Abstract
Research suggests that standalone dialectical behaviour therapy (DBT) skills training results in reduced emotion regulation difficulties, depression, and anxiety across various clinical populations. Although autistic individuals experience heightened emotion regulation difficulties compared to the general population, limited research is available on the effectiveness of DBT for autistic populations, particularly adapted DBT programmes to meet the specific needs of the autistic community. In this study, we aimed to examine the preliminary impact of a neurodiversity-affirming informed DBT group skills training programme (NDA-DBT) on emotion regulation difficulties, psychological distress, mindfulness and quality of life. We collected data from 12 participants using the Difficulties in Emotion Regulation Scale, Depression Anxiety and Stress Scale, Five Facet Mindfulness Questionnaire, and World Health Organization Quality of Life-Brief (WHOQOL-BREF). Preliminary results of the pilot study suggest decreased emotion regulation difficulties and psychological distress, increased mindfulness skills and improvement across the psychological and social quality of life domains. The results of this pilot study suggest that the NDA-DBT programme seems promising for reducing emotion regulation difficulties and psychological distress and increasing mindfulness skills. Future research on neurodiversity-affirming informed DBT will assist the autistic community in accessing psychological support adapted to their unique needs and challenges.
Keywords
Autism Spectrum Disorder (Autism)
Autism is a form of neurodivergence characterised by natural variation in communication, social interaction, sensory processing and patterns of interests and activities, which manifest uniquely across individuals and throughout the lifespan (American Psychiatric Association [APA], 2022; Pellicano & den Houting, 2022). Challenges experienced by autistic individuals are now understood as emerging from the interaction between individual neurodevelopmental differences and environmental demands, rather than being intrinsic deficits. This perspective acknowledges that functional outcomes are shaped by contextual factors, including sensory environments, social expectations and support systems, and underscores the importance of interpreting autistic experiences within a broader, biopsychosocial framework (Bölte et al., 2024; Milton, 2012). The global prevalence of autism has recently been estimated at 61.8 million (95% uncertainty interval = 52.1–72.7 million), or approximately one in every 127 people globally (Santomauro et al., 2024). However, only six of the 105 studies that informed this prevalence estimate focused on adults (Santomauro et al., 2024).
Research shows that autistic individuals experience an elevated prevalence of co-occurring mental health conditions. Lai et al.'s (2019) meta-analysis of 96 studies conducted between 1993 and 2019 examined co-occurring mental health conditions in autistic individuals. Lai et al. (2019) reported pooled prevalence estimates of 28% for attention deficit hyperactivity disorder (ADHD), 20% for anxiety disorders, 13% for sleep-wake disorders, 12% for disruptive, impulse-control, and conduct disorders, 11% for depressive disorders, 9% for obsessive-compulsive disorder (OCD), 5% for bipolar disorders, and 4% for schizophrenia spectrum disorders. Autistic adults who experience co-occurring mental health challenges have an elevated risk of suicidality and report diminished overall quality of life (Kirby et al., 2019; Roestorf et al., 2022). While the research surrounding the co-occurrence of psychiatric conditions among autistic individuals is often focused on and presented as a product of individual vulnerability, it is important to also consider the influence of contextual factors (Greenlee et al., 2020; Taylor & Gotham, 2016). Additionally, autistic individuals have a higher risk of experiencing emotion dysregulation (McDonald et al., 2024). This refers to challenges in utilising adaptive emotion regulation strategies effectively and relying on less effective responses when navigating experiences and their environment (Conner et al., 2020b; Gross & Thompson, 2007). A meta-analysis by McDonald et al. (2024) reported that autistic individuals experience a significantly higher severity of emotion dysregulation, showing a large effect size (Hedges’ g = 1.07). Emotion dysregulation has been found to significantly impact autistic traits, leading to increased challenges with social interactions, heightened reactivity to emotive situations, and increased repetitive behaviours to cope with dysregulated emotions (Conner et al., 2021; Goldsmith & Kelley, 2018; Samson et al., 2015; Swain et al., 2015). Furthermore, emotion dysregulation has been identified as a significant predictor of co-occurring mental health conditions, contributing to heightened anxiety, increased suicidality, diminished social connections, and a lower overall quality of life in autistic children and adults (Beck et al., 2020; Conner et al., 2020a; Conner et al., 2020b; Vasa et al., 2018). A qualitative study of autistic adults’ experiences of support and treatment for mental health challenges found that they had difficulty finding appropriate treatment and support, stating that when they presented with co-occurring mental health difficulties, they were often viewed as “too complex” or perceived as “high-functioning” and coping well with their autistic characteristics (Camm-Crosbie et al., 2019). Furthermore, current evidence-based psychological treatments for autistic individuals are predominantly behaviour-based interventions such as applied behaviour analysis (ABA). While ABA is contested in autistic communities, it still dominates practice (Allen et al., 2024). A qualitative study of adults who received ABA interventions as children reported memories of traumatic events related to ABA, beliefs that they should not be forced to behave like their peers, some benefits but also significant long-term negative consequences, that ABA is an unethical intervention, and that ABA practitioners should listen to autistic individuals and consider alternative interventions (Anderson, 2023).
