Borderline Personality Disorder or First‐Episode Psychosis? Challenges in Assessing Psychotic Features in Early Intervention: A Case Report

ABSTRACT

Introduction

Adolescence is a crucial period for the onset of both borderline personality disorder (BPD) and first‐episode psychosis. Although transient stress‐related paranoid ideation and dissociative symptoms are one of the diagnostic criteria for BPD, determining when these psychotic features progress to early‐stage schizophrenia remains unclear.

Methods

Two case studies aimed to explore the challenges in assessing and following up on psychotic features in BPD over time. Two adolescent girls aged 14 and 15 were initially diagnosed with BPD using semi‐structured clinical interviews for DSM‐IV personality disorder. Five years later, symptom development was explored using comprehensive diagnostic interviews.

Results

Both cases exhibited lapses in reality testing at baseline, interpreted as transient and stress‐related symptoms. The first case illustrated the difficulties in distinguishing normative magical and imaginative thinking in youth from odd beliefs or bizarre fantasies and preoccupations, emphasizing the need for age‐adapted and elaborate, detailed phenomenological assessment of the content, duration and frequency of these experiences and their impact on functioning. The second case demonstrated the progression of psychotic features from adolescence into early adulthood and the challenge of judging whether these features crossed the threshold to psychosis.

Conclusion

Both cases underscore the need for clinical training regarding differential diagnostics and management of psychotic features when present in early manifestations of BPD. This gap in care presents a missed opportunity for inclusion of patients in targeted early intervention programs for first‐episode psychosis. We propose the implementation of a continuous monitoring strategy in BPD with psychotic features.

1. Introduction

The concept of psychosis lacks a singular, operationalised definition in both the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM‐5‐TR; APA ²⁰²²) and the International Classification of Diseases, 11th Revision (ICD‐11; WHO ²⁰¹⁹) that transcends diagnostic categories. Rather, it serves as a descriptive term applied to a range of phenomena involving reality distortion or loss of intersubjective contact with reality (Parnas et al. 2010). The DSM‐5‐TR glossary defines psychotic features as including delusions, hallucinations and formal thought disorder, which are considered central clinical markers for person‐centred early intervention strategies. Schizophrenia is characterised by a combination of positive symptoms (e.g., delusions, hallucinations, disorganised speech and/or behaviours) and negative symptoms (e.g., diminished emotional expression or avolition) and psychotic features fall under the category of positive symptoms (APA ²⁰²²; Tandon et al. ²⁰¹³). The DSM‐5‐TR also includes Attenuated Psychosis Syndrome in Section III as a condition for further study. This syndrome includes psychosis‐like symptoms that warrant clinical attention but fall below the threshold for a full psychotic disorder. Reality distortions concomitant with borderline personality disorder (BPD) are explicitly listed as a differential diagnosis for attenuated psychosis, and a BPD diagnosis should be used when the overall clinical characteristics are better accounted for by this disorder (APA ²⁰²²).

Several different kinds of concepts have been used to elucidate phenomenological experiences at the border of psychotic disorders, that is psychotic features, micro‐psychotic, psychotic‐like, brief psychotic experiences, near‐psychotic phenomena, etc. In the DSM‐IV and DSM‐5‐TR, transient stress‐related paranoid ideation or severe dissociative symptoms is one of the nine diagnostic criteria for BPD (APA ²⁰²²). The psychotic features in BPD occur during periods of extreme stress or emotional dysregulation. These experiences are often reactive and tied to specific triggers, such as interpersonal conflicts or perceived abandonment. As the stress abates, people with BPD can generally correct stress‐related psychotic distortion of reality (Simonsen and Newton‐Howes ²⁰¹⁸), indicating that these phenomena are ego‐dystonic and unstable.

However, following the addition of this ninth criterion in the DSM‐IV, clinicians have faced significant challenges in differentiating psychotic symptoms and personality pathology in a clearly defined and consensual manner (Leclerc et al. 2024; Zandersen et al. ²⁰¹⁹), especially in cases with severe personality disorder (PD; Luyten and Fonagy ²⁰²²; Tyrer et al. ²⁰²²). One significant challenge for clinicians is determining when psychotic features reach the threshold for psychotic disorder. Adding to the complexity, psychotic features also occur in other mental disorders, including depression, bipolar disorder and posttraumatic stress disorder, all of which commonly co‐occur with BPD. Furthermore, dissociative symptoms, such as depersonalisation and derealisation, are particularly sustained in individuals with BPD (Gunderson et al. 1975). Some studies suggest that these dissociative symptoms may be a corollary of trauma, given the high incidence of traumatic experiences among individuals with BPD (Fung et al. 2023; Porter et al. ²⁰²⁰). However, trauma is also prevalent in schizophrenia (Chen et al. 2024) and recent evidence links dissociative phenomena to hallucinations and other positive psychotic symptoms across diagnoses (Chanen et al. 2024; Longden et al. ²⁰²⁰). Clinically, this makes the dissociative symptoms particularly challenging to interpret, as they may emerge both from trauma and from the underlying personality disorder or psychotic disorder.

