Addressing Schemas, Modes and Interoceptive Disconnection in Chronic Binge Eating Disorder: A Temperament‐Informed Schema Therapy Case Series Authors: Phoebe R Joshua, Susan Simpson, Vivienne Lewis, Sally F Kelty Published: 2026 ABSTRACT ABSTRACTBinge eating disorder (BED) is often chronic and accompanied by trauma, comorbidity and neurodiversity, for whom existing treatments can be less effective. This study examined a manualised, temperament‐informed schema therapy protocol (ST‐BED‐TI) targeting early maladaptive schemas (EMS), schema modes and interoception in adults with complex BED. A prospective case series was conducted with five adult women with long‐standing BED and high clinical complexity, who received between 25 and 35 sessions of ST‐BED‐TI in outpatient services in Australia and the United Kingdom. Individual change was evaluated with reliable change indices and clinical significance for global eating disorder (ED) symptom scores. This trial was pre‐registered on the Australian New Zealand Clinical Trials Register (ID: ACTRN12623001259639). Four of five participants showed reliable and clinically significant improvement in global ED symptoms at post‐treatment; three of the four measured at 6‐month follow‐up maintained these gains. All participants reported reductions in binge eating frequency. Most demonstrated reliable improvements in key EMS, schema modes and at least one interoceptive awareness domain, with one case showing marked interoceptive gains despite limited measurable EMS or mode change. Within the limits of a small uncontrolled case series, ST‐BED‐TI appears acceptable, clinically manageable and promising in this pilot context. These preliminary findings warrant larger controlled and mechanism‐focused studies.Keywords: binge eating disorder, interoception, neurodiversity, schema therapy, temperament Introduction Binge eating disorder (BED) is characterised by recurrent episodes of consuming unusually large amounts of food with a feeling of loss of control, without regular compensatory behaviours. People with BED experience substantial psychological distress, functional impairment and high rates of psychological and medical comorbidity (American Psychiatric Association 2022; Kessler et al. 2013; Udo and Grilo 2018). Within contemporary affect‐regulation and escape models, binge eating is seen as a negatively reinforced strategy for regulating overwhelming or aversive emotions, temporarily attenuating self‐awareness and negative affect through distraction, sensory soothing or numbing (Heatherton and Baumeister 1991; Leehr et al. 2015). Individuals with BED experience intensified emotional reactivity, difficulties identifying and managing negative affect and a tendency to use eating to cope with distress (Dingemans et al. 2017; Gianini et al. 2013; Lavender et al. 2015), emphasising the importance of interventions that target emotion regulation processes. Developmental Origins: Early Maladaptive Schemas (EMS) Schema theory supplies a developmental framework that conceptualises binge eating as a strategy for emotion regulation. EMS are enduring cognitive and emotional patterns that arise when core emotional needs in childhood (e.g., safety, secure attachment, autonomy, realistic limits and emotional validation) are not sufficiently met and subsequently shape how people see themselves, others and the world (Aloi et al. 2025; Cunningham et al. 2024). EMS are elevated across eating disorder (ED) diagnoses compared with both non‐clinical and other clinical populations (Aloi et al. 2025; Meneguzzo et al. 2021). EMS of Insufficient Self‐Control, Emotional Deprivation, Abandonment and Vulnerability to Harm are particularly prominent in bulimia nervosa (BN) and BED, in contrast to EMS such as Social Undesirability, Failure and Punitiveness, which are more characteristic of restrictive presentations (Basile et al. 2019; Pauwels et al. 2016; Rania et al. 2020). A recent network analysis indicated that central EMS in BED include Social Isolation, Abandonment, Subjugation and Negativity/Pessimism, highlighting the pervasive loneliness and emotional distress foundational to this disorder (Aloi et al. 2025). These EMS foster chronic shame, perceived defectiveness and expectations of rejection or loss, increasing the appeal of rapid, externally driven strategies such as binging to cope with distress (Aloi et al. 2025; Spirou et al. 2022). These associations do not establish that EMS cause binge eating in individual cases. EMS can also explain the link between childhood adversity and EDs. Mediation studies suggest that EMS such as Defectiveness, Failure and Negativity/Pessimism partially account for associations between early abuse or neglect and later ED psychopathology (Fasolato et al. 2024; Meneguzzo et al. 2021). Individuals with higher‐weight and binge eating presentations report more numerous and severe EMS than normal‐weight controls, alongside higher rates of childhood trauma and symptoms of post‐traumatic stress disorder (PTSD), suggesting that entrenched EMS and trauma‐related processes drive the chronicity and complexity that are often seen in BED (Spirou et al. 2022). Recent work indicates that PTSD and complex PTSD (C‐PTSD), typically rooted in prolonged childhood trauma, are highly prevalent in ED populations and negatively affect treatment progress and outcomes (Day et al. 2024; Day et al. 2026). Consequently, interventions that address EMS alongside emotion regulation may be relevant to some complex presentations, although comparative and longitudinal evidence is needed to determine whether such changes produce more durable reductions in binge episodes. Coping Responses: Emotional Avoidance and Schema Modes When EMS are activated by current interpersonal or situational stressors, individuals experience intense schema‐congruent affects (e.g., shame, fear and loneliness) that are hypothesised to originate in childhood experiences of unmet emotional needs. In effect, present situations are interpreted through the implicit lens of earlier experiences. The here‐and‐now expression of these EMS is termed schema modes: configurations of emotions, cognitions, bodily states and behaviours that manifest from activated EMS and habitual coping styles (Brockman et al. 2023). ED samples show a distinctive schema mode profile, with higher levels of maladaptive modes and lower levels of adaptive modes such as the Healthy Adult and Happy Child (Simpson et al. 2025; Talbot et al. 2015). In the Schema Mode Inventory for Eating Disorders (SMI‐ED and EN‐SMI‐ED‐SF), Detached Protector and Detached Self‐Soother modes are particularly elevated in BED, Demanding and Punitive Critic modes are more prominent in restrictive presentations, and the Eating Disorder Overcontroller mode is most strongly expressed in anorexia nervosa (AN) and BN (Simpson et al. 2025; Talbot et al. 2015). These detached modes map closely onto phenomenological descriptions of binge eating to ‘switch off’, numb emotions or self‐soothe through food (Marney et al. 2024) and are more frequent in individuals with higher weight and trauma, who also endorse higher maladaptive modes (Spirou et al. 2022). Qualitative work on schema therapy for EDs highlights that helping clients recognise the link between EMS, modes and ED cognitions can increase self‐awareness and nurture compassionate, agency‐building responses in place of ED patterns (Cunningham et al. 2024; Marney et al. 2024). Interoceptive Disconnection Emotion‐driven binge eating is also linked to disturbances in interoceptive awareness, which is one's capacity to notice, interpret and respond to bodily signals, including hunger, satiety and affective arousal (Herbert 2020; Herbert 2020). Many individuals with EDs experience alexithymia, a difficulty decoding these emotional and bodily signals, and rely on rigid rules or behaviours over internal cues to manage their state (Brucar et al. 2025; Makowski et al. 2025). Large‐scale