Neurodiversity-Affirming Framework
In 1998, the word ‘neurodiversity’ was introduced by autistic sociologist Judy Singer and journalist Harvey Blume (Blume, 1998; Singer, 1998), establishing a cohesive framework for advocacy pertaining to autistic individuals and the broader neurodivergent community. The neurodiversity-affirming framework acknowledges and values the natural diversity of human neurodevelopment, seeing autism and other conditions as variations that should be respected and embraced, rather than cured (Singer, 1998; Walker, 2012). The strengths and unique skills of neurodivergent individuals are highlighted, fostering respect and dignity irrespective of deviations from neurotypical norms (Walker, 2014). This approach encompasses all neurodivergent individuals, including those requiring complex support, and it follows the social model of disability. This model proposes that disability results from societal barriers rather than individual impairments (Oliver, 1996). This framework promotes autism acceptance, supports a positive sense of autistic identity, and fosters greater acceptance, which is linked to improved mental health outcomes (Cage et al., 2018). It also stresses the importance of involving neurodivergent individuals in decision-making processes, ensuring that their voices are heard and that their needs are prioritised in neurodivergent research and practice (Pellicano et al., 2014). Neurodiversity-affirming frameworks mark a significant shift from viewing autism through a deficit-based lens to recognising it as a broad spectrum of distinct abilities, strengths and differences (Cherewick & Matergia, 2024).
Pantazakos and Vanaken's (2023) paper on neurodiversity-affirming clinical practices centres the lived experiences and perspectives of autistic individuals as equal partners in clinical decision making, recognising their unique knowledge and expertise. It resists interventions aimed at normalising behaviour or suppressing autistic traits (i.e., ABA), instead prioritising environmental adaptations that reduce barriers and stigma. These approaches emphasise respect for consent, autistic self-narratives and validating diverse ways of being. Finally, Pantazakos and Vanaken theorise that neurodiversity-affirming practice seeks to balance respect for the autistic self with appropriate therapeutic challenge, ensuring that interventions remain meaningful, collaborative and genuinely supportive from the autistic person's perspective. Their paper highlights cognitive behavioural therapy (CBT) as a promising approach for addressing mental health challenges, particularly anxiety, in autistic individuals, while emphasising the need for systematic adaptations to accommodate sensory, cognitive and communication differences (Pantazakos & Vanaken, 2023). However, the need for clinicians to adapt CBT protocols has been reported as challenging due to limited clinician confidence, knowledge and ability to make necessary adaptations (Cooper et al., 2018; Spain & Happé, 2020).
Dialectical Behaviour Therapy
Dialectical behaviour therapy (DBT) is grounded in CBT and principles from dialectical philosophy and Zen practices and was developed to assist individuals experiencing severe emotional dysregulation (Linehan et al., 1991). This therapeutic approach integrates a dual focus on validation and acceptance, alongside cognitive and behavioural transformation (Linehan, 1993a). The acceptance-change dialectic acknowledges that self-acceptance and personal growth can simultaneously exist (Hartmann et al., 2012). DBT assumes that dysfunctional behaviour stems from emotion dysregulation or the downregulation of negative emotions attributable to skill deficits (Linehan, 1993a; Neacsiu et al., 2010). As such, DBT is considered a strengths-based model, as it focuses on developing and enhancing existing skills and capabilities while also facilitating the acquisition of new skills and abilities (Lew et al., 2006; Linehan, 1993b; Neacsiu et al., 2010). Skill training, often delivered in groups, focuses on four core modules: mindfulness, distress tolerance, emotion regulation and interpersonal effectiveness. DBT was initially developed for individuals with borderline personality disorder (BPD) who exhibit chronic suicidal behaviour. When delivered in full, it integrates individual therapy, group skill training, telephone coaching and a therapist consultation team (Linehan et al., 1991). Randomised control trials and systematic reviews have shown that DBT reported significant improvements in suicidality, general psychopathology and BPD symptoms 1- to 2-year post-intervention, and significantly reduced suicide attempts, hospitalisation for suicide ideation, dropout rates, psychiatric hospitalisations, emergency department visits and depression symptoms (Gillespie et al., 2022; Hernandez-Bustamante et al., 2024; Linehan et al., 2006; McMain et al., 2022). A benchmarking study by Azevedo et al. (2024) found that DBT was consistently associated with moderate to large improvements in health-related quality of life for both adolescents and adults with borderline personality symptoms, supporting its effectiveness beyond symptom reduction alone. Furthermore, recent studies have highlighted the effectiveness of standalone DBT skills training in addressing emotion regulation difficulties and enhancing psychological wellbeing. For instance, Kells et al. (2020) found that DBT skills training significantly reduced emotion dysregulation and improved mindfulness skills in adults with BPD, emerging BPD traits, or broader difficulties with emotional regulation. Similarly, Lee et al. (2022) reported that participation in DBT skills training was associated with improvements in quality of life for women presenting with BPD traits.