In summary, these complexities contribute to the ambiguity in differentiating psychotic features in BPD from schizophrenia and other disorders involving psychotic symptoms or psychotic features. Although such features may simply meet criterion 9 of BPD, they could alternatively signal: (a) a prodromal phase of psychotic disorder, (b) the presence of a co‐morbid or primary psychotic disorder, or (c) transient psychotic experiences commonly observed in adolescence. In the following section, we will explore these possibilities in further depth.

1.1. Fluctuations Within the Same Underlying Disorder or Distinct Disorders That May Co‐Occur?

Early empirical research found that the presence of psychotic features provided useful information in defining the borderline syndrome (Gunderson and Kolb 1978). Initially, the disorder was conceptualised as “borderline schizophrenia” but in the DSM‐III, this was reclassified as schizotypal personality disorder (SPD) (APA ¹⁹⁸⁰; Perry and Klerman ¹⁹⁸⁰; Spitzer et al. ¹⁹⁷⁹). The introduction of the DSM‐III marked a shift in the diagnostic process from prototypical narrative descriptions (as seen in DSM‐II) of the various disorders to a polythetic approach focusing on a list of symptoms with equal weight. Furthermore, it entailed a drift away from considerations of the profound disturbances of the personality in schizophrenia (Simonsen and Newton‐Howes ²⁰¹⁸) that were found in the earliest descriptions of schizophrenia (e.g., Kraepelin ¹⁸⁸³), where personality traits were considered fundamental as premorbid predispositions or vulnerabilities that influenced the course of the illness (Keshavan et al. 2005; Simonsen et al. ²⁰⁰⁸; Simonsen and Newton‐Howes ²⁰¹⁸).

Discrete overlaps in symptoms do exist between BPD, SPD and schizophrenia, as these disorders all involve symptoms that exhibit some intermittent loss of reality. Additionally, they all encompass personality dysfunction or changes in personality (Jaspers 1963) as well as emotional disturbances. Specifically, BPD is characterised by emotion dysregulation, schizophrenia by diminished emotional expression or avolition and SPD by inappropriate or constricted emotional experiences (APA ²⁰²²). Each disorder also involves impairments in social and general functioning, with unstable relationships being characteristic of BPD, lack of close friends and confidants, social anxiety and odd behaviour for SPD and significant social withdrawal for schizophrenia (Fulford and Holt 2023). The schizoid, avoidant and schizotypal personality traits have been considered the typical “bridge” between personality and psychotic symptoms, due to evidence on familial psychiatric history that suggests a common genetic diathesis (Leclerc et al. 2024). For example, in a Danish register study of 2539 participants, 33.1% with SPD (and 58.2% of patients with comorbid SPD and cannabis use disorder) developed schizophrenia over a period of 20 years (Hjorthøj et al. 2018), underscoring the vulnerability of certain forms of personality pathology for the development of severe mental disorders.

The potential overlaps of symptoms between BPD, SPD and schizophrenia complicate the demonstration of independent diagnoses in research and clinical settings. Consequently, BPD and schizophrenia have long been considered incompatible (Boberg et al. 2022), with schizophrenia taking precedence due to its well‐established evidence base in psychopharmacology (Simonsen and Newton‐Howes ²⁰¹⁸). The DSM‐IV from 1994 allowed for greater flexibility in diagnosing co‐occurring conditions in the presence of symptom overlap. In contrast, while the DSM‐5 includes an exclusion criterion stating that the general PD criterion “is not better explained as a manifestation or consequence of another mental disorder” (APA ²⁰²²), it still recognizes the complexity of comorbidity. However, the ICD‐10 presents a slightly different formulation of the general PD criteria, stating that the symptoms “are not better explained” by another mental disorder (WHO ²⁰⁰⁴). This reflects a more traditional hierarchical approach to psychiatric classification, emphasizing distinct categories.

When one accepts the broader non‐hierarchical approach to diagnoses, some studies indicate relatively high rates of comorbidity. For instance, studies indicate that 38% of people with BPD meet criteria for a comorbid diagnosis of psychotic disorder (Slotema et al. 2018), whereas 17%–25% with schizophrenia or first‐episode psychosis (FEP) also meet criteria for BPD (Bahorik and Eack 2010; Francey et al. ²⁰¹⁸; Schandrin et al. ²⁰²³). Recent evidence has shown that auditory hallucinations and delusions are prevalent in BPD (26%–59%) (D'Agostino et al. ²⁰¹⁹; Niemantsverdriet et al. ^{2017, 2022}). Importantly, auditory hallucinations are reported to be similar in BPD and schizophrenia regarding most phenomenological characteristics (Merrett et al. 2016).

Given the high levels of comorbidity, Meisner et al. (2024) and others have suggested using the term “co‐occurrence” instead of “comorbidity”. This terminology reflects the idea that multiple diagnoses may not signify separate disorders per se but could instead reflect fluctuations within the same underlying disorder. The ICD‐11 further supports this notion by using “co‐occurrence” when two or more disorders are present (WHO ²⁰¹⁹). The findings of high comorbidity or co‐occurrence have blurred the traditional dichotomy between psychosis and neurosis. Furthermore, by adding the ninth criterion to the BPD diagnosis in the DSM‐IV in 1993, the historical link to schizophrenia was reopened, despite keeping the SPD diagnosis and by that again further complicating the accuracy of clinical diagnostics (Simonsen and Paris 2025). This notion of fluctuations within the same underlying disorder, however, aligns well with recently proposed models of personality pathology, which argue that BPD lies on the severity pathway between internalising/externalising and psychotic psychopathology (Sharp and Wall 2018).

