Structured clinical management is provided by non-specialist clinicians, usually as weekly individual and group therapy sessions [226].

A comparison of mentalization-based therapy with structured clinical management found substantial improvements with both treatments across a range of clinical outcome measures. Mentalization-based therapy achieved steeper and somewhat larger effect sizes after 18 months, but structured clinical management was equally effective over the initial 6 months and patients receiving structured clinical management showed faster reduction in self-harming behaviors [220].

Good clinical care is a CBT-informed approach that incorporates a problem-solving paradigm as the core treatment intervention and stresses the importance of effective organizational structure. Psychologists trained in CBT provide therapy and case management. Patients are discussed in a weekly team meeting, with each team including a psychiatrist. A therapy session is typically offered once per week, which is flexible if patients need an additional contact, up to a maximum of 24 therapy sessions over six months. Case management is also flexibly provided, with clinical trials averaging around three management sessions for every therapy session. This underscores the point that effective intervention involves more than formal psychotherapy [160,226]. A study that randomized adolescents with BPD or BPD traits to cognitive analytic therapy or good clinical care found that subjects in both groups showed significant improvements across a range of clinical outcome measures, with little difference in benefit between therapies [160,227].

Two approaches for the primary care setting are motivational interviewing and problem solving. These are not treatment approaches per se, but instead are concrete, problem-focused tools that allow the clinician to improve the outcomes of patients with BPD while navigating some of the problematic behaviors and attitudes driven by the personality traits of BPD. Both were developed to address the very legitimate provider concerns over emotional endurance and job satisfaction when managing patients with BPD. When treating patients with BPD, clinicians should also consider a collaboratively developed crisis and safety plan and should use an overall approach of active listening, mindfulness, and strengthening patients' connections to their most important values [216,232,233].

One of the biggest challenges in the successful treatment of patients with BPD and other personality disorders is the patient's tendency for irregular treatment attendance, disengagement and premature discontinuation of treatment, and resistance to help and intervention. Motivational interviewing has shown promise in assisting the provider to motivate patients with BPD to engage in therapy and effect positive change (Table 6) [200,232].

Medications may effectively reduce a single or narrow range of targeted symptoms in BPD but have not yet shown convincing efficacy in addressing the core features (i.e., frantic efforts to avoid abandonment, emptiness, identity disturbance, and dissociation). Medication is considered adjunctive to psychotherapy, and prescribing psychotropic medication can help build a positive alliance with the patient with BPD. The selection and prescribing of medication for BPD is more complicated than in patients with other psychiatric conditions [18,236]. For example, patients with BPD are often highly perceptive to physiologic stimuli and medication side effects. Several strategies can be used to help optimize pharmacotherapy response and minimize interference from BPD pathology [18,184]:

Among the more rigorously designed clinical trials of drug therapy for BPD, no drug agent has been found effective in improving the core features of abandonment fears and hypersensitivity, chronic feelings of emptiness, identity disturbance, or dissociation. This may be due to outcome assessments lacking the ability to detect change in these symptoms or because these core BPD symptoms are not treatable with currently available drug therapies [237].

Aside from the finding of amitriptyline efficacy in reducing depression associated with BPD, tricyclic antidepressants have been found ineffective across a range of outcome measures [184]. This may be due to the prominent anticholinergic side effects further compromising the already tenuous behavioral control over impulsivity, aggression, and suicidality [238].

Second-generation atypical antipsychotics are more frequently prescribed for BPD, because they possess greater tolerability and show a broader therapeutic range due to their serotonergic and noradrenergic activity. Broadly, atypical antipsychotics are efficacious in reducing impulsive aggression, mood instability, anxiety, anger, impulsivity, and cognitive symptoms. Olanzapine and aripiprazole have accounted for a sizeable proportion of positive clinical trial results and have shown significantly improved affective instability, impulsivity, psychosis, and interpersonal dysfunction [222,235,238].

Aripiprazole has the added benefit of a long half-life and favorable metabolic profile, making administration easier and possibly increasing adherence and therapeutic benefit. As a partial agonist at D2 and 5-HT1A receptors and an antagonist at 5-HT2A receptors, aripiprazole may possess greater efficacy in reducing the impulsivity and aggression associated with BPD [244,245]. Despite a similar mechanistic profile, ziprasidone has not shown benefit in BPD [246].

Use of benzodiazepines has not received empirical support and is generally contraindicated for patients with BPD due to the risks of behavioral disinhibition, addiction, and overdose [184]. Benzodiazepines also impose the risks of inhibiting learning and interfering with skills acquirement [206]. Tricyclic antidepressants also pose a high risk of toxicity in overdose, including death due to fatal arrhythmia, and are generally avoided [257].

There is broad awareness that families of patients with BPD should, in most cases, be involved in the therapeutic process. One reason is that destructive family dynamics can greatly contribute to treatment drop-out by patients with BPD. In addition, families typically experience significant distress from living with and trying to cope with the problems of the patient with BPD. Regardless of the role family played in life adversity of the patient with BPD, they can become entangled in dysfunctional relationships with the patient that impede treatment. Family intervention can include providing psychoeducation concerning BPD and its origin, course, and treatment; teaching family members problem-solving skills to address difficult patient dynamics and provide the patient with validation; and transmit other communication skills to address the emotional reactivity of the relative with BPD [258,259,260].