Furthermore, DBT has been effectively used in diverse populations, including those with eating disorders, mood disorders, substance use disorders, intellectual disabilities, depression, non-suicidal self-injury, ADHD and post-traumatic stress disorder (Courbasson et al., 2012; Fleming et al., 2015; Harned et al., 2014; McNair et al., 2016; Safer et al., 2010; Sakdalan et al., 2010; Van Dijk et al., 2013). DBT has been shown to reduce self-harming behaviours and emotion dysregulation in autistic adolescents (Phillips et al., 2024) and emotion dysregulation in autistic adults exhibiting self-harming or suicidal behaviours (Bemmouna et al., 2022). A randomised controlled trial of autistic outpatients aged 18 to 65 years (N = 123) exhibiting suicidal behaviour found that DBT, compared to treatment as usual, reduced suicidal ideation, suicide attempts and depression severity. However, the study reported no significant effect on social anxiety and the effects were not sustained at the 12-month follow up (Huntjens et al., 2024). DBT, including DBT skills training only, was also found to have high retention, attendance and satisfaction rates among autistic adults (Bemmouna et al., 2022; Ritschel et al., 2022). While DBT is the gold standard evidence-based treatment for BPD (May et al., 2016), thoughtful adaptations are required to meet the specific needs of autistic adults. Although autism and BPD share some overlapping characteristics, it is crucial to recognise that these conditions arise from distinct underlying mechanisms (Dudas et al., 2017). For example, an individual with BPD may experience temporary issues with social interactions during episodes of intense emotional distress. In contrast, an autistic person may experience pervasive and consistent challenges with social interactions because of difficulties in interpreting social cues (Gordon et al., 2020). Consequently, when considering the use of DBT with autistic individuals, it is crucial to consider the specific skill development and acquisition required by autistic adults (Hartmann et al., 2012) and further consider that autistic individuals do not experience a skill deficit, but rather a skill difference (e.g., the double empathy problem; Milton, 2012; Sasson et al., 2017).
Limitations of DBT
Although DBT is an evidence-based intervention for emotion dysregulation, its standard delivery is not inherently neurodiversity-affirming. Traditional DBT was designed to help clients manage distress and adapt to social environments, often using concepts and strategies rooted in neurotypical norms. As a result, some elements of DBT may implicitly or explicitly promote the suppression of autistic traits (such as stimming, direct communication styles, or sensory needs) to fit social expectations, a process known as masking. Masking has been linked to poorer mental health outcomes and increased stress among autistic people (Cage & Troxell-Whitman, 2020; Hull et al., 2017). Additionally, DBT's focus on teaching ‘adaptive’ social and emotional behaviours can inadvertently prioritise normalisation and encourage clients to conform to societal standards of behaviour rather than affirming their authentic neurodivergent identities. Without explicit adaptations or a neurodiversity-affirming framework, DBT risks reinforcing the message that autistic ways of being are undesirable or need to be changed to gain social acceptance. Furthermore, when delivered in full, DBT is resource-intensive and expensive, which can limit accessibility (Abraham et al., 2024). DBT's structured nature, group format, and use of abstract concepts can be cognitively demanding and require emotion recognition and expression, which is not suitable for some autistic individuals (Cooper et al., 2018).
Neurodiversity-Affirming DBT-Informed Skills Training Programme
The neurodiversity-affirming DBT (NDA-DBT) informed group skills training programme is a 10-week programme for autistic adults, primarily based on Marsha Linehan's DBT Skills Training Manual (1993b), developed by a clinical psychologist and a clinical neuropsychologist, and an autistic clinical psychology registrar with lived experience. This programme was adapted based on the neurodiversity-affirming framework and explicitly designed to support autistic adults in maintaining their identity while effectively navigating the challenges posed by a primarily neurotypical environment.
It includes 90 minute sessions designed to assist clients in managing emotional dysregulation and psychological distress while developing mindfulness skills (Sakdalan & Weerakoon, 2024). Although DBT was developed for clients with complex, multi-diagnostic, high-risk disorders (Linehan & Wilks, 2015) and has been successfully adapted for other conditions, its application within a neurodiversity-affirming framework remains unexplored. To the best of our knowledge, no existing neurodiversity-affirming adaptations of DBT have been researched to date. While several DBT adaptations have been trialled with autistic individuals these programmes focused on creating a predictable environment, addressing sensory and social anxiety needs, simplifying materials with visual supports, offering concrete and structured instructions, providing planning support, incorporating participants’ focused interests to enhance engagement and motivation, and supporting alexithymia (Bemmouna et al., 2022; Ritschel et al., 2022; Weiner et al., 2025). To our knowledge, this pilot study is the first evaluated DBT-based intervention designed specifically to align with neurodivergent identity, values and the social model of disability. A summary of the adaptations and reconceptualisations of the NDA-DBT programme is presented in Table 1.
NDA-DBT Program Reconceptualisation: Summary Table.
ABA= applied behaviour analysis; ADHD= attention deficit hyperactivity disorder; NDA-DBT= Neurodiversity-Affirming Dialectical Behaviour Therapy.
This pilot study aimed to assess the preliminary impact of NDA-DBT in improving emotion regulation, psychological distress, mindfulness and quality of life in autistic adults. It was hypothesised that preliminary results would show emotion regulation and psychological distress would decrease, and mindfulness skills and quality of life would increase after completing the 10-week NDA-DBT informed group skills training programme for autistic adults.