1.2. Psychotic Features in Early‐Stage Psychopathology: A Developmental Perspective

In recent years, there has been notable progress in the development of clinical staging models that address the developmental and transdiagnostic aspects of psychopathology. Clinical staging models offer a framework for understanding how psychopathology develops and progresses over time (Chanen et al. 2016; Mei et al. ²⁰¹⁹). They enable clinicians to assess and intervene with tailored interventions at different stages of disorder, aiming to prevent or mitigate the severity and functional impairment associated with mental disorders such as BPD and schizophrenia.

Adolescence marks a critical period for the early emergence of severe mental disorders such as BPD and psychotic disorders. Although clinical staging models and early intervention strategies are gaining traction, the complexity of distinguishing between transient psychotic‐like experiences that are common in this age group and early signs of schizophrenia in young individuals poses ongoing challenges. In a cohort of 984 youths, as many as 17.2% reported one or more psychotic experiences (Rimvall et al. 2024). The transient nature of these features in young people, combined with the stigma associated with psychotic disorders and the potential adverse effects of anti‐psychotic medications, has led to a cautious approach in diagnosing and treating psychotic disorders in young people. This cautious approach, which can be particularly prevalent in child and adolescent psychiatry (Bartlett 2014), can result in extended periods of undetected and untreated psychosis, despite strong support for early intervention for FEP (Correll et al. 2018; Hegelstad et al. ²⁰¹²).

Despite emerging evidence on the progression and impact of psychotic features in adolescents with BPD, there remains a significant gap in our understanding of the developmental trajectories and clinical implications of these features. In studies of adolescents with BPD, psychotic features have been found to be prevalent, adding another layer of complexity to the diagnostic process. Determining when these psychotic features reach the threshold for psychotic disorders is particularly challenging in the early stages of both disorders, as the phenotypes are still evolving and transient psychotic‐like experiences are common during this developmental period. Adolescents with BPD describe (auditory) hallucinations that may be difficult to differentiate from those in schizophrenia regarding physical (frequency, duration, location and loudness), cognitive (beliefs regarding the origin of voices, disruption to life and controllability) and emotional (negative content and distress) characteristics (Cavelti et al. 2019). Delusions may, however, be less severe but more hostile and negative symptoms and symptoms of formal thought disorder appear less severe in young people with BPD compared to schizophrenia (Thompson et al. 2019). Co‐occurrence of these symptoms in young people who also meet criteria for BPD is, similar to findings among adults with BPD, related to significantly higher levels of psychopathology (Chanen et al. 2024; Cavelti et al. ²⁰¹⁹; Niemantsverdriet et al. ²⁰²²; Schandrin et al. ²⁰²³; Slotema et al. ²⁰¹⁸; Thompson et al. ²⁰¹⁹). Accordingly, some have argued that psychotic features might better be understood as a transdiagnostic risk marker for severe psychopathology and impairment (Hartmann et al. 2021; Kaess and Cavelti ²⁰²²).

Studies that refine our understanding regarding the relationship between these two disorders in their early stages, and the developmental trajectory from BPD to schizophrenia will help inform more targeted and preventative interventions for individuals with symptoms of both disorders. Treatment approaches for BPD and schizophrenia differ significantly: first‐line treatment for BPD is psychotherapy (Storebø et al. 2020), with limited evidence for pharmacological treatments (Stoffers‐Winterling et al. ²⁰²²), whereas schizophrenia typically requires pharmacological treatment (McDonagh et al. ²⁰¹⁷). Accurate diagnosis is crucial not only to ensure that individuals receive appropriate and effective treatment, avoiding potentially harmful or ineffective interventions, but also to target the underlying phenotype early and prevent the progression of the disorder.

The variability in symptom persistence and the differential responses to treatment underscore the need for more nuanced research and critical considerations of conceptual differences and nuances in diagnostic systems. Studies that explore the intricacies of clinical assessment, diagnostics, and treatment planning for adolescents presenting with symptoms of BPD and psychotic features are crucial, particularly studies that track the progression of these features in BPD from adolescence into early adulthood. Prospective studies that examine early signs of personality pathology and later development of schizophrenia are, however, sparse, but the limited existing evidence suggests that there is a probable association with increased risk (Newton‐Howes et al. ²⁰¹⁵).

In a recent 5‐year longitudinal study of 111 adolescents with BPD in adolescence, we found that 16% had developed schizophrenia 5 years later (Jørgensen et al. 2024). At baseline, 89% of the sample reported transient stress‐related paranoid ideation or severe dissociative symptoms, but after 5 years, these symptoms only persisted among 28% of the participants. The psychotic features thus remitted in most of the participants but remained in a subgroup of participants, some of whom experienced persistent and pervasive psychotic symptoms indicative of schizophrenia spectrum disorder, some of whom had gone undetected. In the current study, we selected two such cases of adolescents with BPD from the M‐GAB randomised clinical trial (RCT, Beck et al. ²⁰²⁰; ClinicalTrials.gov identifier: NCT02068326) who also attended the 5‐year follow‐up study to gain a richer understanding of the complexities involved in diagnosing and treating psychotic features in young people with early‐stage BPD.