studies synthesising EMS, emotion regulation, stress and social support show that EMS are associated with emotional overeating and poorer diet quality, with emotion regulation difficulties mediating this and perceived stress amplifying it (Gerges et al. 2022; Obara‐Gołębiowska 2025). In other words, EMS reflecting disconnection, rejection or impaired limits make people more prone to intense negative affect and emotion‐regulation difficulties, which in turn promote stress‐related overeating and unhealthy eating patterns (Gerges et al. 2022; Obara‐Gołębiowska 2025). Neurofunctional models identify binge eating subtypes characterised by high negative emotionality, impulsive approach behaviour or over‐controlled restraint, each displaying distinct resting‐state connectivity profiles (Brucar et al. 2025). From a schema therapy perspective, such profiles can be conceptualised as different constellations of modes interacting with interoceptive and regulatory capacities, emphasising the importance of therapeutic approaches that target this (Cunningham et al. 2024; Marney et al. 2024). Temperament, Complex Trauma and Individual Vulnerability Temperament, trauma and neurodiversity all influence one's vulnerability to developing an ED. Temperamental traits such as harm avoidance, perfectionism, behavioural inhibition and emotional dysregulation appear to predate ED onset and persist after recovery, reflecting genetic and neurocircuitry differences in reward, interoception and executive control (Makowski et al. 2025). Under‐controlled profiles (impulsivity, high negative emotionality, low effortful control) are more often associated with binge‐purge and binge eating presentations, whereas over‐controlled profiles (rigidity, inhibition, harm avoidance) predominate in restrictive disorders (Brucar et al. 2025; Makowski et al. 2025). EMS and temperament jointly predict disordered eating, with Disconnection/Rejection schemas and emotion regulation difficulties playing key roles (Gerges et al. 2022; Güler and Özgörüş 2022). ED patients with childhood trauma histories and higher‐weight individuals with PTSD symptoms tend to exhibit more severe EMS, maladaptive modes and greater symptom burden than those without trauma (Fasolato et al. 2024; Meneguzzo et al. 2021; Spirou et al. 2022). Neurodiversity adds a further layer of complexity to BED treatment. Individuals with autism and attention deficit/hyperactivity disorder (ADHD) experience higher rates of EDs, PTSD and other mental health conditions (Cobbaert et al. 2024; Spicer et al. 2024). ADHD is particularly prevalent in binge eating, with estimates of 10%–36% in BED and 15%–54% in BN (Appolinario et al. 2022; Kessler et al. 2013). Neurodivergent individuals are at increased risk of EMS due to neurocognitive differences, sensory sensitivities, attachment disruptions, adverse childhood experiences and minority stressors such as stigma and lack of accommodations (Cobbaert et al. 2024; Spicer et al. 2024). Qualitative research suggests that neurodivergent adults with BED often binge to regulate sensory overload, under‐stimulation and strong emotions, especially when their eating patterns are irregular and medication influences appetite (Makin, Meyer, Mondelli, and Tchanturia 2025). Adults with BN and ADHD similarly describe binge eating as a means for managing boredom, emotional chaos and inhibition difficulties (Makin, Meyer, Chubinidze, et al. 2025). Schema Therapy as a Mechanism‐Targeted Intervention Schema therapy is well positioned to address this multi‐layered pathway from early adversity, temperament and neurodiversity to binge eating. Schema therapy conceptualises EDs as maintained by EMS and critic and coping modes and aims to reduce symptoms by weakening EMS and maladaptive modes while strengthening adaptive modes and Healthy Adult functioning (Simpson and Smith 2019). Empirical work suggests that schema therapy can reduce ED symptoms, decrease EMS severity and improve general psychopathology, particularly in severe and enduring presentations where standard cognitive behavioural therapy (CBT) shows limited efficacy (Cunningham et al. 2024; Joshua et al. 2023). ED‐specific schema mode inventories show that maladaptive modes are strongly associated with neuroticism, whereas adaptive modes correlate with psychological flexibility. This supports the promise of mode‐focused interventions to target symptomatology and core patterns (Simpson et al. 2025). Collectively, these developments support the rationale for examining a temperament‐informed schema therapy approach in BED that addresses EMS, schema modes, emotion regulation and interoception while accommodating complex trauma histories and neurodivergent processing styles. Context and Aim of the Present Study Despite the promise of temperament‐informed schema therapy approaches for BED, few studies have examined how schema therapy protocols may alleviate binge eating, and none have evaluated temperament‐informed protocols in highly complex BED presentations. Most existing BED interventions are anchored in cognitive behavioural or behavioural weight‐loss frameworks, and relatively little is known about how manualised schema therapy can be adapted for clients with long‐standing binge eating, complex trauma histories and neurodivergent profiles or about the extent to which changes in EMS, modes and interoception accompany symptom improvement (Grilo 2024; Hay 2020). The present study addressed this gap by examining a manualised, temperament‐informed schema therapy intervention for binge eating presentations (ST‐BED‐TI). The proposed theoretical framework informing ST‐BED‐TI is illustrated in Figure 1. The model conceptualises perceived parenting and related early experiences, trauma exposure, temperament and neurodivergent context as vulnerability and contextual factors associated with the development and activation of EMS. These EMS are linked to schema modes and to interoceptive disconnection or body mistrust, which may in turn be associated with binge eating and broader ED symptoms. The treatment manual, which is currently being prepared for publication, was designed to accommodate individual differences in emotional sensitivity, neurodivergence, trauma exposure and sensory processing and maintain fidelity to core schema therapy principles through EMS and mode awareness, emotion regulation work, attention to unmet emotional needs, interoceptive reconnection and the strengthening of Healthy Adult functioning and corrective attachment experiences. The model is intended to clarify the theoretical targets of the intervention and to guide the interpretation of individual patterns of change and is conceptual rather than a tested mediation model. Given the uncontrolled case‐series design, the present study does not establish temporal precedence, causal pathways or mediation between changes in EMS, modes, interoceptive awareness and ED symptoms. Conceptual model of the hypothesised associations and proposed intervention targets of ST‐BED‐TI in BED. Using a small‐sample case series design with multiple time‐point assessments, the study evaluated changes in ED psychopathology, EMS, schema modes and interoceptive awareness across the course of ST‐BED‐TI in adults with BED. The primary aim was to evaluate preliminary change in binge eating associated with ST‐BED‐TI, rather than to estimate efficacy or establish causal mechanisms. Secondary aims were to examine whether changes in EMS, schema modes and interoceptive awareness occurred and whether these met criteria for reliable and clinically significant improvement at the individual level. Method Design We conducted a prospective case series study of ST‐BED‐TI in adults with BED. Participants completed assessments at pre‐treatment (baseline), mid‐treatment, post‐treatment and 3‐ and 6‐month follow‐up. Outcomes were evaluated at the individual level using reliable change and clinical significance indices, consistent with