Method
Participants
The pilot study population consisted of autistic adults aged 28 to 47 years old (M = 35.9, SD = 7.1), 71% identified as female, 21% non-binary and 7% male, who identify as autistic, either through a formal diagnosis or self-identification. Participants were referred by their treating clinicians and/or general practitioners and were required to have a Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R; Ritvo et al., 2011) score > 65. The clinicians collected data before each NDA-DBT group commenced and again after each group was concluded. Data were collected from 15 participants; one participant was excluded because of complications with co-occurring conditions outside the scope of this study, resulting in a final sample size of 14 participants. The final sample size following data screening is detailed in the Data Screening section.
Measures
Ritvo Autism Asperger Diagnostic Scale-Revised
The Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R), an 80-item questionnaire developed for autistic adults aged 18 years and over with average or above-average intellectual abilities, was administered to support the identification of autism-related traits among participants. The RAADS-R assesses social relatedness challenges, circumscribed interests, language and sensory motor subscales. Responses were recorded on a four-point Likert scale ranging from 0 (Never true) to 3 (True now and when I was young). Scores ranged from 0 to 240, with scores of 65 or above linked to autism-related traits. None of the non-autistic individuals exceeded this threshold, confirming a specificity of 100%. Furthermore, only 3% of participants in the autistic group failed to meet or surpass this cut-off, leading to a sensitivity of 97% (Ritvo et al., 2011). The RAADS-R is a psychometrically sound, unidimensional tool, with Sturm et al. (2024) reporting no systematic bias by age, diagnosis/identity or gender. The RAADS-R also exhibited high test–retest reliability, with a coefficient of 0.987, and demonstrated substantial concurrent validity (96%; Ritvo et al., 2011).
Adult ADHD Self-Report Scale V1.1
The Adult ADHD Self-Report Scale V1.1 (ASRS v1.1), an 18-item self-report instrument, was developed to assess traits associated with ADHD in adults aged 18 years and older. The questionnaire aligns with the diagnostic criteria outlined in the DSM-IV and DSM-5-TR, with particular attention given to the expression of traits relevant to adult populations (Kessler et al., 2005). Responses were recorded on a five-point Likert scale ranging from 0 (Never) to 4 (Very Often). The results were categorised into parts A and B, total score and subscale scores. This study utilised part A scores only, as they are the most predictive of ADHD diagnosis (Kessler et al., 2007). Part A scores were calculated from items 1 to 6, with scores ranging from 0 to 24. Part A scores of 14 or above indicate that the profile aligns with a DSM-5-TR ADHD diagnosis in adults, and is considered clinically significant (Adler et al., 2006; Kessler et al., 2007). Fusar-Poli et al. (2024) examined the internal consistency of the ASRS v1.1 in autistic participants and reported Cronbach's alpha of 0.76 for the total score, indicating acceptable reliability for this population. Given the estimated co-occurrence of ADHD in autistic individuals ranges from 28% to 38.5% (Lai et al., 2019; Rong et al., 2021), this measure was used to assess what impact this co-occurrence might have on the effectiveness of the NDA-DBT.
Toronto Alexithymia Scale
The 20-Item Toronto Alexithymia Scale (TAS-20; Bagby et al., 1994) is an extensively used self-report measure designed to identify issues related to alexithymia, such as difficulty identifying and describing feelings and the degree to which individuals tend to prioritise attention on external events rather than internal experiences. The participants responded on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The total score ranged from 20 to 100.
Higher scores indicate greater difficulty identifying, describing and processing emotions. The ranges for clinical significance (Bagby et al., 1997) include no alexithymia (20–51), possible alexithymia (52–60) and alexithymia (61–100). The TAS-20 has demonstrated strong psychometric properties across various populations, including the general public and clinical groups (Bagby et al., 2020). This measure was used to assess the potential impact of this co-occurrence on the effectiveness of NDA-DBT.
Camouflaging Autistic Traits Questionnaire
The Camouflaging Autistic Traits Questionnaire (CAT-Q) is a 25-item self-report measure used to assess engagement in camouflaging strategies for navigating social interactions (Hull et al., 2019). The scale measures masking behaviours, compensation and assimilation. It provides an overall camouflaging score as well as three subscale scores, with responses recorded on a seven-point Likert scale ranging from ‘strongly disagree’ to ‘strongly agree’. The total of all items provided the total camouflaging score, with higher scores indicating higher levels of self-reported camouflaging. Possible scores ranged from 25 (no endorsement of camouflaging strategies) to 175 (strong endorsement of camouflaging strategies). The psychometric properties of the CAT-Q demonstrate its high reliability and validity for use in autistic populations (Hull et al., 2019; Hull et al., 2021). The CAT-Q was administered to examine whether masking might impact the effectiveness of the NDA-DBT.