1.3. Aims

We aim to explore the complexities and challenges with psychotic features in patients diagnosed with early‐stage BPD, presenting two cases to illustrate fluctuations in and progression of symptoms and different developmental pathways for psychopathology. Additionally, we will consider the implications for recurrent assessment and strategies as part of treatment. Based on these cases, we will suggest an optimised approach to suitable monitoring and treatment for these young people. This approach is particularly relevant to facilitate the transition into adult psychiatry, which typically operates with diagnosis‐specific assessment and treatment and lacks the capacity for extensive re‐assessment of psychopathology.

2. Materials and Methods

2.1. Case‐Study Design

In the current study, we have used what Stake (1995) refers to as an instrumental case study design, that is where cases are used to gain a broader appreciation of an issue or phenomenon. The cases were selected to answer our research questions following the CARE recommendations for case reports (Gagnier et al. 2013). Please find the CARE checklist in the Supporting Information. Theory‐driven approaches to case studies can help generate knowledge that is potentially transferable to a range of clinical contexts and accordingly apply to real‐life situations in the clinic and offer practical guidance for clinicians and students (Crowe et al. 2011).

2.2. Participants

In the original study, 111 adolescents with BPD were included at baseline. This cohort was then followed over 5 years (cf. Figure 1). In the first year after inclusion, five participants were diagnosed with a psychotic disorder. In the second year, another three participants were diagnosed with a psychotic disorder. By the 5‐year follow‐up, a total of 15 participants fulfilled criteria for a psychotic disorder. From the 5‐year follow‐up cohort, we identified two cases that met criteria for a schizophrenia spectrum disorder at follow‐up, which had either not yet developed or not been detected at baseline and had not been identified in routine care during the follow‐up period. The first case involved a participant who likely had a psychotic disorder at baseline, highlighting the challenges of assessing psychotic symptoms in adolescence when they co‐occur with BPD symptoms. The second case focused on a participant who exhibited psychotic features at baseline and progressively developed full psychotic symptoms, providing insight into the developmental trajectory of these symptoms over time. For the present study, we will focus on assessments carried out at baseline and after 5 years.

FIGURE 1.

Prevalence of psychotic disorders over 5 years.

2.3. Assessments and the Diagnostic Process

The two cases were referred from two child and adolescent outpatient clinics within the mental health services of Region Zealand, Denmark. One of the exclusion criteria for participating in the study included “current psychosis, diagnosis of schizophrenia or schizotypal personality disorder” (Beck et al. 2020). Before referral, they were screened for eligibility during their psychiatric reviews to ensure they did not meet any exclusion criteria. Following referral, participants underwent baseline assessment to double‐check eligibility. The following assessment instruments were used at baseline and after 5 years (and are detailed in previous publications on the sample e.g., Beck et al. ²⁰²⁰; Jørgensen et al. ²⁰²⁴).

2.3.1. Baseline

Diagnostic interviews on PDs were carried out by a trained clinical psychologist, and mental disorders were assessed by research assistants who were psychologists. 10% of the interviews were selected for inter‐rater reliability analyses. BPD was assessed by use of the Childhood Interview for DSM‐IV Borderline Personality Disorder (CI‐BPD; Zanarini ²⁰⁰³) and the Zanarini Rating Scale for Borderline Personality Disorder (ZAN‐BPD; Zanarini and Frankenburg ²⁰⁰¹). PDs were assessed using the Structured Clinical Interview for DSM‐IV‐Axis II (SCID‐II; First et al. ¹⁹⁹⁷). The SCID‐II assesses the presence of DSM‐IV PDs. Mental disorders were assessed with the Mini‐International Neuropsychiatric Interview for Children and Adolescents (MINI‐KID 6.0; Sheehan et al. ¹⁹⁹⁸). The MINI‐KID is a structured diagnostic interview for children aged 6–17 years based on 24 DSM‐IV and ICD‐10 criteria. Once all the assessments had been performed, we collated the information along with information from the participants' medical journals to rate their level of psychosocial functioning using the Children's Global Assessment Scale (CGAS; Shaffer et al. ¹⁹⁸³). Interrater reliability (IRR) for the CI‐BPD, the ZAN‐BPD, the MINI‐KID and the CGAS was excellent in the M‐GAB trial (Beck et al. 2020). IRR for the SCID‐II interview was found to be moderate in the M‐GAB trial (Beck et al. 2020).

2.3.2. 5‐Year Follow‐Up

At the 5‐year follow‐up, all diagnostic assessments were carried out by trained clinical psychologists. Mental disorders were assessed with the WHO diagnostic interview Schedules for Clinical Assessment in Neuropsychiatry (SCAN; WHO ¹⁹⁹⁴). PDs were assessed with the Structured Clinical Interview for DSM‐5 Personality Disorders (SCID‐5‐PD; First et al. ²⁰¹⁶). General functioning was assessed with the Work and Social Adjustment Scale (WSAS) (Mundt et al. 2002), which is a self‐report measure. A WSAS score below 10 indicates low functional impairment, 10–19 moderate functional impairment, and 20–40 severe functional impairment. After the participants were assessed by semi‐structured interviews, the first author (M.S.J., clinical psychologist) and the last author (E.S., chief physician) carefully went through each participant's data and made diagnoses according to a consensus process. All interviews were audio‐taped and 10% were randomly chosen to assess for IRR by a clinical psychologist. For the SCAN interview, IRR showed excellent agreement. For the SCID‐5‐PD, IRR showed moderate agreement, but almost perfect agreement with regards to the BPD criteria on the SCID‐5‐PD. Internal consistency on the WSAS was in the good to excellent range.