Several interventions are available for family members of patients with BPD, with varying levels of intensity [184]. The first level is basic psychoeducation. This should be offered to all parents, spouses, and involved others and has the lowest intensity. The next level is counseling, which involves meeting with a therapist who assists family members with advice and problem solving. Families usually welcome these sessions. Support groups are offered in the community where available and include Family Connections, sponsored by the National Education Alliance for Borderline Personality Disorder (NEA-BPD), and various support groups through the National Alliance on Mental Illness. Attendance and involvement can be ongoing and long-term. Conjoint therapy sessions with the patient and parents can be useful for planning, problem solving issues related to budget, sleep hygiene, treatment adherence, emergencies, and provider vacations. This intervention can be very helpful in sustaining the holding environment and decreasing patient splitting. It is usually led by a family counselor, primary care clinician, or both.

Addressing family as the secondary client affirms the importance of assessing family needs as an aspect of care that engages the entire family system. The clinician can inform and support the family in gaining coping strategies for their distress and confusion resulting from chronic exposure to their relative's BPD symptoms [139]. For patients with severe BPD symptoms, the negative emotional impact on family members is best addressed by referral to family therapy. In family therapy, all parties participate to resolve communication problems and other family system stressors.

Medication is often part of the broader treatment plan. In some cases, the patient with BPD requires a medication choice that carries greater toxicity or lethality risk with overdose. Family members should be fully informed about medications, including the specific target symptoms, anticipated symptom changes from the medication, potential side effects, and actions to take in case of emergency [139].

Although MDD is virtually ubiquitous as a comorbidity in BPD and despite some overlap between MDD and BPD symptoms (such as chronic dysphoria in BPD and sadness and worthlessness in MDD), patients meeting full criteria may not benefit from antidepressants. The reasons for reduced antidepressant response remain unclear. Clinical trials have found that in patients with BPD and a co-occurring major depressive episode, improvement in BPD symptoms resulted in later improvements in major depressive symptoms, but the reverse was not found. In addition, patients with BPD receiving psychotherapy have shown reductions in self-reported depressive symptoms. Thus, specific treatment that targets BPD may be effective treatment for both disorders [38,222,263].

Although some pharmacotherapy trials have found reductions in comorbid depressive symptoms, these results are difficult to interpret, as most of the studies excluded patients with comorbid MDD. Furthermore, while the interventions led to reductions in subsyndromal depression symptoms, remission rates in patients with comorbid BPD and MDD were not evaluated. Studies combining medication and specialized psychotherapy have shown mixed results, generally supporting the conclusion that treatment of BPD leads to improvement in depressive symptoms. Psychotherapy is generally more important than pharmacotherapy in positive treatment outcomes of BPD and comorbid MDD [38,222].

Comorbid PTSD is a more complex problem. It is important that BPD is not treated as a variant of PTSD. Evidence-based psychotherapies for BPD, such as dialectical behavioral therapy, tend to focus on the present and short-term future, and these alone are not helpful for PTSD. In some cases, time-limited, evidence-based CBT for PTSD has been useful [38].

An ongoing substance use disorder is highly important in the course of BPD and can be very problematic. There is some evidence that active substance use disorder is associated with more severe BPD symptoms and a worse intermediate-term prognosis, but the adverse effects on prognosis may attenuate over time. Clinical trials of patients with BPD and comorbid substance use disorder suggest that successful psychotherapy can reduce BPD and substance use disorder symptoms. Added benefits have been found using a dialectical behavioral therapy-based smartphone application in reducing substance use urges [264,265].

When eating disorders are comorbid with BPD, rates of the eating disorder tend to decline over time, although change to another eating disorder may also occur. This suggests that eating disorder symptoms reflect the core impulsivity of BPD. Patients who exhibit serious weight loss from anorexia nervosa require treatment for the eating disorder before starting BPD therapy. Binging and purging behaviors can be addressed by the same treatment approaches used to reduce self-harming behaviors in specialized psychotherapies for BPD [38,266].

However, the course of BPD rarely, if ever, shows a simple linear improvement. The frequent alternation between progress and setback is emotionally draining for everyone involved. Long-term studies of the course of BPD found the first five years of treatment are typically the most crisis-ridden. A series of intense, unstable relationships that end angrily, with subsequent self-destructive or suicidal behaviors, are characteristic. Although such a pattern may persist for years, decreasing frequency and seriousness of self-destructive behaviors, decreased suicidal ideation and acts, and declining frequency and duration of hospitalization are early indicators of improvement. Following hospitalization, roughly 60% of patients with BPD are readmitted in the first 6 months; this declines to 35% 18 to 24 months after initial hospitalization. In aggregate, utilization of psychiatric care gradually diminishes over time to involve briefer, less intensive interventions [3].