Difficulties in Emotion Regulation Scale
The Difficulties in Emotion Regulation Scale (DERS) is a 36-item self-report instrument designed to assess clinically relevant difficulties in emotion regulation. Its integrative theoretical framework conceptualises emotion regulation as comprising emotional awareness and understanding, emotional acceptance, goal-directed behaviour, impulse control during negative emotions and access to effective regulation strategies (Gratz & Roemer, 2004). Participants were asked to rate their agreement to statements on a five-point Likert scale ranging from 1 (Almost Never) to 5 (Almost Always) such as ‘When I’m upset, I become angry with myself for feeling that way’ and ‘When I’m upset, I have difficulty concentrating’. The total raw scores range from 36 to 180, with higher scores indicating more significant difficulties in emotion regulation. The DERS also includes six subscales and has demonstrated strong psychometric properties across diverse populations. Good-to-excellent internal consistencies for the DERS subscales were reported for autistic adults by McVey et al. (2022) in their study of 156 autistic adolescents, with Cronbach's alpha values ranging from 0.80 to 0.90 and McDonald's Omega values from 0.84 to 0.91. Construct validity analyses revealed moderately strong to strong positive associations between the DERS subscales and measures of anxiety and depression, suggesting that the DERS effectively captured emotion regulation difficulties in this population. However, the awareness subscale was not correlated with anxiety or depression, but showed a moderate correlation with alexithymia, indicating that it may measure a related but distinct construct. In this pilot study, due to the limited sample size, we opted for the total DERS score for simplicity and to detect general response trends. While acknowledging potential issues with unidimensionality in autistic populations, this exploratory approach allows us to identify overall trends in response to treatment. Future research is planned to investigate specific subdomains for a more nuanced understanding of changes in emotion regulation among autistic adults.
Depression Anxiety and Stress Scale
The Depression Anxiety and Stress Scale (DASS-42) (Lovibond & Lovibond, 1995) is a 42-item self-report scale created to measure the intensity of three negative emotional states (depression, anxiety, and stress) in adults and adolescents aged ≥ 17 years. The Depression Scale evaluates dysphoria, hopelessness, devaluation of life, lack of interest or involvement, anhedonia, inertia and self-deprecation. The Anxiety scale examines autonomic arousal, skeletal musculature effects, situational anxiety and the subjective experience of anxious feelings. The Stress Scale assesses difficulty relaxing, nervous arousal and being easily agitated, irritable, or impatient. Responses were recorded on a four-point scale, ranging from 0 (Never) to 3 (Almost Always), to rate the degree to which they had experienced each symptom over the past week. The DASS-42 total scores are interpreted as normal (0–32), mild (33–39), moderate (40–49), severe (50–57) and extremely severe (58–126). The DASS has been reported to be a viable self-report measure for assessing depression, anxiety and stress in autistic individuals without intellectual disabilities (Nah et al., 2018; Park et al., 2020).
Five Facet Mindfulness Questionnaire
The Five Facet Mindfulness Questionnaire (FFMQ-15) is a self-report tool designed to measure mindfulness related to thoughts, experiences and daily actions. It assesses five mindfulness subscales: observing, describing, acting with awareness, non-judgment and nonreactivity (Baer et al., 2012). The questionnaire consisted of 15 items that measured levels of mindfulness, and participants responded to statements on a five-point Likert scale ranging from 1 (Never or very rarely true) to 5 (Very often or always true). The items included in the FFMQ-15 were derived from the original extensive FFMQ-39. The validity of the FFMQ-15 was supported by a study of 538 university students (Kim et al., 2023). However, since the reliability and validity of the FFMQ-15 remain unestablished in an autistic population, and considering the limited sample size of this pilot study, the total FFMQ-15 score was used.
World Health Organisation Quality of Life-Brief
The World Health Organisation Quality of Life-Brief (WHOQOL-BREF) is a 26-item self-report measure that assesses quality of life across four domains: physical health, psychological health, social relationships and environment (The WHOQOL Group, 1998). Higher scores indicate a better quality of life. Responses are given on a five-point Likert scale, which varies throughout the questionnaire, from ‘Not at all’ = 1 to ‘Extremely’ = 5. The WHOQOL-BREF has demonstrated strong psychometric properties for use with autistic adults (McConachie et al., 2018). The measure also demonstrated strong convergent and discriminant validity, indicating that the WOHQOL-BREF is a reliable and valid tool for autistic adults (McConachie et al., 2018).
Intervention
The DBT skills modules were delivered over 10 weeks in 90 min sessions. Modules were systematically adapted for autistic adults by integrating neuro-affirmative language, increasing the use of concrete and explicit communication, and incorporating sensory, interoceptive, and somatic regulation strategies throughout content and delivery. Specific adaptations also included the reconceptualisation of traditional DBT skills to recognise autistic communication styles, validate neurodivergent lived experience, and prioritise self-compassion and acculturation approaches over behavioural conformity.
Design and Analysis
This pilot study used a longitudinal pre–post design with two time points to examine the preliminary impact of NDA-DBT on emotion regulation, psychological distress, mindfulness, and quality of life. Participants completed the baseline (T1) and immediate post-intervention (T2) assessments. Researchers employed a within-subjects approach, and the statistical analyses included paired samples t-tests and Wilcoxon signed-rank tests (due to the limited sample size) to assess changes over time. We interpreted effect size as follows: d = 0.2 (small), 0.5 (medium) and 0.8 (large; Cohen, 1988). We analysed all data using Jamovi v.2.3.28.0.