3. Case Illustrations

We refer to the clients for these cases as Sienna and Marlene. We have modified important details of the case reports, such as name, age and other identifying information, to protect their confidentiality while preserving the illustrative value of the cases.

3.1. Case 1, Sienna

Sienna is a 21‐year‐old woman currently studying at the University. She was referred to child and adolescent mental health services at the age of 15. Upon entry, she was diagnosed with BPD and referred to the M‐GAB trial, where she was assessed with the baseline diagnostic interviews described above. On the CI‐BPD, she met all nine diagnostic criteria for BPD (i.e., intense anger, affective instability, chronic feelings of emptiness, fear of abandonment, identity disturbance, unstable and intense relationships, transient stress‐related paranoia ideation or severe dissociative symptoms, impulsivity and self‐harm). Her BPD severity was high, with a total score of 26 out of a possible 36 on the ZAN‐BPD. On the SCID‐II, she also met diagnostic criteria for paranoid PD and on the MINI‐KID, she met criteria for depression and social phobia. Her level of social functioning on the CGAS was 34, indicating major impairment in functioning, but in line with the total sample mean of 35.5 (Beck et al. 2020).

As a part of the baseline assessment, Sienna was interviewed about psychotic symptoms using the MINI‐KID. She reported childhood fears such as believing dog walkers were child abductors and feeling that someone was following her, although she could dismiss these worries when she saw no one was there. She said that at ages 10–11, she believed that her mom could read her thoughts and that her father could read her mother's thoughts, causing her to avoid thinking about certain things. At the time of the baseline assessment, she no longer held these beliefs. She also said that she once believed she was “the chosen one” among a group of friends, with a mission to save Santa Claus and protect the world from perceived dangerous and paedophilic teachers. Additionally, she experienced epiphanies, such as looking at a tree and feeling like it was a sign to “do something”. She said that her friends had commented on her unusual relationship with her teddy bears, but she was not asked to elaborate on that relationship. She denied having such experiences at the time of assessment as well as she denied having ever experienced hallucinations. Furthermore, she said that the experiences were not dependent on the presence of depressive symptoms. At intake, these symptoms were interpreted as transient and stress‐related, consistent with her BPD diagnosis. Sienna was then offered a 1‐year treatment with mentalization‐based group treatment (MBT‐G) and treated with anti‐depressants for some months until her symptoms of depression improved. Although Sienna felt that she had profited from MBT‐G, she was referred to adult mental health services at age 18 with a diagnosis of BPD. She recalled not being thoroughly reexamined upon referral, with the assessment focusing on BPD. Despite informing a psychologist and a psychiatrist about her symptoms, they were again deemed transient and stress‐related. Sienna then received another 3 years of weekly MBT (combined individual and group treatment) in adult psychiatry without follow‐up on the progression of psychotic symptoms. Sienna ended her treatment program for BPD 6 months prior to the 5‐year follow‐up assessment.

As part of the M‐GAB 5‐year follow‐up study, Sienna was reassessed to evaluate her current psychopathological symptoms. When asked about psychotic symptoms, Sienna reported experiencing several psychotic symptoms throughout her life. She recalled an epiphany in the 4th grade where she believed she was chosen by a God named Moon to singlehandedly save Santa Claus, and there was a war involving elves. She worked on writing a Bible continuously through childhood and adolescence. She was convinced that the teachers at school were in disguise. In reality, they were Demons trying to obstruct her divine cause. Sienna received these divine messages through her teddy bears, and she talked to them every day at home, but also telepathically when she was not home until the age of 19. From her teenage years until she was 17, Sienna became increasingly convinced that her close friends conspired against her and were out to get her. She was convinced people around her were surveilling her through cameras and that they were waiting for her outside her home. In the same period, she was convinced that her father could read her thoughts, prompting her to avoid thinking certain things. She also had instances where her close ones suddenly sounded demonic, and their faces changed to demonic faces. When she was 17, she started taking 25 mg Quetiapine (off‐label) for unrest and persistent sleep problems, and this medication had—and still has—relieved her of many of the psychotic symptoms mentioned above.