recommendations for small‐sample psychotherapy research (Kazdin 2023; Tate et al. 2016). Ethical Considerations Ethical approval was obtained from the University of Canberra Human Research Ethics Committee (HREC ID: 13243) and the UK's National Health Service (NHS ID: 326901). The trial was pre‐registered on the Australian New Zealand Clinical Trials Register (ID: ACTRN12623001259639). All participants provided informed consent for participation and use of anonymised data. Pseudonyms and identifiers were used to protect confidentiality. Participants Eligibility Criteria Eligible participants were adults (≥ 18 years), fluent in English, who met DSM‐5‐TR criteria for an ED characterised by binge eating, including BED, BN or other specified feeding or ED with binge eating or bulimic features of low frequency or duration. Exclusion criteria were active psychosis, acute severe suicidality, intellectual disability, current substance abuse, pregnancy, underweight status with active weight loss and prior schema therapy. Diagnoses and eligibility were determined through file reviews and a clinical psychologist's assessment rather than structured diagnostic interviews or comprehensive assessment batteries across the entire sample. Participants were recruited from two Australian private psychology practices and two UK NHS ED services. The first five women who met eligibility were informed about the project and reassured that choosing not to participate would not affect their treatment within that service. All five women consented to participate in the study, commenced ST‐BED‐TI and completed treatment. Four completed all assessment time points; one completed pre‐, mid‐ and post‐treatment assessments but declined the 3‐ and 6‐month follow‐ups due to executive functioning challenges. Sample Characteristics Five women in their 30s (M = 33.8 years) who completed ST‐BED‐TI were included. All met criteria for BED and had longstanding binge eating difficulties (minimum duration 16 years). Clinical complexity was high: four had complex trauma histories or symptoms consistent with C‐PTSD, four had mood disorders, two had anxiety disorders, one had a formal autism diagnosis (with one further suspected), one had ADHD (with two further suspected), and one had borderline personality disorder (BPD). Three had previously received psychological treatment, but none had received schema therapy or BED‐specific treatment. Although some participants were prescribed psychotropic medication, no medication changes were made for any participant during the treatment period. Three participants lived in Australia and two in the United Kingdom. Detailed clinical characteristics and treatment responses for each participant are presented in the Results section. Intervention Protocol ST‐BED‐TI is a manualised temperament‐informed schema therapy protocol for binge eating presentations, adapted from existing schema therapy programs for EDs (Simpson and Smith 2019). It integrates core schema‐therapy techniques with a weight‐neutral, trauma‐ and temperament‐informed framework. The protocol emphasises affect regulation, interoceptive awareness and meeting unmet emotional needs and allows clinicians to modify pacing, language and experiential work to individual temperamental and neurodivergent profiles (e.g., high emotional reactivity, sensory sensitivities and attentional variability). The ST‐BED‐TI protocol, currently being prepared for publication, is structured into four phases (see Table 1) and comprises 23 weekly sessions plus two fortnightly consolidation sessions (intended total: 25). Up to 10 additional sessions were available under the protocol when participants required extended trauma processing or management of significant life events. For such participants, the third phase was extended with a specific focus on the reason for extension (e.g., additional imagery rescripting for further trauma processing) while maintaining the tone and aims of the ST‐BED‐TI protocol. In the present sample, participants received 25 (n = 3), 30 (n = 1) or 35 (n = 1) sessions depending on clinical needs. This variation in treatment duration reflects the flexibility of the protocol, rather than allowing a uniform treatment dose; however, this does mean that additional trauma‐focused work may have contributed to outcomes. TABLE 1Temperament‐informed schema therapy (ST‐BED‐TI) protocol.PhaseSessionsAims16 Assessment Psychoeducation Mode mapping Early eating interventions 25 Intertwine trauma history and temperament style with schema model Orient to imagery rescripting Paradoxical diet 38 Imagery Rescripts focused on unmet needs in early life Reparenting and chairwork Strengthen Healthy Adult Interpersonal pattern breaking 46 Consolidate gains Habituate self‐nourishment Client‐led imagery rescripting Temperament‐informed schema therapy (ST‐BED‐TI) protocol. Assessment Psychoeducation Mode mapping Early eating interventions Intertwine trauma history and temperament style with schema model Orient to imagery rescripting Paradoxical diet Imagery Rescripts focused on unmet needs in early life Reparenting and chairwork Strengthen Healthy Adult Interpersonal pattern breaking Consolidate gains Habituate self‐nourishment Client‐led imagery rescripting Delivery Treatment was delivered in outpatient contexts across two Australian private practices and two UK NHS ED services by four experienced schema therapists (clinical psychologists). Therapy was provided via a combination of in‐person and telehealth modalities, depending on participant needs (e.g., level of disability, geographic area, accessibility and personal preferences) and service context. Delivery modality therefore varied across participants, with some receiving fully in‐person treatment, some fully online treatment, and others a mixed format. All therapists received monthly supervision from an accredited schema therapy trainer‐supervisor, experienced in the design and delivery of schema therapy interventions for EDs (S.S.). Therapist effects could not be estimated because each therapist treated a small and non‐equivalent set of participants and therapist competence was not independently rated. Treatment fidelity was not formally rated; however, in supervision, all therapists reported progress and session focus consistent with the manual as assessed by an observing clinician (P.J.). Procedure Following informed consent, participants completed baseline measures, which informed individualised schema formulations (including EMS, parenting experiences and temperament). Mid‐treatment assessment occurred following session 12, post‐treatment assessment occurred following the final therapy session, and follow‐up assessments were conducted at 3 and 6 months post‐treatment. Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Canberra (Harris et al. 2019; Harris et al. 2009). Measures The primary outcome was ED psychopathology, assessed with the Eating Disorder Examination Questionnaire (EDE‐Q 6.0; Fairburn and Beglin 2008). Secondary outcomes of EMS, schema modes and interoceptive awareness were measured using the Young Schema Questionnaire—Rasch Version (YSQ‐R; Yalcin et al. 2022), Schema Mode Inventory for Eating Disorders (SMI‐ED; Simpson et al. 2018) and the Multidimensional Assessment of Interoceptive Awareness—2 (MAIA‐2; Mehling et al. 2018), respectively. Eating Disorder Examination Questionnaire (EDE‐Q 6.0) The EDE‐Q (Fairburn and Beglin 2008) is a 28‐item self‐report measure assessing ED psychopathology over the past 28 days. Items are rated on a 7‐point Likert scale (0–6), yielding four subscales (Restraint, Eating Concern, Shape Concern, Weight Concern) and a Global score, calculated as the mean of the four subscales. Higher scores indicate greater ED severity. Item 14 (overeating episodes with loss of control) was used to index binge eating frequency. This item refers to the past 28 days, so it was divided by four to calculate average weekly binge