Data Screening
Before analysis, we examined the data for missing values, normality, and outliers and undertook a missing data analysis, which indicated that values were missing at random. Little's MCAR test could not be conducted due to insufficient patterns of missingness. Therefore, we applied pairwise deletion (excluding case-by-analysis) to allow cases to be retained in analyses where relevant data were available. Consequently, sample sizes varied across tests, with n = 12 for DERS, n = 11 for DASS-42, n = 11 for FFMQ-15 and n = 10 for the WHQOL-BREF domains. We identified two outliers in the data, as assessed by inspecting boxplots for values greater than 1.5 box lengths from the edge of the box for the WHOQOL-BREF-Psychological Health domain. These were examined and retained due to the expected variation in the experiences of autistic adults. Normality was violated (Shapiro–Wilk's test of normality p = .005) for WHOQOL-BREF-Psychological Health and (Shapiro–Wilk's test of normality p = .008) for WHOQOL-BREF-Social. The assumption of normality was not violated for the remaining measures, as determined by inspection of histograms and the Shapiro–Wilk test of normality (DERS, p = .709; DASS-42, p = .103; FFMQ-15, p = .566; WHOQOL-BREF-Physical Health, p = .801; and WHOQOL-BREF-Environment, p = .112).
Results
Participant characteristics are detailed in Table 1. The mean RAADS-R score for the group (M = 158, SD = 21.20) indicated that the group's average traits were approximately 2.5 times higher than the diagnostic threshold of 65, indicating pervasive autistic traits across domains (social relatedness problems, circumscribed interests, language and sensorimotor). The mean ASRS score for the (n = 12) group (M = 15.8, SD = 4.90) indicated high ADHD characteristics, with 75% of the participants recording ASRS scores consistent with an ADHD diagnosis. Camouflaging of autistic traits was observed in all participants who completed the CAT-Q (n = 12) with 67% scoring extremely high and 33% scoring high. Alexithymia was reported in 67% of the participants who completed the TAS (n = 12).
The researchers ran one-tailed paired sample t-tests to determine whether post-treatment scores were significantly lower for emotion regulation and psychological distress, and significantly higher for mindfulness. Additionally, although the Shapiro–Wilk test suggested normality, excluding WHOQOL-BREF-Psychological Health and Social Relationships, Wilcoxon signed-rank tests were also performed as a robustness check due to the limited sample sizes (n = 12 for DERS, n = 11 for DASS-42, n = 11 for FFMQ-15 and n = 10 for WHOQOL-BREF).
We found a difference in emotion regulation (DERS) before (M = 125.75, SD = 23.58) and after (M = 104.08, SD = 21.53) NDA-DBT, with a statistically significant mean decrease of 21.67, 95% CI [15.23, ∞], SE = 3.56, t(11) = 6.04, p=<.001, d = 1.75. The Wilcoxon signed-rank test also revealed a significant decrease between pre-(Mdn = 125) and post-treatment (Mdn = 98.5) emotion regulation difficulties (W = 78, p = .001, rb = 1.00). Similarly, a difference in psychological distress (DASS-42) before (M = 72.64, SD = 16.85) and after (M = 56.36, SD = 15.03) NDA-DBT was found, with a statistically significant mean decrease of 16.28, 95% CI [7.63, ∞], SE = 4.77, t(10) = 3.41, p = .003, d = 1.03. The Wilcoxon signed-rank test also revealed a significant decrease between pre-(Mdn = 76) and post-treatment (Mdn = 64) psychological distress (W = 45, p = .005, rb = 1.00). Two participant pairs had identical pre- and post-treatment scores and were excluded from the Wilcoxon analysis.
We also found an increase in mindfulness (FFMQ-15) scores from pre- (M = 2.52, SD = 0.42) to post-treatment (M = 2.92, SD = 0.43), t(10) = 3.02, p = .006, d = 0.91. The mean difference (M = 0.40, 95% CI [0.16, ∞], SE = 0.13) reflected improved post-treatment scores. The Wilcoxon signed-rank test also revealed a significant increase in pre-(Mdn = 2.40) and post-treatment (Mdn = 2.93) mindfulness skills (W = 60, p = .009, rb = −0.82). A paired samples t-test indicated that the increase in WHOQOL-BREF-Physical Health scores from pre-treatment (M = 17.70, SD = 4.11) to post-treatment (M = 25.20, SD = 18.31) was not statistically significant, t(9) = 1.35, p = .106, d = 0.43. The mean difference was 7.50 (95% CI [−2.72, ∞], SE = 5.57), suggesting an improvement in physical health quality of life scores following treatment. In contrast, a Wilcoxon signed-rank test revealed a statistically significant increase from pre-treatment (Mdn = 18.50) to post-treatment (Mdn = 20.50), W = 40, p = .022, rb = 0.78. One participant pair had identical pre- and post-treatment scores and was excluded from the Wilcoxon analysis.
An increase in WHOQOL-BREF-Social Relationships scores from pre-treatment (M = 7.90, SD = 3.57) to post-treatment (M = 8.50, SD = 3.41), t(9) = 2.71, p = .012, d = 0.86 was found. The statistically significant mean difference was 0.60 (95% CI [0.20, ∞], SE = 0.22), indicating improved social relationships following the intervention. The Wilcoxon signed-rank test also yielded a statistically significant increase from pre- to post-treatment, W = 15, p = .024, rb = 1.00. However, it is important to note that five out of 10 participant pairs had identical pre- and post-treatment scores and were therefore excluded from the Wilcoxon analysis, substantially reducing the effective sample size. As a result, the robustness of the Wilcoxon test result is limited, and the finding should be interpreted with caution.