Sienna tells the assessor (first author) that she has improved over the years, but she is still very invalidated by symptoms of severe anxiety, but she is uncertain what she is afraid of. Although she is enrolled at the university, she rarely attends classes, because she feels like she is under surveillance while she is in class and in public places like train stations and supermarkets, leading her to stay home most of the time. Sienna is not currently in any psychosocial treatment. She has tried to discontinue the Quetiapine several times, but whenever she tries, her anxiety spirals upwards, she becomes aggressive and gets into conflicts, and she starts to see familiar people's faces turn into demonic faces again. She occasionally feels unreal or that things around her turn blurry, but is uncertain whether these symptoms only occur in situations where she is distressed. Occasionally, she sees yellow dots around her but does not know if it is related to migraine. Furthermore, she tells the assessor that her mind is compartmentalised into two spaces; in one room, there are random thoughts and not in the other, but the thoughts in the two spaces can influence each other, and sometimes her thoughts block for a while. Whilst she feels like her symptoms of BPD have improved over the years, she still meets all criteria for BPD as measured with the SCID‐5‐PD apart from the emptiness criterion. Sienna also meets the criteria for a DSM‐5 schizophrenia spectrum disorder. Her level of functioning on the WSAS is 30, which is equivalent to severe functional impairment (Weekers et al. 2023).

3.2. Case 2, Marlene

Marlene is a 20‐year‐old woman who was referred to the M‐GAB trial at the age of 14. Marlene lives by herself, did not complete primary school and is currently not in education, employment, or training. Her interests include visiting family and a few friends, but maintaining social relationships has become increasingly difficult as her peers have moved on with education and jobs. Marlene is currently on psychotropic medication, taking 100 mg Quetiapine (off‐label) and 60 mg Fluoxetine.

At the time of referral to M‐GAB, Marlene was only taking off‐label anti‐psychotic medication. At baseline, she met all nine diagnostic criteria for BPD on the CI‐BPD, and the severity of BPD was at a midrange level (total score of 14 out of 36 on the ZAN‐BPD). She did not meet criteria for any other PDs other than BPD. On the MINI‐KID, she met criteria for depression, panic disorder, agoraphobia, social phobia and obsessive‐compulsive disorder. Marlene's level of functioning on the CGAS was 31, indicating major impairment in functioning, similar to the total sample mean of 35.5 (Beck et al. 2020). On the MINI‐KID, Marlene was interviewed about psychotic symptoms and reported that she felt like she had so many thoughts that her friends must be able to see them and know what she was thinking about. Marlene was not asked to elaborate on this. On the ZAN‐BPD, Marlene described that her surroundings sometimes felt unreal and her sense of reality became blurred when she was emotionally numb. Conversely, being overwhelmed by intense feelings also made it difficult for her to distinguish what was real. These symptoms were considered transient and stress‐related, consistent with her BPD diagnosis. Marlene was then offered a 1‐year treatment program with MBT‐G, but dropped out midway, feeling it was not helpful.

At the 5‐year follow‐up, Marlene was re‐assessed with SCAN and the SCID‐5‐PD. She describes that she has a basic, stable sense of who she is, but often is “putting on a mask” in front of others, to the extent that she finds it difficult to return to her true self. She experiences constant anxiety, with social and health anxiety being the most prominent. Marlene frequently visits the doctor with various suspicions of somatic disorders. Moreover, she has a lot of urges to hurt herself or others, such as thinking “I need to put my hands and feet on the hotplate”, “I could cut off my finger”, or “it would be easy to stab my sister's throat” when holding a knife. Yet, she has not self‐harmed for over 3 years and does not have suicidal urges. She describes that she has been under a “constant grey cloud” for the past 2 weeks, with increased sleep and low energy. She experiences intense and rapidly shifting emotions, from feeling everything is “pink and unicorn‐like” to extreme lows where she feels sad and empty. Marlene refers to periods with elevated mood as “glimmer flips” and says that she feels like “Einstein” during these episodes, but they only last a maximum of 5 min and are infrequent.

Regarding the assessment of psychotic symptoms, Marlene describes feelings of unreality where she questions if she really exists. She says, “It is as if I am unconscious, sometimes I sit and stare at nothing for hours, not knowing what I am doing”. She also describes experiencing time slowing down, cars driving slower and songs playing at a lower speed. Occasionally, she feels alienated from her own body, attributing it to low self‐esteem. When feeling bad, she becomes what she describes as “hypersensitive” and an “apparatus that is turned off and on”, most often off when alone. She also says that she feels “overstimulated” when she is around people, needing to use earphones to block out stimuli and denies that it is related to anxiety. A couple of years ago, she saw her teeth as yellow with holes and transparent, which persisted for over a year despite several dentists assuring her that they were normal. Marlene did not smile in front of others and insisted that the light be switched off when she was in a room with others (so they did not see her teeth). In that same period, her senses were “off”. For instance, a friend touched her arm, but she felt it in her toes. Lights and sounds were also blurred, and she felt restless. Currently, she describes that her thoughts are very vivid, saying she sees them in front of her or hears them as if they are spoken or sees them as written in text in front of her eyes. She says that she can hear her own thoughts loudly, and she thinks people around her must hear them or at least know that something is going on in her brain. She also hears her name being called both inside and outside her head, and sometimes needs to remove her headphones to check if someone is calling her. At times, her thoughts are also repeated like an echo or are absent for a while.