frequency. The EDE‐Q demonstrates good reliability and sensitivity to change in BED samples (Dingemans et al. 2012; Reas et al. 2006). Young Schema Questionnaire—Rasch Version (YSQ‐R) The YSQ‐R (Yalcin et al. 2022) assesses 20 EMS on a 6‐point scale, with higher scores indicating stronger schema endorsement. All schema subscales were analysed. Baseline EMS scores informed formulation, whereas repeated administration allowed assessment of EMS change. Published internal consistency coefficients (α = 0.79–0.93) and standard deviations from a clinical schema therapy sample were used to calculate reliable change indices (RCIs) for each EMS (Yalcin et al. 2022). Schema Mode Inventory for Eating Disorders (SMI‐ED) The SMI‐ED is a 190‐item self‐report that assesses 16 schema modes in ED populations, rated on a 6‐point Likert scale (Simpson et al. 2018). Mode‐specific means, standard deviations and internal consistencies (α ≈ 0.78–0.93) from clinical ED samples (Simpson et al. 2018) were used to compute RCIs for each mode. Because formal test–retest reliability data are not yet available for the SMI‐ED, internal consistency coefficients were used as conservative proxies for reliability. This is consistent with prior RCI applications in psychological research (Lambert and Ogles 2009). Multidimensional Assessment of Interoceptive Awareness—2 (MAIA‐2) The MAIA‐2 (Mehling et al. 2018; Mehling et al. 2012) assesses eight interoceptive domains with subscale scores that are computed as mean item scores. RCIs were calculated for each subscale using standard deviations and reliability estimates (α = 0.67–0.86) from validation samples reported by Mehling et al. (2018). Although these data come from mixed clinical and community samples instead of ED‐specific cohorts, the MAIA‐2 has demonstrated sensitivity to change in clinical interventions and has been widely used in ED research, supporting its use for RCI calculations in the present study. Baseline Measures The Adult Temperament Questionnaire—Short Form (ATQ‐SF; Evans and Rothbart 2007), which assesses individual temperament traits related to emotional reactivity and regulation, and the Young Parenting Inventory—Revised (YPI‐R; Louis et al. 2018), which assesses perceived parenting patterns that are associated with schema development, were administered at baseline. These measures were used clinically to inform individualised case formulation but were not included as outcome variables. Consistent with the ATQ‐SF authors' recommendations to avoid categorisation or comparison to norms (Evans and Rothbart 2007), subscales were interpreted on a scale from 1 (i.e., extremely low) to 7 (i.e., extremely high). Participants were screened for the presence of other clinical diagnoses such as mood, anxiety, personality‐ and trauma‐related disorders, ADHD and autism in their initial assessment with the assigned clinical psychologist against DSM‐5‐TR criteria (American Psychiatric Association 2022). This was combined with a thorough review of their clinical history and the addition of further assessments at the discretion of their clinical psychologist. No uniform structured diagnostic interview or standardised autism and ADHD assessment battery was administered across the sample. Formal diagnoses, suspected diagnoses and clinically observed features are therefore distinguished where relevant. Data Analysis Given the small sample size and exploratory aims, data were analysed at the individual level using a case‐series framework. This approach is recommended for psychology case series and early‐phase intervention studies, where heterogeneity and idiographic change are central (Kazdin 2023; Tate et al. 2016). RCIs (Jacobson and Truax 1991) were calculated for all primary and secondary outcomes to determine whether individual changes were higher than what would be expected from measurement error. Reliability parameters were drawn from different clinical, community and validation populations, and formal test–retest estimates were unavailable for some secondary measures. Accordingly, RCIs were used descriptively within an exploratory case‐series framework to characterise reliable change for each participant. They were not interpreted as evidence of group‐level treatment efficacy or causal effects. An absolute RCI value of ≥ 1.96 was interpreted as reliable change under the conventional two‐sided p < 0.05 criterion, indicating that the observed change was unlikely to be attributable to measurement error alone. The most conservative estimate was used in the case where multiple reliability estimates were published. For the EDE‐Q Global score, clinical significance was evaluated using a cut‐off of 2.8 to distinguish clinical from community levels of ED psychopathology (Mond et al. 2008; Velkoff et al. 2023). Participants were classified as Recovered (reliable improvement and Global score < 2.8); Improved (reliable improvement but Global ≥ 2.8); Unchanged (no reliable change); and Deteriorated (reliable worsening). RCIs and clinical‐significance classifications were computed for pre‐ to post‐treatment and pre‐treatment to 6‐month follow‐up. All participants completed pre‐, mid‐ and post‐treatment measurements; however, one declined to complete the 3‐ and 6‐month follow‐up measures. No imputation was used, and follow‐up interpretations are therefore based on four participants only. As such, RCIs and clinically significant change were computed from pre‐post treatment and pre‐treatment to 6‐month follow‐up so that the group could be compared on one of these measures. Results Participant Characteristics All five participants completed the full course of ST‐BED‐TI. Four completed all assessment time points; one declined 3‐ and 6‐month follow‐up but completed pre‐, mid‐ and post‐treatment assessments. At baseline, all participants met criteria for BED and reported elevated EDE‐Q Global scores, multiple EMS, pronounced maladaptive schema modes and reduced interoceptive awareness. All had long‐standing binge‐eating (≥ 16 years), most had complex trauma histories, and several had neurodivergent features or diagnoses. Table 2 summarises participant demographics and baseline clinical characteristics, whereas Figure 2 and Table 3 outline their eating‐related changes. Table 4 details their changes in EMS, schema modes and interoception. TABLE 2Participant demographic and baseline clinical characteristics.PseudonymAge (years)BED duration (years)ComorbiditiesBaseline EDE‐Q globalWeekly binges a Key EMS (mean subscale score) b Dominant schema modes (mean subscale score) 2 Temperament (mean factor scores) c Sessions (n)Delivery modeKelly3124 PDD, GAD, C‐PTSD Underactive thyroid 3.5910 Self‐Sacrifice (4.83) Punitiveness to Others (4.75) Emotional Constriction (4.4) Vulnerability to Harm (4.33) Insufficient Self‐Control (4.14) Detached Self‐Soother (5.67) Helpless Surrenderer (4.8) Vulnerable Child (4.45) Demanding Parent Mode (4.4) Undisciplined Child (4.38) Negative affectivity (5.19) Orienting sensitivity (4.60) Effortful control (3.26) Extraversion/surgency (2.53) 25All in personMaggie3820 None at time of intake Previous: BN 3.771.5 Self‐Sacrifice (5.50) Social Isolation (4.20) Approval Seeking (4.20) Vulnerability to Harm (4.17) Insufficient Self‐Control (3.86) Compliant Surrenderer (4.6) Healthy Adult (4.06) Happy Child (3.91) Demanding Parent Mode (3.2) Helpless Surrenderer (3.00) Orienting sensitivity (6.13) Negative affectivity (4.62) Effortful control (3.21) Extraversion/surgency (2.53) 25All via videoBlair3216 None at time of intake Previous: MDD Complex trauma history 4.183.75 Failure (6.00) Approval Seeking (5.80) Punitiveness to Self (5.40) Emotional Constriction (5.20) Self‐Sacrifice (5.17) Demanding Parent Mode (5.73) Compliant Surrenderer (5.70) Vulnerable Child (5.15) Punitive Parent Mode (4.84) Detached Self‐Soother (4.83) Angry Child (4.80) Orienting sensitivity (5.73) Negative affectivity (4.81) Effortful control (3.47) Extraversion/surgency (3.41) 35All but one via videoTiana3317 BPD, Recurrent mood disorder with bipolar features, ADHD Ehlers Danlos syndrome, Fibromyalgia 4.327 Emotional Deprivation (5.00) Self‐Sacrifice (5.00) Abandonment (4.88) Failure (4.83) Mistrust/Abuse (4.60) Demanding Parent Mode (4.60) Detached Self‐Soother (4.50) Vulnerable Child (4.1) Angry Child (4.00) Undisciplined Child (3.88) Orienting sensitivity (5.60) Negative affectivity (5.58) Extraversion/surgency (2.94) Effortful control (2.11) 25Mostly via videoJosie3520 MDD, GAD, Agoraphobia, Autism PCOS Complex trauma history 4.415.25 Defectiveness/Shame (6.00) Failure (6.00) Social Isolation (6.00) Emotional Deprivation (5.80) Approval Seeking (5.40) Demanding Parent Mode (5.53) Vulnerable Child (5.45) Punitive Parent Mode (5.16) Detached Self‐Soother (5.00) Detached Protector (4.64) Negative affectivity (6.04) Orienting sensitivity (5.60) Extraversion/surgency (2.53) Effortful control (2.42) 30All via videoAbbreviations: ADHD = attention deficit/hyperactivity disorder, BN = bulimia nervosa, BPD = borderline personality disorder, C‐PTSD = complex post‐traumatic stress disorder, EMS = early maladaptive schemas, GAD = generalised anxiety disorder, PCOS = poly‐cystic ovarian syndrome, PDD = persistent depressive disorder. a Calculated using item 14 of EDE‐Q (episodes of overeating with loss of control) divided by four. b The five highest EMS and schema mode subscales. c Presented from highest to lowest. Participant demographic and baseline clinical characteristics. PDD, GAD, C‐PTSD Underactive thyroid Self‐Sacrifice (4.83) Punitiveness to Others (4.75) Emotional Constriction (4.4) Vulnerability to Harm (4.33) Insufficient Self‐Control (4.14) Detached Self‐Soother (5.67) Helpless Surrenderer (4.8) Vulnerable Child (4.45) Demanding Parent Mode (4.4) Undisciplined Child (4.38) Negative affectivity (5.19) Orienting sensitivity (4.60) Effortful control (3.26) Extraversion/surgency (2.53) None at time of intake Previous: BN Self‐Sacrifice (5.50) Social Isolation (4.20) Approval Seeking (4.20) Vulnerability to Harm (4.17) Insufficient Self‐Control (3.86) Compliant Surrenderer (4.6) Healthy Adult (4.06) Happy Child (3.91) Demanding Parent Mode (3.2) Helpless Surrenderer (3.00) Orienting sensitivity (6.13) Negative affectivity (4.62) Effortful control (3.21) Extraversion/surgency (2.53) None at time of intake Previous: MDD Complex trauma history Failure (6.00) Approval Seeking (5.80) Punitiveness to Self (5.40) Emotional Constriction (5.20) Self‐Sacrifice (5.17) Demanding Parent Mode (5.73) Compliant Surrenderer (5.70) Vulnerable Child (5.15) Punitive Parent Mode (4.84) Detached Self‐Soother (4.83) Angry Child (4.80) Orienting sensitivity (5.73) Negative affectivity (4.81) Effortful control (3.47) Extraversion/surgency (3.41) BPD, Recurrent mood disorder with bipolar features, ADHD Ehlers Danlos syndrome, Fibromyalgia Emotional Deprivation (5.00) Self‐Sacrifice (5.00) Abandonment (4.88) Failure (4.83) Mistrust/Abuse (4.60) Demanding Parent Mode (4.60) Detached Self‐Soother (4.50) Vulnerable Child (4.1) Angry Child (4.00) Undisciplined Child (3.88) Orienting sensitivity (5.60) Negative affectivity (5.58) Extraversion/surgency (2.94) Effortful control (2.11) MDD, GAD, Agoraphobia, Autism PCOS Complex trauma history Defectiveness/Shame (6.00) Failure (6.00) Social Isolation (6.00) Emotional Deprivation (5.80) Approval Seeking (5.40) Demanding Parent Mode (5.53) Vulnerable Child (5.45) Punitive Parent Mode (5.16) Detached Self‐Soother (5.00) Detached Protector (4.64) Negative affectivity (6.04) Orienting sensitivity (5.60) Extraversion/surgency (2.53) Effortful control (2.42) Abbreviations: ADHD = attention deficit/hyperactivity disorder, BN = bulimia nervosa, BPD = borderline personality disorder, C‐PTSD = complex post‐traumatic stress disorder, EMS = early maladaptive schemas, GAD = generalised anxiety disorder, PCOS = poly‐cystic ovarian syndrome, PDD = persistent depressive disorder. Calculated using item 14 of EDE‐Q (episodes of overeating with loss of control) divided by four. The five highest EMS and schema mode subscales. Presented from highest to lowest. Average weekly binges across participants and time points. TABLE 3EDE‐Q scores across time points.ParticipantPre‐treatmentMid‐treatmentPost‐treatment3‐month follow‐up6‐month follow‐upRCI a Clinical significanceKelly3.592.771.232.051.05 −3.66 RecoveredMaggie3.772.682.412.362.77−1.51ImprovedBlair4.184.412.232.232.73 −2.19 RecoveredTiana4.322.411.36 −4.27 b RecoveredJosie4.414.002.052.812.09 −3.76 Recovered a Those in bold met the absolute RCI value of ≥ 1.96. b RCI from pre‐treatment to post‐treatment used, as no follow‐up measures were provided. EDE‐Q scores across time points. Those in bold met the absolute RCI value of ≥ 1.96. RCI from pre‐treatment to post‐treatment used, as no follow‐up measures were provided. TABLE 4Reliable change summary across all measures.ParticipantEDE‐Q RCI a Clinical significanceYSQ‐R RCI c , d SMI‐ED RCIMAIA‐2 RCIKellyYesRecovered Vulnerability to harm Self‐sacrifice Insufficient self‐control Emotional constriction Punitiveness (other) Mistrust/abuse Defectiveness/shame Dependence/incompetence Enmeshment Unrelenting standards Approval seeking Negativity/pessimism Punitiveness (self) Vulnerable child Demanding parent mode Detached self‐soother Helpless surrenderer Angry child Impulsive child Happy child Punitive parent mode Healthy adult Detached protector Attention regulation Emotional awareness Body listening Trusting MaggieNoImproved Noticing Attention regulation Emotional awareness Body listening Trusting BlairYesRecovered Failure Approval seeking Emotional constriction Punitiveness (self) Subjugation Vulnerable child Punitive parent mode Demanding parent mode Compliant surrenderer Detached self‐soother Angry child Eating disorder overcontroller Body listening Trusting Tiana b YesRecovered Emotional deprivation Abandonment Mistrust/abuse Failure Self‐sacrifice Social isolation Defectiveness/shame Dependence/incompetence Vulnerability to harm Subjugation Insufficient self‐control Approval seeking Fear of loss of control Vulnerable child Angry child Demanding parent mode Detached self‐soother Happy child Punitive parent mode Healthy adult Self‐aggrandiser Eating disorder overcontroller Not worrying Attention regulation Emotional awareness Body listening Trusting JosieYesRecovered Approval seeking Insufficient self‐control Punitiveness (self) Vulnerable child Punitive parent mode Demanding parent mode Detached self‐soother Angry child Impulsive child Eating disorder overcontroller Not worrying a Yes = Absolute RCI value of ≥ 1.96. No = Absolute RCI value of < 1.96. b RCI from pre‐treatment to post‐treatment used, as no follow‐up measures were provided. c Subscales meeting an RCI value of ≥ 1.96 at the 6‐month follow‐up. Improvement was identified as an RCI of −1.96 in all YSQ‐R subscales, −1.96 in all SMI‐ED subscales except Happy Child and Healthy Adult, which required an RCI of +1.96 and an RCI of +1.96 in all MAIA‐2 subscales. d Those in bold are the subscales identified in Table 2 as the most elevated for each participant at baseline. Reliable change summary across all measures. Vulnerability to harm Self‐sacrifice Insufficient self‐control Emotional constriction Punitiveness (other) Mistrust/abuse Defectiveness/shame Dependence/incompetence Enmeshment Unrelenting standards Approval seeking Negativity/pessimism Punitiveness (self) Vulnerable child Demanding parent mode Detached self‐soother Helpless surrenderer Angry child Impulsive child Happy child Punitive parent mode Healthy adult Detached protector Attention regulation Emotional awareness Body listening Trusting Noticing Attention regulation Emotional awareness Body listening Trusting Failure Approval seeking Emotional constriction Punitiveness (self) Subjugation Vulnerable child Punitive parent mode Demanding parent mode Compliant surrenderer Detached self‐soother Angry