An increase in WHOQOL-BREF-Psychological Health scores from pre-treatment (M = 13.10, SD = 3.78) to post-treatment (M = 15.50, SD = 4.17), t(9) = 2.06, p = .035, d = 0.65 was also found. The statistically significant mean difference was 2.4 (95% CI [0.26, ∞], SE = 1.17), indicating improved psychological health following the intervention. The Wilcoxon signed-rank test also yielded a statistically significant increase from pre- to post-treatment, W = 20, p = .003, rb = 0.91. However, it is important to note that four out of ten participant pairs had identical pre- and post-treatment scores and were therefore excluded from the Wilcoxon analysis, substantially reducing the effective sample size. As a result, the robustness of the Wilcoxon test result is limited, and the finding should be interpreted with caution.
Finally, while post-treatment scores were higher than pre-treatment scores for WHOQOL-BREF environment (pre-treatment, Mdn = 25.00, post-treatment, Mdn = 27.50, W = 30, p = .202) the t-test and Wilcoxon signed-rank tests were not significant.
Bonferroni correction was applied to control for multiple comparisons. As seven tests were conducted, the adjusted significance threshold was set at p < .007 (α=0.05/7). After applying Bonferroni correction, physical health, social relationships and psychological health were no longer significant; therefore, these results should be interpreted cautiously. The DERS, DASS-42, and FFMQ-15 (t-test only) results remained significant. Despite the observed effect being statistically significant, the infinite upper bound of the confidence interval for the DERS and DASS-42 scores and the infinite lower bound for the FFMQ-15 and WHOQOL-BREF indicate extreme uncertainty in the estimate, suggesting that the results may be unstable or overestimated because of the limited sample size.
Discussion
This study aimed to explore the preliminary impact of NDA-DBT on emotion regulation, psychological distress, and mindfulness skills in a sample of autistic adults. This was the first study to examine a neurodiversity-affirming adaptation of DBT-informed skills groups. Understanding the impact of neurodiversity-affirming therapies on mental health outcomes for autistic adults could provide greater insight into the specific needs of neurodivergent populations and allow for the development of therapies that support the unique abilities and skills of different neurotypes.
Although this pilot study had a limited sample size, preliminary support for our hypothesis that there would be a significant decrease in emotion regulation difficulties in autistic adults after completing NDA-DBT skills training was reinforced. These results are consistent with those of Bemmouna et al. (2022), who reported a significant decrease in emotion regulation difficulties in autistic adults exhibiting self-harm and suicidal behaviours, and Phillips et al. (2024), who reported a reduction in emotion regulation difficulties in autistic adolescents. Although these studies delivered interventions based on the four components of comprehensive DBT, the current study's adapted skills training-only intervention suggests similarly positive results and effect sizes found by Bemmouna et al. (2022). Despite its known limitations, DBT is recognised as a gold standard, evidence-based, transdiagnostic intervention for supporting emotion regulation (Bemmouna et al., 2022). Our preliminary findings suggest that, when DBT is adapted within a neurodiversity-affirming framework that respects and values autistic identity and lived experience, it could potentially be a suitable and supportive approach for autistic adults experiencing emotion regulation challenges. The second hypothesis that there would be a significant decrease in psychological distress in autistic adults after completing the NDA-DBT informed skills training was also supported by our preliminary findings. This is somewhat consistent with Cornwall et al. (2021), who reported preliminary findings that radically open DBT significantly reduced global distress in autistic adults. Similarly, Huntjens et al. (2024) reported a significant decrease in depression severity, but no significant decrease in social anxiety, for autistic adult outpatients exhibiting suicidal behaviour, and Neal (2023) reported that four of six autistic adults who completed a DBT skills programme reported significant decreases in depression, anxiety and stress. Currently, there is limited research assessing psychological distress as an outcome measure when considering the effectiveness of DBT in autistic adults. However, significant reductions in anxiety, but not depression, have been reported post-DBT skills training for anxious and/or depressed, non-BPD adults with high emotion dysregulation (Neacsiu et al., 2014). The inconsistent findings may result from the various co-occurring conditions or sources of distress among participants in different studies, or from the limited sample sizes, which may affect the validity of the results.
Our hypothesis that mindfulness skills would increase significantly after NDA-DBT treatment was somewhat supported. While mindfulness skills have not been examined in studies assessing the effect of DBT on autistic adults, significant increases in mindfulness skills have been reported in post-DBT skills training for adults with BPD and severe emotion dysregulation (Kells et al., 2020). Similarly, a scoping study assessing mindfulness outcomes across DBT studies found increases in the level of mindfulness, particularly for the facets of accepting without judgment/non-judging and non-reactivity (reduced impulsivity) post DBT intervention (Eeles & Walker, 2022). These preliminary findings contribute to the small but growing body of literature on the effectiveness of DBT skills training for autistic adults. Notably, this is the first study to examine a neurodiversity-affirming adaptation of DBT skills for autistic adults.