Marlene says that she has always experienced these symptoms and that she told clinicians at the child and adolescent mental health services about them, but no psychotic disorder was suspected at the time. At the 5‐year follow‐up, Marlene meets criteria for a schizophrenia spectrum disorder. She also continues to meet diagnostic criteria for BPD, exhibiting five BPD criteria (fear of abandonment, impulsivity, affective instability, emptiness, transient stress‐related paranoid ideation or severe dissociative symptoms) as well as two at a subthreshold level (identity disturbance and inappropriate anger). She does not meet criteria for any other PDs. Additionally, she meets criteria for obsessive‐compulsive disorder, social phobia and mild depression. With a WSAS score of 21, her level of functioning is severely impaired.

4. Discussion

This study presented two adolescent cases with BPD followed over 5 years, focusing on the diagnostic challenge of identifying psychosis when co‐occurring with BPD. Both BPD and FEP typically emerge as evolving phenotypes in adolescence (Cavelti et al. 2021), and early differentiation between transient, stress‐related psychotic‐like symptoms and emerging psychotic disorder remains difficult. Although conceptually distinct, these syndromes can overlap in phenomenology, especially in early stages.

From a clinical staging perspective, psychotic symptoms often present in attenuated forms before progressing into more persistent and impairing experiences (Mei et al. 2019). The p‐factor model supports a dimensional understanding of psychopathology, placing BPD and psychosis along a continuum of severity (Caspi and Moffitt 2018). For some individuals, BPD may increase vulnerability to developing schizophrenia spectrum disorders (e.g., Ryan et al. ²⁰¹⁷). However, longitudinal data examining the developmental trajectories from BPD to psychosis remain scarce.

Both cases describe symptoms that could indicate the kind of multiple fluctuating symptoms seen in SPD or early‐stage schizophrenia, that is self‐reference, unformed hallucinations, dissociative symptoms such as derealisation and depersonalisation, fear of bodily contact and loss of self‐demarcation or basic self. The first case (Sienna) illustrates the challenges of disentangling psychotic phenomena from normal magical and imaginative thinking often seen in youth. Additionally, it highlights the difficulty in differentiating these phenomena from odd beliefs, bizarre fantasies, and preoccupations. Finally, it raises the challenge of determining whether these symptoms are transient or above the threshold to warrant a diagnosis and treatment of a psychotic disorder. These challenges are amplified by the fact that many young people in the early stages of psychotic disorder often deny or attempt to conceal their psychotic experiences from others (Corcoran et al. 2007). In the case of Sienna, the presence of multiple symptoms that could indicate psychosis should have led to a more thorough assessment of psychotic disorder using comprehensive assessment tools. The second case (Marlene) demonstrates the progression of psychotic features over the adolescent period into early adulthood. In her case, the symptoms were considered transient and dissociative at baseline, but over time evolved into symptoms that might require diagnosis‐specific treatment for FEP. Both cases met all nine BPD criteria at baseline, supporting the finding that individuals with BPD who also experience psychotic features may represent a particularly severe subset of BPD (Cavelti et al. 2019; Chanen et al. ²⁰²⁴; Niemantsverdriet et al. ²⁰²²; Schandrin et al. ²⁰²³; Slotema et al. ²⁰¹⁸; Thompson et al. ²⁰¹⁹). This is further supported by the severe functional impairment observed in both cases. WSAS scores in both cases were higher than average for the broader sample, supporting this interpretation.

A core clinical dilemma is how to approach the co‐occurrence of BPD and psychotic features. Three strategies could be considered: (1) a hierarchical approach, prioritizing treatment of psychotic disorders, (2) simultaneous diagnosis and treatment of both disorders, and (3) a pragmatic monitoring strategy while in treatment for BPD. The traditional hierarchical model risks neglecting personality pathology. In some countries, adolescents with subthreshold psychotic symptoms are referred to Clinical High‐Risk (CHR) or FEP services. However, CHR and FEP services are not routinely designed to assess or treat co‐occurring personality pathology, despite growing evidence of overlap between CHR and BPD populations (Ryan et al. 2017; West et al. ²⁰²¹). As a result, youth may receive fragmented care: those referred to CHR or FEP services may not receive evidence‐based BPD treatment, and those referred to BPD services may not be systematically assessed or followed for emerging psychotic symptoms.

Simultaneous diagnosis demands integrated care models that are often unavailable. RCTs that examine treatment for co‐occurring BPD and psychotic features are currently lacking. At the moment, we do not have evidence‐based treatment programs for people with symptoms of both disorders, and in clinical practice, the limited evidence hinders appropriate clinical decision‐making (Chanen et al. 2024). This may result in prolonging the period of untreated psychosis (Cavelti et al. 2021; Francey et al. ²⁰¹⁸), or prescription of off‐label anti‐psychotic medication despite limited evidence. For instance, 15% of the adolescents in the M‐GAB trial were on off‐label anti‐psychotic medication at baseline even though psychotic disorder was an exclusion criterion for participation. Gleeson et al. (2012) showed in a pilot study that youth with co‐occurring FEP and BPD appeared to benefit more from integrated treatment than from specialist FEP treatment alone. Therefore, a clear take‐home message from this study is the importance of developing integrated, developmentally sensitive treatment pathways for youth presenting with overlapping symptoms of BPD and emerging psychosis.

A pragmatic approach, monitoring psychotic features over time, may be the most feasible. Our cases support this third strategy, highlighting the risk of missed or delayed psychotic disorder diagnosis when BPD is the initial focus.