child Eating disorder overcontroller Body listening Trusting Tiana b Emotional deprivation Abandonment Mistrust/abuse Failure Self‐sacrifice Social isolation Defectiveness/shame Dependence/incompetence Vulnerability to harm Subjugation Insufficient self‐control Approval seeking Fear of loss of control Vulnerable child Angry child Demanding parent mode Detached self‐soother Happy child Punitive parent mode Healthy adult Self‐aggrandiser Eating disorder overcontroller Not worrying Attention regulation Emotional awareness Body listening Trusting Approval seeking Insufficient self‐control Punitiveness (self) Vulnerable child Punitive parent mode Demanding parent mode Detached self‐soother Angry child Impulsive child Eating disorder overcontroller Yes = Absolute RCI value of ≥ 1.96. No = Absolute RCI value of < 1.96. RCI from pre‐treatment to post‐treatment used, as no follow‐up measures were provided. Subscales meeting an RCI value of ≥ 1.96 at the 6‐month follow‐up. Improvement was identified as an RCI of −1.96 in all YSQ‐R subscales, −1.96 in all SMI‐ED subscales except Happy Child and Healthy Adult, which required an RCI of +1.96 and an RCI of +1.96 in all MAIA‐2 subscales. Those in bold are the subscales identified in Table 2 as the most elevated for each participant at baseline. Participant 1 Kelly (Participant 1) described a childhood with loving parents but experiencing emotional neglect from an early age due to her mother's poor mental health. Parenting experiences reported on the YPI‐R suggested experiences of emotional inhibition and deprivation, controlling behaviour and degradation and rejection. She reported binge eating frequency with behavioural and psychological symptoms that would have met criteria for BED diagnosis at the age of 7 years had she been assessed, although the limits of this retrospective diagnosis are acknowledged. Kelly had engaged with a variety of psychological supports across her life, although none for binge eating, but continued to experience high levels of distress until engaging with the ST‐BED‐TI program. Participant 2 Maggie (Participant 2) had always experienced a contentious relationship with her mother, which was characterised by controlling behaviour, competitiveness and status seeking on the YPI‐R Maternal scales. Maggie's primary focus upon commencing treatment was to not feel that food controlled her and be able to eat in a way that supported her well‐being instead of being driven by episodes of binge eating and attempts at restriction. Towards the end of treatment, she reflected feeling that she had met this goal and now was able to actively and healthily engage with dietetic and exercise supports. Although Maggie reported no longer meeting clinical criteria for BED and some change in her EMS and schema modes, none of these met the criteria for RCI. Participant 3 Blair (Participant 3) described a highly contentious relationship with her mother with YPI‐R responses of excessive competitiveness and status seeking, emotional inhibition and deprivation, overprotection and overindulgence, punitiveness and controlling behaviour, a complex trauma history and suspected neurodivergence. She disclosed previously unrecognised sexual trauma mid‐treatment, which resulted in a 10‐session extension of phase three of the protocol. This allowed consolidation of her developmental trauma, sexual assault and attachment traumas in the context of the schema model. Participant 4 Tiana (Participant 4) had a complicated mental health history and suspected autism. She had received several episodes of psychological treatment; however, none were specific to BED. Like the other participants, Tiana reported more maladaptive parenting from her mother in her responses to the YPI‐R, particularly in the domain of punitiveness, and markedly elevated levels of negative affectivity and orienting sensitivity on the ATQ‐SF. Tiana experienced difficulties in executive functioning and did not complete follow‐up assessments, so her pre‐treatment results were compared to her post‐treatment scores only. Participant 5 Josie (Participant 5) entered treatment largely housebound and socially isolated. Among other diagnoses, Josie experienced autism, agoraphobia, C‐PTSD‐consistent symptoms and a highly enmeshed and controlling maternal relationship featuring degradation and rejection, emotional inhibition and deprivation, punitiveness and controlling behaviour as measured by the YPI‐R. She attended an additional five sessions to accommodate clinical complexity and comorbidities: assessment of autism, risk management and imagery rescripting for complex trauma. The ST‐BED‐TI treatment had vast impacts, with Josie voluntarily leaving her house several times across the course of treatment and engaging with further supports to support her in leaving her abusive and enmeshed home environment. Outcome Measures Eating Disorder Psychopathology At baseline, all participants scored above the EDE‐Q Global clinical cut‐off of 2.8. At post‐treatment, four of five participants showed reliable improvement on EDE‐Q Global scores and fell below the clinical cut‐off, meeting criteria for recovery. The fifth participant demonstrated a substantial reduction in symptoms that did not reach the RCI threshold. At baseline and below the clinical cut‐off at post‐treatment and the 6‐month follow‐up. Of the four participants who completed 6‐month follow‐up, three maintained reliable improvement and Global scores below 2.8, showing sustained recovery; the fourth remained below the clinical cut‐off but did not show further reliable change from pre‐treatment to follow‐up, indicating stable improvements. These classifications describe within‐person change against prespecified thresholds and do not establish intervention efficacy. Average reported binge eating frequency decreased across participants, with most moving from near‐daily binge episodes to rare or absent binges at post‐treatment and follow‐up. These changes paralleled reductions in global ED psychopathology and are shown in Figure 2. Because binge frequency was indexed through retrospective self‐report, these patterns should be interpreted cautiously. EDE‐Q scores, RCI and clinical significance across time points are shown in Table 3. Schema Processes Participants reported reductions in several EMS assessed by the YSQ‐R. Schema mode changes were also observed on the SMI‐ED, with decreases in maladaptive coping modes and increases in healthy adult functioning. EMS and schema modes that met RCI criteria are presented in Table 4. Interoceptive Awareness Improvements in interoceptive functioning were observed on the MAIA‐2, particularly in attention regulation, emotional awareness and body listening. All participants reported reliable improvements on at least one MAIA‐2 subscale, showing increased ability to recognise and respond to internal bodily states during treatment. Domains of interoceptive awareness that met RCI criteria are presented in Table 4. Discussion This case series examined a manualised, temperament‐informed schema therapy protocol (ST‐BED‐TI) for adults with longstanding BED and high clinical complexity. Across five participants, changes in ED psychopathology occurred alongside changes in EMS, schema modes and interoceptive awareness. Given the uncontrolled design, these findings are best interpreted as preliminary indicators of change and treatment acceptability rather than causal treatment effects. Symptom Change and Clinical Significance Four of five participants showed reliable improvement and clinically significant improvement on EDE‐Q Global scores at post‐treatment, meeting criteria for recovery. Three of the four participants who completed the 6‐month follow‐up maintained reliable and clinically significant gains; the fourth remained below the clinical cut‐off with stable symptomatology. All participants showed considerable reductions in binge eating frequency. These outcomes are notable given