Our final hypothesis, that the quality of life would significantly increase after NDA-DBT treatment, was partially supported, with improvements found in the psychological and social domains, but not in the physical and environmental domains. However, after applying the Bonferroni correction to control for multiple comparisons, no quality of life domains were found to be significant. Our findings, while not significant, showed a reported increase in scores across all domains indicating some support for Huntjens et al. (2024), who found that comprehensive DBT significantly improved overall quality of life, as measured by the Manchester Short Assessment of Quality of Life (MANSA), for autistic adults with suicidal behaviour compared to treatment as usual.
From a theoretical perspective, the preliminary findings of this pilot study may be viewed as consistent with the neurodiversity paradigm (Singer, 1999; Walker, 2021), which recognises autism as a valid form of neurocognitive diversity. This paradigm emphasises that autistic people are likely to benefit from tailored, affirming supports that respect individual strengths and respond to the unique barriers encountered in a predominantly neurotypical society (Milton, 2012; Sasson et al., 2017). The current results also appear to align with Linehan's (1993a) biosocial theory of emotion dysregulation, which posits that emotional distress can emerge from the interaction between biological sensitivities and environments that do not accommodate individual needs. This framework may be particularly relevant for autistic individuals, who frequently describe experiences of social misunderstanding, sensory invalidation, and external pressures to mask or suppress their authentic selves. The observed value of a neurodiversity-affirming adaptation of DBT in this study tentatively suggests that prioritising validation, supporting authentic self-expression and fostering self-acceptance within a structured therapeutic context may contribute to enhanced emotional wellbeing for autistic adults.
Limitations and Future Recommendations
Several methodological factors limit the interpretation of this study's findings. Violations of normality and tied pairs reduced the reliability of statistical analyses, while the small sample size increased the risk of sampling bias and reduced generalisability. The absence of a control group and the possible effect of confounding variables mean that improvements cannot be conclusively linked to the NDA-DBT intervention. Participants presented with diverse co-occurring conditions and clinical profiles (Table 2), which may have influenced outcomes independently of the NDA-DBT intervention.
Participant Gender, Age and Levels of Autistic Characteristics, ADHD Characteristics, Camouflaging, and Alexithymia.
ADHD= attention deficit hyperactivity disorder; PTSD= posttraumatic stress disorder; CPTSD= complex posttraumatic stress disorder.
Additionally, the intervention was limited to the DBT-informed skills training component, rather than the full DBT protocol, and the programme duration was restricted to 10 sessions due to financial and Medicare constraints. Programme developers aim to lengthen the sessions to a minimum of 16 weeks.
Future research should address these limitations by recruiting larger samples, including control or comparison groups, using covariate analyses, and collecting qualitative data. To strengthen internal validity, future studies should employ mixed-methods approaches to capture both quantitative outcomes and participants’ lived experiences. Statistical control for relevant covariates is recommended to isolate intervention effects more precisely. Moreover, the selection of outcome measures should prioritise instruments validated for use with autistic populations to ensure sensitivity and relevance. Implementation of the full DBT model and extension of programme duration are recommended. Longitudinal designs and exploration of the relationship between emotion regulation and psychological distress will further clarify intervention effects and inform programme adaptation for autistic adults.
It is also important to acknowledge that there is a paucity of research on screening instruments for autistic adults that are considered neurodiversity-affirming, as such the instruments used in this research may not be considered neurodiversity-affirming by the autistic community.
Longitudinal research should assess the impact of treatment over time on emotion regulation, psychological distress, mindfulness and quality of life. The relationship between emotion regulation and psychological distress should also be examined to better understand the mediating role of emotion regulation in further enhancing future programme adaptation. Based on the high levels of masking behaviours and alexithymic characteristics observed in the sample, future research should explore how these may change post-NDA-DBT. Finally, qualitative feedback from autistic clients should be included in future research to enhance internal validity by providing a deeper understanding of participants’ experiences, perceptions and contextual factors that may influence outcomes.
Conclusions
The preliminary results of this pilot study tentatively suggest that NDA-DBT may influence the reduction of emotion regulation difficulties and psychological distress, and possibly increase mindfulness. While the findings align with the limited existing literature, further research is needed to establish causality and investigate the effects of neurodiversity-affirming adaptations in more detail. However, given the challenges and obstacles faced by autistic adults in accessing appropriate psychological treatments, this vital research has contributed to existing knowledge by showing that the DBT-informed group skills training programme format, appropriately adapted to be neurodiversity affirming, has the potential to be used as a standalone intervention for autistic adults. It has also extended the limited knowledge on adapted DBT treatments informed by neurodiversity affirming frameworks for autistic adults.
Footnotes
Ethical Approval Statement
Ethical approval was granted by The Cairnmillar Institute Human Research Ethics Committee (approval number: 2024041602).
Author Contributions
All authors contributed to the written work presented in this article and approved of the final manuscript. JAS conceptualised this project, data collection, critically edited and contributed to the drafting of this manuscript. MS conceptualised the specific questions and hypotheses and drafted the manuscript. . MS undertook the analysis under supervision of JAS. All authors were involved in the conceptual design of the study and the review of the study findings.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The datasets generated during and/or analyzed during the current study are not publicly available due to confidentially and participants have not consented to data release.
Anonymize Data
Cairnmillar Institute Human Research Ethics Committee (approval number: 2024041602)