Importantly, the diagnosis‐specific structure of adult psychiatry means that when patients are referred from child and adolescent mental health services on a specific diagnosis, they will most often only be thoroughly assessed for that specific mental disorder upon entry into adult services, which was also the case for Sienna and Marlene. This means that psychotic symptoms may remain undetected and untreated, which is associated with serious consequences for the affected individual, including enduring symptoms, impaired functioning, long‐term disability, social isolation, poor physical health and possible harm to self or others (Drake et al. 2020).

5. Clinical Implications

A developmental, stage‐informed approach is essential. When adolescents with BPD report possible psychotic features, they should undergo comprehensive assessment, ideally supplemented with phenomenological assessments that follow patients' narratives more closely (Nordgaard et al. 2021; Parnas et al. ²⁰⁰⁵). If symptoms do not meet full diagnostic criteria for psychotic disorder, structured monitoring should continue across service transitions. Upon referral to adult psychiatry, early psychotic phenomena must be re‐evaluated alongside personality pathology to ensure accurate diagnosis and timely treatment.

Furthermore, longitudinal cohort studies are needed to track outcomes in adolescents with BPD and early psychotic features. A dimensional, transdiagnostic perspective could clarify whether such features are temporary, prodromal, or indicative of a separate disorder. Tailored, integrated care models must be developed for this complex clinical population.

6. Limitations

Case studies offer valuable insights into specific phenomena, but they also come with inherent limitations. The most significant limitation is the limited generalisability of the findings. Furthermore, case studies often rely on the researchers' subjective judgements, and we may have emphasised aspects of the cases that align with our preconceived notions. Additionally, there is a lack of clear definitions of the concept of psychosis and delineation of different concepts such as micro‐psychotic experiences, psychotic‐like experiences and psychotic features. This ambiguity can affect the interpretation and assessment of psychotic symptoms. The expertise of the raters in the psychiatric assessments varied between baseline and follow‐up. At baseline, the MINI interviews were conducted by recently graduated psychologists who were research assistants, and at follow‐up, the SCAN was administered by trained clinical psychologists with a PhD. Adding to that, the instrument differed at the two time points, which could influence the consistency and reliability of the assessments.

Another limitation concerns the conceptual framework. This study relied on categorical diagnostic criteria for BPD, whereas more recent conceptualisations emphasise dimensional models of personality pathology, such as the Alternative Model for Personality Disorders (AMPD) in DSM‐5 Section III. Dimensional models may better reflect underlying traits and symptom domains that cut across traditional categorical boundaries, including those separating personality pathology and psychotic features. These frameworks may offer a more nuanced understanding of the overlap between BPD and psychotic phenomena (Kotov et al. 2020; Longenecker et al. ²⁰²⁰; Phalen et al. ²⁰²²). Future research will benefit from incorporating such dimensional approaches.

Additionally, the research interviews did not include reports and observations from ongoing clinical or therapeutic work, and as such, we had limited access to detailed accounts of real‐time relational dynamics that could help clear up whether the psychotic features were transient and stress‐related or more stable mental states.

Another potential limitation of the study is the possibility of confirmation bias. Since the M‐GAB trial specifically targeted adolescents with BPD and excluded those with psychotic disorders, there is a risk that psychotic symptoms may have been overlooked or undetected due to this focus, which may have conflated the challenges in differentiating the two disorders. Our cases were selected from Danish mental health services, where the ICD‐10 is applied, but they were assessed using DSM‐5 criteria. The study therefore has some practical limitations since there are differences in how diagnostic criteria for schizophrenia are defined in ICD‐10 and DSM‐5, as well as how subtypes of schizophrenia are addressed. Furthermore, the ICD‐10 follows a more hierarchical understanding of psychiatric classification, whereas the DSM‐5 adopts a more horizontal approach.

7. Conclusion

This case study highlighted significant challenges in the assessment, differential diagnostics and follow‐up of psychotic features in two young women with BPD in the transition period from adolescence into early adulthood. The cases illuminated the intricate task of distinguishing between normative developmental experiences, psychotic features and potential indicators of emerging psychosis. The failure to address this gap in care represents a missed opportunity for timely and targeted early intervention. Therefore, we advocate for the integration of continuous monitoring strategies tailored to young people with BPD who exhibit psychotic features.

Ethics Statement

The RCT from which the two cases were derived was approved by the Regional Ethics Committee of Region Zealand (no: SJ‐371). No additional ethical approval was required for the 5‐year follow‐up study including this case study due to local guidelines.

Consent

Written informed consent was obtained from the participants for publication of details from their assessments. Personal identifiers, such as age and other potentially identifiable information, have been anonymised or altered.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: eip70103‐sup‐0001‐supinfo.docx.

Jørgensen, M. S. , Austin S. F., Storebø O. J., et al. 2025. “Borderline Personality Disorder or First‐Episode Psychosis? Challenges in Assessing Psychotic Features in Early Intervention: A Case Report.” Early Intervention in Psychiatry 19, no. 10: e70103. 10.1111/eip.70103.

Data Availability Statement

The data supporting the findings of this case study are contained within the article. Due to the nature of case studies, confidential information from an individual participant cannot be shared publicly to ensure privacy and ethical compliance. Further enquiries can be directed to the corresponding author.