the chronicity of BED (≥ 16 years) and the high levels of comorbidity, trauma and neurodiversity in the sample, but they may also reflect non‐specific therapeutic factors, and they are consistent with emerging evidence that schema therapy approaches can be effective in severe, enduring ED presentations (Joshua et al. 2023). Beyond questionnaire scores, symptom improvement was rooted in significant functional changes, including increased autonomy, social participation and movement towards safer, less enmeshed living environments. Such changes are consistent with schema‐theoretical aims of strengthening Healthy Adult functioning and meeting core emotional needs, rather than focusing solely on behavioural symptom management (Brockman et al. 2023; Simpson and Smith 2019). Change in EMS and Schema Modes: Preliminary Evidence for Mechanism‐Targeted Effects A key aim of the study was to explore whether symptom change co‐occurred with shifts in EMS, schema modes and interoceptive awareness. Most participants exhibited reliable reductions in EMS, particularly in those central to binge eating presentations (e.g., Defectiveness/Shame, Emotional Deprivation, Failure, Vulnerability to Harm and Insufficient Self‐Control; Aloi et al. 2025; Basile et al. 2019; Rania et al. 2020). Notably, many of the EMS showing reliable change overlapped with those most elevated at baseline (e.g., Kelly's Vulnerability to Harm and Insufficient Self‐Control, Blair's Failure and Punitiveness and Josie's Defectiveness/Shame and Social Isolation, shown in Tables 2 and 4). Overall, the pattern of results is consistent with the hypothesised targets of ST‐BED‐TI, such that it is associated with change in each participant's core schema vulnerabilities rather than only peripheral beliefs. However, these findings cannot establish that the manualised treatment produced these changes, that they preceded symptom improvement, or that they were specific to the temperament‐informed component. Similarly, several participants exhibited reductions in maladaptive modes alongside increases in Healthy Adult and Happy Child modes. For Blair, extending Phase 3 after the disclosure of previously unrecognised sexual trauma allowed intensive imagery rescripting and reparenting work that coincided with reliable reductions in Vulnerable Child, Punitive Parent and Eating Disorder Overcontroller modes. Josie's large mode shifts were reflected in reduced Detached Self‐Soother and Punitive Parent scores and in her increased capacity to challenge maternal control and move towards a less abusive environment. Taken together, the individual patterns are compatible with the hypothesised mechanisms of ST‐BED‐TI: As EMS and maladaptive modes reduce, and adaptive modes strengthen, people with BED may become less reliant on binge eating to manage schema‐driven distress and more able to attend to emotional needs through compassionate, agency‐building strategies. This reflects recent findings indicating that understanding the impact of modes helps clients reframe binge eating as a coping attempt and develop alternative ways of meeting needs (Cunningham et al. 2024; Marney et al. 2024). However, without temporal sequencing capable of distinguishing whether process changes preceded symptom change, followed it or reflected broader therapeutic factors, these findings cannot demonstrate that EMS or mode changes mediated symptom improvement. Maggie's case provides a counterpoint to this account. She achieved recovery‐level improvements in ED symptoms and robust interoceptive gains but did not meet RCI thresholds for any EMS or mode. This pattern suggests that, for some individuals, symptom change may be driven more by enhanced interoceptive awareness and emotion‐regulation capacity than by measurable reductions in EMS or maladaptive modes, at least as captured by current instruments. It is therefore possible that interoceptive, behavioural, relational or adaptive schema changes were more relevant for Maggie, rather than substantially weakening her existing EMS, which began at relatively moderate levels. In this context, standard EMS‐focused measures may have limited sensitivity to growth in early adaptive schemas, secure‐attachment representations or nuanced shifts in Healthy Adult functioning, particularly when baseline adaptive resources are already relatively high (van Donzel et al. 2026). Future work would benefit from incorporating measures of early adaptive schemas to capture resilience‐building changes that may underpin symptom improvement in BED even when EMS scores remain relatively stable, in addition to more frequent process assessment. Interoceptive Awareness and Emotion Regulation Interoceptive awareness improved reliably in all participants on at least one MAIA‐2 domain. These improvements complement the observed EMS and mode changes and are consistent with theoretical accounts linking interoceptive disconnection, emotional avoidance and binge eating (Cella et al. 2019; Romano et al. 2026). As ST‐BED‐TI helped participants name and rework EMS and modes, it may also have facilitated increased tolerance of bodily sensations and emotions, making it safer to ‘stay with’ internal experience rather than escape via binge eating. The present findings broaden the extant literature linking EMS, emotion regulation and binge eating (Gerges et al. 2022; Obara‐Gołębiowska 2025), suggesting that interoceptive and emotion regulation changes warrant further study, but do not establish that schema or mode work caused the observed changes in bodily awareness or symptoms. Can ST‐BED‐TI Tolerate Complexity? Temperamental traits, complex trauma histories and neurodivergent processing styles are increasingly recognised as principal contributors to BED vulnerability and to the heterogeneity of treatment response (Bryant et al. 2025; Kowalewska et al. 2024). All participants had numerous EMS and maladaptive modes, and most reported complex trauma histories, consistent with evidence that EMS mediate the association between childhood trauma and ED severity (Fasolato et al. 2024; Meneguzzo et al. 2021). Despite this complexity, ST‐BED‐TI appeared clinically manageable and safe in these outpatient services. The protocol's flexibility in pacing, session number, delivery format and explicit attention to trauma, temperament and neurodiversity allowed therapists to manage high risk, treatment ruptures and practical barriers without abandoning the core framework. Although preliminary, it is encouraging that participants with C‐PTSD‐consistent symptoms, BPD, autism, ADHD and significant functional impairment achieved recovery‐level improvements on ED outcomes. However, the design cannot determine whether these adaptations, the broader schema therapy model, common therapeutic factors or concurrent care accounted for improvement. In this study, neurodiversity‐related processes were particularly salient: Josie's late‐identified autism and severe agoraphobia, Blair's suspected neurodivergence and Tiana's suspected autism all demonstrate the need for adaptations in modality, sensory pacing and explicit validation of neurodivergent needs within a weight‐neutral, non‐stigmatising framework. However, neurodivergence was not assessed with a uniform standardised battery, and the small sample prevents further statistical inference of these findings—consequently, these examples cannot establish neurodiversity‐specific treatment effects. In line with the extant literature, ST‐BED‐TI's emphasis on validating and accommodating sensory and emotional needs, adopting a weight‐neutral stance and integrating interoceptive training may be especially helpful for neurodivergent people (Cobbaert et al. 2024; Makin, Meyer, Chubinidze, et al. 2025; Makin, Meyer, Mondelli, and Tchanturia 2025; Spicer et al. 2024). However, given the small number of neurodivergent participants and the absence of formal neurodiversity‐specific measures, these observations require replication in a larger, targeted sample. Strengths and Limitations This study has several notable strengths. These include the manualised ST‐BED‐TI protocol; multiple assessment time points; and a spotlight on individual‐level outcomes that are well suited to small‐sample, early‐phase psychotherapy research (Kazdin 2023; Tate et al. 2016). Assessing ED pathology, EMS, schema modes and interoceptive awareness within the same participants provided a preliminary multi‐level view of change processes. At the same time, major limitations must be acknowledged. The sample size was very small (n = 5), uncontrolled and limited to women in their 30s, restricting generalisability and precluding causal inference. It is acknowledged that the variability of delivery across in‐person and online formats had the potential to influence therapeutic alliance, emotional processing, access to experiential exercises and measurement completion, although therapists anecdotally reported that they did not observe this. Further, delivery of therapy via videoconferencing is well established as equally effective compared with in‐person (Amaro et al. 2025; Simpson and Slowey 2011). This mixed format represents ‘real‐life’ therapeutic delivery, which increases equitable access to treatment with similar outcomes (Wolfe et al. 2025). Similarly, improvements could reflect non‐specific therapeutic factors, regression to the mean, natural fluctuations, therapist effects or individualised extensions of treatment. The study relied exclusively on self‐report measures; there were no behavioural or physiological indices of interoception, and standardised measures of neurodivergence or other diagnoses, such as C‐PTSD, were not administered to the whole sample. Diagnoses were based on file review and clinical psychologist assessment rather than structured diagnostic interviews. One participant did not complete follow‐up, leaving four participants for longitudinal interpretation. Finally, although all therapists were experienced schema therapists and received regular supervision, treatment fidelity, therapist competence and therapist effects were not formally assessed, limiting conclusions about which components of ST‐BED‐TI were most active and about the model's replicability across settings. RCI estimates were also based on psychometric parameters drawn from different populations for some secondary measures, so they should be interpreted as provisional indicators rather than precise estimates of clinically meaningful change. Future Directions These preliminary findings point to several priorities for future research. Larger, controlled trials are needed to compare ST‐BED‐TI with established BED treatments and standard schema therapy for EDs, while directly testing the added value of temperament‐informed adaptations to evaluate relative efficacy, durability of gains and cost‐effectiveness (Grilo 2024; Hay 2020). Mechanism‐focused designs, such as single‐case experimental designs with frequent measurement (Kazdin 2023), might provide more information on the relationships between changes in EMS, modes, interoceptive awareness and binge eating frequency and test whether shifts in these processes mediate symptom improvement. Study designs would benefit from staggering baseline assessments across participants to improve our assessment of whether symptom improvements occurred in response to treatment and from measuring binge frequency through regular diaries or session‐level measurement, while incorporating formal fidelity monitoring. Further work should also incorporate preregistered moderator and subgroup analyses to examine whether temperament profiles, trauma or neurodivergent status predict who benefits most from ST‐BED‐TI, building on emerging distinctions between under‐ and over‐controlled ED presentations (Brucar et al. 2025; Makowski et al. 2025). In addition, there is a need to evaluate and improve neurodiversity‐informed adaptations of schema therapy, drawing on lived experience and co‐design methods to optimise language, pacing, modality and sensory accommodations for autistic and ADHD clients with BED (Cobbaert et al. 2024; Makin, Meyer, Mondelli, and Tchanturia 2025; Spicer et al. 2024). Integrating qualitative data, including the companion qualitative study of participant and clinician perspectives referenced in the methods (Joshua et al. 2026), will be important for deepening understanding of acceptability, perceived mechanisms of change and barriers to engagement and for informing further development of ST‐BED‐TI. Conclusion Within the constraints of a small, uncontrolled case series, this study provides preliminary evidence that a temperament‐informed schema therapy protocol was associated with changes in EMS, modes and interoceptive awareness, alongside BED symptom improvement. These findings are consistent with the proposed theoretical mechanisms of ST‐BED‐TI. For most participants, recovery‐level improvements on ED outcomes were accompanied by reductions in core vulnerability EMS and maladaptive modes and by enhanced capacity to notice and trust internal bodily and emotional messages. One participant's trajectory, characterised by marked gains in interoceptive awareness without reliable EMS or mode change, suggests that for some individuals, strengthening early adaptive schemas, Healthy Adult functioning and bodily awareness may be more crucial for symptom change than reducing EMS. Across cases, ST‐BED‐TI appeared able to accommodate complex trauma, high negative affectivity, neurodivergent processing and practical barriers using flexible dosing, telehealth delivery and targeted rupture‐repair work, within an overarching trauma‐ and temperament‐informed framework. Although larger, controlled and mechanism‐driven studies are needed to confirm efficacy, feasibility, clarify pathways of change and refine neurodiversity‐informed adaptations (Aloi et al. 2025; Kazdin 2023; Tate et al. 2016), the present findings suggest that ST‐BED‐TI may be a promising and clinically manageable treatment approach for adults with longstanding and complex BED. Author Contributions Phoebe R. Joshua: conceptualisation, data curation, formal analysis, investigation, methodology, project administration, resources, validation, writing – original draft and writing – review and editing. Susan Simpson: conceptualisation, ST‐BED‐TI protocol development, data curation, investigation methodology, project administration, resources, supervision, validation and writing – review and editing. Vivienne Lewis: conceptualisation, project administration, supervision, validation and writing – review and editing. Sally F. Kelty: project administration, supervision, validation and writing – review and editing. Funding This work was supported by the Australian Government Research Training Program. Ethics Statement This study was approved by the University of Canberra Human Research Ethics Committee (ID: 13243) and the UK's National Health Service (ID: 326901). Conflicts of Interest The authors declare no conflicts of interest. Acknowledgements We would like to thank the clinicians who delivered this intervention and participated in this study: Dr. Rachel Gold, Dr. Emma DeCicco and Dr. Isabelle Parks. It would not have been possible without your experience and generosity. This research was supported by the Australian Commonwealth through an Australian Government Research Training Program Scholarship. Open access publishing facilitated by University of Canberra, as part of the Wiley ‐ University of Canberra agreement via the Council of Australasian University Librarians Data Availability Statement The data that support the findings of this study are available from the lead author via phoebe.joshua@canberra.edu.au upon reasonable request. References Associated Data Data Availability Statement The data that support the findings of this study are available from the lead author via phoebe.joshua@canberra.edu.au upon reasonable request.